quarterly bulletin - 2015 q4

Quarterly Bulletin2015 Q4 | Volume 55 No. 4

Quarterly Bulletin2015 Q4 | Volume 55 No. 4

Topical articles

Bonus regulation: aligning reward with risk in the banking sector 322Box The ‘agency problem’ 324Box European Banking Authority Regulatory Technical Standard and its link to the new deferral rules and the Senior Managers Regime 327

The Prudential Regulation Authority’s secondary competition objective 334Box Competition responsibilities in the regulation of UK financial services 336Box The relationship between competition and stability in the banking sector 340

Trends in UK labour supply 344Box A brief description of equilibrium unemployment theory 351

The potential impact of higher interest rates and further fiscal consolidation on households: evidence from the 2015 NMG Consulting survey 357Box Survey method 359

Recent economic and financial developments

Markets and operations 370Box Developments in swap spreads 372Box Risk reversals 375Box Cross-currency basis swaps 377

Report

BoE-HKMA-IMF conference on monetary, financial and prudential policy interactions in the post-crisis world 382

Appendices

Contents of recent Quarterly Bulletins 386

Bank of England publications 388

Contents

The contents page, with links to the articles in PDF, is available atwww.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx

Author of articles can be contacted [email protected]

Except where otherwise stated, the source of the data used in charts and tables is the Bank of England or the Office for National Statistics (ONS). All data, apart from financialmarkets data, are seasonally adjusted.

Research work published by the Bank is intended to contribute to debate, and does notnecessarily reflect the views of the Bank or members of the MPC, FPC or the PRA Board.

www.flickr.com/bankofengland

www.youtube.com/bankofenglanduk

www.twitter.com/bankofengland

Topical articles

Quarterly Bulletin Topical articles 321

322 Quarterly Bulletin 2015 Q4

• Remuneration policies in the banking sector incentivised excessive risk-taking, therebycontributing to the financial crisis. Since the crisis, remuneration rules have come into force tobetter align employees’ incentives with the long-term health of banks and the financial system.

• This article explains the key components of the Prudential Regulation Authority’s remunerationrules for banks and considers the direction of the global policy agenda.

Bonus regulation: aligning reward withrisk in the banking sectorBy Marilena Angeli and Shahzad Gitay from the Bank’s Cross-Sectoral Policy Division.(1)

(1) The authors would like to thank Christopher Gynn for his help in producing this article.

Overview

In the years leading up to the financial crisis, bonus paymentsmore than doubled in the finance and insurance industry —significantly outpacing growth in overall spending in theeconomy (shown by nominal GDP in the summary chart).There is broad consensus that banks’ remuneration policieswere a contributing factor to the financial crisis. Thisincluded rewarding high short-term profits with generousvariable remuneration awards (like bonuses) by encouragingexcessive risk-taking that did not consider the long-term riskscreated for banks and, ultimately, wider society. The globalregulatory response sought to align incentives with riskstaken. In the United Kingdom, the Bank of England’sPrudential Regulation Authority has implemented rules toachieve this with a view to improving the safety andsoundness of individual firms and the resilience of thefinancial system.

Requiring the size of awards to employees to be determinedby a balanced suite of metrics, and requiring awards to bepaid at least 50% in non-cash instruments which tie the sizeof the award to the longer-term performance of the firm,should help to align incentives from the outset. Deferringawards means they can be adjusted to reflect longer-termrisk horizons. From 2016, deferral periods of up toseven years from the date of payment will apply. Throughreduction of unvested awards (‘malus’) and reclaiming vestedawards (‘clawback’), ex-post adjustments can be made up toten years from the initial date of award. Finally, variableremuneration payments must not limit firms’ abilities tostrengthen their capital bases.

Looking ahead, remuneration remains on the agenda forregulators around the world. As Governor Carney said inNovember 2014, ‘Senior manager accountability and new

compensation structures will help rebuild trust in financialinstitutions’. One consequence of the regulation ofremuneration, particularly the introduction in theEuropean Union of the bonus cap, has been an increase infixed remuneration as a proportion of total remuneration.As with excessive variable remuneration without appropriateincentives, this can also impact negatively on resiliencewithin the financial system. Higher fixed pay limits theproportion of total remuneration that can be used to absorblosses in a downturn and that which is aligned to long-termrisks. The global regulatory focus is on the need to ensure asufficient portion of remuneration remains ‘at risk’ of beingreduced or eliminated and on the role of incentives inreducing misconduct.

Click here for a short video that discusses some of the keytopics from this article.

Summary chart Timeline of UK bonus payments

0

50

100

150

200

250Indices: 2003 = 100

Bonus payments in finance and insurance(b)

Nominal GDP(c)

2003 04 05 06 07 08 09 10 11 12 13 14

(a)

Sources: ONS and Bank calculations.

(a) From 2007–08 there was a large drop in the value of the FTSE 100 UK banks.(b) See footnote (b) on Chart 1.(c) GDP at current prices on a financial-year basis.

https://youtu.be/k29LK7jGRUM

Topical articles Bonus regulation: aligning reward with risk 323

Introduction

In the years leading up to the financial crisis, the quantities ofvariable remuneration paid out within the UK financial sectorincreased substantially with a similar trend seen elsewhere, forexample in the United States (Chart 1). There is broadconsensus that, particularly in banks, remuneration policieswere one of the contributing factors to the financial crisis.Lord Turner, former Chair of the Financial Services Authority(FSA), has highlighted the role that inappropriate incentivestructures played in encouraging imprudent behaviour. TheTreasury Select Committee’s report on the banking crisisargued that, in too many cases, ‘the design of bonus schemesin the banking sector were flawed and not aligned with theinterests of shareholders and the long-term sustainability ofthe banks’. The report criticised the bonus-drivenremuneration structures prevalent in the City of London andother financial centres that led to ‘reckless and excessiverisk-taking’.

The perceived role of remuneration in incentivising excessiverisk-taking and undermining sound and effective riskmanagement brought remuneration onto internationalregulatory agendas. This article explains the broad types ofremuneration that are currently used by firms, and some ofthe ‘incentive effects’ to which they give rise. It explains theinternational and domestic remuneration rules that have beenbrought into force since the crisis with the objective of betteraligning the incentives of individuals with the risks taken.Finally, the article describes how global regulators are workingto ensure the effective implementation of these reforms andstrengthen them further going forwards. Click here for a shortvideo that discusses some of the key topics from this article.(1)

What is variable remuneration?

Employees’ remuneration packages can be broken down intotwo parts. Fixed remuneration is the contracted, fixed inadvance amount, which is paid to an employee for their workas a reflection of their professional experience and theresponsibilities associated with their role. Variableremuneration includes any remuneration awarded in additionto the fixed element, which varies according to some measureof performance (the main example being bonuses).

The primary purpose of variable remuneration is, therefore, toincentivise performance, but there are other reasons why firmsaward variable remuneration. Having a substantial proportionof remuneration that can be varied provides greater flexibilityin the management of banks’ costs. Variable pay may also beawarded due to the expectations of individuals working in thebanking sector — particularly those working in investmentbanking. Deferred variable remuneration, where payment ofthe award is delayed for a set time period, may also play a rolein staff retention. Furthermore, variable remuneration canhelp to address the ‘agency problem’ by enabling a betteralignment between the interests of risk-takers within firmsand their shareholders, and between the interests ofshareholders and wider society. The box on page 324 explainsthis in more detail.

Variable remuneration can consist of a combination ofbonuses and long-term incentive plans (LTIPs). Bonuses areusually delivered at the end of the firm’s reporting year and,by reflecting the performance over the year, are used toincentivise the delivery of targets and/or assess employees’performance against a range of financial and non-financialmetrics. They are therefore based on prior performance over arelatively short period. LTIPs, by contrast, seek to align theinterests of the employee with the long-term interests of afirm, and performance is determined with reference toforward-looking metrics assessed over a multi-yearframework. These metrics effectively act as a ‘scalar’ to theinitial award value: if defined targets are not met, then theinitial value of the award will be adjusted downwards.Exceeding or meeting targets would result in the full value ofthe initial award remaining. The size of any adjustment willdepend on the outturn of performance compared to thetarget.

For the finance and insurance industry, variable remunerationreached a peak of 34% of total remuneration in 2006(Chart 2). Despite falling back following the onset of thefinancial crisis, it has continued to be a sizable proportion, onaverage comprising one quarter of total remunerationbetween 2007 and 2013, before falling to 20% of total

100

0

50

150

200

250Indices: 2003 = 100

UK finance and insurance(b)

New York securities industry(c)

2003 04 05 06 07 08 09 10 11 12 13 14

(a)

Sources: Office of the State Comptroller (OSC), ONS and Bank calculations.

(a) From 2007–08 there was a large drop in the value of the FTSE 100 UK banks. See theHouse of Commons Treasury Committee (2009).

(b) Finance and insurance industry as defined by the ONS. Bonus payments paid during thefinancial year, figures derived from the ONS Labour Market Statistics release. The base yearis the 2003/04 financial year.

(c) Data derived from the OSC estimations of bonus pools in the New York securities industry.The base year is set for year-end 2003.

Chart 1 Bonus payments across the UK finance andinsurance industry and bonus pools in the New Yorksecurities industry

(1) https://youtu.be/k29LK7jGRUM.


remuneration in 2014. For most companies in the economy,by contrast, pay largely comprises of fixed remuneration andonly a relatively small proportion of total pay — less than 5%— is variable.

Regulating remuneration

This section begins by explaining how the regulation ofremuneration can support resilience in the banking sector andfinancial stability. It looks at how remuneration has featuredin the international and UK regulatory agenda following the

crisis. It then focuses on the main components of the rulesthat have been introduced in the UK financial sector since thecrisis, which include: rules relating to the form that variableremuneration takes (for instance, cash versus non-cashawards); the use of broader performance metrics; longerdeferral periods between when an award is made and when itis paid out; and the conditions under which an award can bescaled back by ‘malus’ or ‘clawback’. The infographic onpage 331 summarises the main changes in the remunerationrules since the onset of the financial crisis.

The ‘agency problem’

This box introduces the agency problem. It then explains thetheory behind the use of certain forms of variableremuneration — such as payment via non-cash instrumentslike shares or debt — which can serve as a solution to thisproblem. The following section of the article discusses theactual changes that have occurred since the crisis in the formthat variable remuneration typically takes.

What is the agency problem?The origin of the ‘agency problem’ (sometimes known as the‘principal-agent problem’) lies in the separation of ownershipand management of the firm. Executive directors (agents) acton behalf of shareholders (principals) in making decisions andcarrying out activities for the firm.

The theory assumes a conflict of interest between a firm’sshareholders and the executive directors. Executive directorsacting as the agents are hired to make decisions that willmaximise shareholder wealth. However, as agents, they maybe tempted to maximise their own wealth. By taking more (orgreater) risks than the shareholders would, they stand toreceive immediate gains (such as high bonuses) but will not beliable for any costs or losses the firm and shareholders maysubsequently incur.

Shareholders may therefore wish to monitor executivedirectors’ actions to ensure they accord with their own riskappetite. In reality, however, constant monitoring is costlyand impractical.

Variable remuneration as a way of aligning employeeand shareholder incentivesA solution would be to use variable remuneration to attemptto better align the executive directors’ risk appetites to thoseof the shareholders. For instance, variable remunerationawarded as shares transfers a proportion of the bank’sownership onto the executive directors. This could increaselong-term performance considerations in executive directors’decision-making, better aligning the time horizons which they

consider with the time horizons over which risks may manifestthemselves.

Shareholders’ incentives and financial stabilityThere are instances, however, when shareholders’ incentivesmay not lead to the strengthening of financial stability. Someshareholders may have a very high appetite for risk. In suchcases, aligning employee incentives with those of shareholderscould lead to risk-taking levels beyond those which thePrudential Regulation Authority considers prudent. Hence, asecondary agency problem arises from the divergence ofincentives between shareholders and the wider society,reflected by financial stability. Payment in debt instrumentsmight provide an alternative way to link variable remunerationto the long-term health of the firm, and, hence, financialstability. The European Banking Authority has issued aregulatory technical standard which provides for the paymentin convertible debt instruments which reflect the credit qualityof an institution as a ‘going concern’ — that is, where a firm isviewed to be operational and profitable for the foreseeablefuture, and meets their regulatory requirements.

In a debt-based award, the future gains are fixed when theaward is made, with adjustments only possible in the form ofreduction — in contrast to an equity award which has thepotential of unlimited upside gain should the price of theequity increase. Debt instruments could, for example, bestructured to automatically absorb losses when a firm’s capitallevel falls below a certain threshold. The awards could beconverted to equity to help ensure that the capital returned toabove the breached threshold. It is, therefore, sometimesargued that payment in debt instruments better aligns theincentives of bank employees with those of widerstakeholders, in particular the firm’s creditors. However, todate, there has been very limited use of such instruments.This reflects the fact that institutions find them costly todevelop and implement, they are not popular withshareholders (because of their large coupon payments, forexample), their lack of transparency and certainty, and the factthat eliminating upside gains can make them unpopular withrecipients.


How can regulators ensure that remunerationsupports financial stability?Following the financial crisis there was considerable publicfrustration about remuneration in the banking sector. Asdiscussed at the recent Open Forum event held by the Bank, inorder for the markets to maintain their ‘social licence’, theyneed to operate in a fair and accountable way, working withthe interests of society in mind.(1) In regulating remuneration,the Prudential Regulation Authority (PRA) helps to maintainthis social licence: the aim is to better align risk and reward byencouraging good risk management and discouragingexcessive risk-taking, including via the deferral of a proportionof variable pay. As such, it is intended to contribute to abetter level of resilience within banks and therefore supportfinancial stability.

A lot of the public’s frustration centred on the overall level ofbankers’ bonuses and pay. The PRA is interested in the totallevel of pay awarded, in as far as it seeks to ensure that it isconsistent with retaining an appropriate level of capital withinthe bank in question. Effective competition within the bankingsector and the labour market will determine total levels ofpay.

In its ‘Principles and standards for sound compensationpractices’, issued in 2009, the Financial Stability Board (FSB)stated that variable remuneration schemes should be designedto work in concert with overall risk management. The metricsthat determine variable remuneration awards should providesignals of the firm’s risk appetite which, in turn, shouldtranslate into a given level of risk-taking by employees. Thesemetrics should be structured so as to align employeeincentives with the long-term interests of the business whiletaking account of the time frame over which financial riskscrystallise.

Variable remuneration also contributes to the flexibility ofbanks’ staff cost bases. In times of stress, costs can be reducedto help maintain the financial health of the firm. Variableremuneration can thus act as a form of loss-absorbingcapacity for the financial system. The FSB has also said that asubstantial proportion of remuneration should be variable andpaid on the basis of individual, business unit and firmperformance.

The UK remuneration rulesThe FSB Principles & Standards (FSB P&S) were implementedin the United Kingdom through remuneration rules introducedby the FSA in 2009 which took effect from 1 January 2010.(2)

These standards have subsequently been extended throughthe European Union (EU) Capital Requirements Directive(CRD). Figure 1 provides a timeline of the key remunerationregulatory developments for the banking sector in theUnited Kingdom.

The UK remuneration rules, which implement theseinternational requirements, are set and supervised by theBank of England’s PRA and the Financial Conduct Authority(FCA).(3) In June 2013, the report of the ParliamentaryCommission on Banking Standards (PCBS) includedrecommendations to the regulators for strengthening theremuneration rules.(4) The PRA and FCA consulted jointly onnew rules which were published in June 2015.(5) In particularthese put in place tougher requirements for deferral andclawback (which are explained in more detail onpages 328–30).(6)

Whose pay is regulated?The remuneration rules apply to banks(7) on a firm-wide leveland to the variable remuneration of all employees who canhave an impact on a firm’s risk profile, known as materialrisk-takers (MRTs). Firms must identify their populations ofMRTs using the criteria in the regulatory technical standardpublished in 2014 by the European Banking Authority (EBA) toensure consistency of identification across EU jurisdictions.The box on page 327 outlines the identification criteria forMRTs and their link to the new deferral rules and SeniorManagers Regime. The application of this standard has led toapproximately a threefold increase in the number of

0

5

10

15

20

25

30

35

40

Finance and insurance(b)

Rest of the economy(c)

Per cent

(a)

2003 04 05 06 07 08 09 10 11 12 13 14

(a) From 2007–08 there was a large drop in the value of the FTSE 100 UK banks.(b) Finance and insurance industry as defined by the ONS.(c) Rest of the economy as defined by the ONS.

Chart 2 Bonus payments as a percentage of totalremuneration in the United Kingdom

(1) See Open Forum: building real markets for the good of the people, June 2015;www.bankofengland.co.uk/markets/Documents/openforum.pdf.

(2) See Financial Stability Forum (2009) and Financial Stability Board (2009).(3) The PRA is part of the Bank of England and responsible for the prudential regulation

and supervision of around 1,700 banks, building societies, credit unions, insurers andmajor investment firms. The FCA is an independent body and responsible for theconduct regulation in the UK financial services industry and the prudential regulationof firms not covered by the PRA.

(4) See Parliamentary Commission on Banking Standards (2013).(5) See Bank of England and Financial Conduct Authority (2015).(6) The PRA and FCA work together in implementing the regulation of remuneration

rules in the United Kingdom for firms regulated by both authorities and forconsistency in the wider industry. The FCA’s solo initiatives on remuneration arebeyond the scope of this article and further information is available at www.the-fca.org.uk/remuneration.

(7) Banks in this article are defined as the PRA-regulated banks and building societies,and PRA-designated investment firms.


individuals in the major UK banks subject to the remunerationrules(1) (Chart 3).

Composition of variable remuneration: cash andnon-cashThe crisis prompted regulators to consider whether non-cashawards could be used as an effective way to align risk-takingincentives. Prior to the crisis banks were able to award theirbonuses as cash awards at the end of the year. Since 2010, atleast 50% of variable remuneration awards are required to bedelivered in non-cash instruments such as shares or debt,which create incentives aligned with long-term value creationand the time horizons of risk.

The EBA has developed a regulatory technical standard on theclasses of instruments that would qualify. In practice,institutions have predominantly paid the non-cash portion ofvariable remuneration in shares or share-linked instruments.As discussed in the box on page 324, there has been anincreasing focus on the case for payment in debt rather thanequity.

Incentives and performance metricsThe performance measures on which variable remunerationawards are determined should be risk-adjusted and designedto encourage sustainable business practices. A balanced suiteof metrics lessens the likelihood of individual short-termism,and also allows consideration of the bank as a whole ratherthan just bottom-line profit figures. These metrics caninclude, but are not be limited to:

• Risk-adjusted return metrics where risk is calculated as ameasure of the return relative to the risk taken over aspecific period. Some measures used include economicprofit and return on risk-weighted assets.(2)

• Prudential metrics which reflect the financial strength ofthe firm, such as a healthy balance sheet or capital levels asa share of risk-weighted assets.

• Strategic metrics which focus on the forward-lookingdirection of the firm on a financial (such as market growthor cost savings) and non-financial basis (such as investmentin human resources).

• Conduct metrics which reflect behaviours that have theinterests of customers in mind. Measures can includecustomer outcomes and compliance with regulation.

Since the introduction of the rules, the PRA has sought todiscourage firms from determining remuneration using anarrow set of metrics based on non risk adjusted returnmetrics.(3) The PCBS concluded that before and during thefinancial crisis, banks had over-relied on return metrics thatwere not adjusted for the risks taken, in particular return onequity, when setting remuneration. This encouragedindividuals to focus on short-term, leveraged growth ratherthan more sustainable business models. In 2015 the PRA

12 August 2009:The FSA publishesremuneration rules inthe United Kingdom

April/September2009:The FSB publishesits Principles andStandards

1 January 2010:The FSA implements the FSB Principlesand Standards through remunerationrules for the UK-based banks

1 January 2011:CRD III has remuneration requirements applicable toall EU jurisdictions

26 June 2013:CRD IV introduces additionalrules on remuneration

1 January 2014:The new CRD provisions comeinto force including the bonuscap and criteria for MRTs

23 June 2015:The PRA and FCA publishnew rules includingextended deferral andclawback periods

30 July 2014:The PRA introduces a clawback rule for awardsmade from 1 January 2015

19 June 2013:PCBS publishes a reportincluding recommendationsfor longer deferral periods

Figure 1 Timeline of remuneration rules

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

2010 11 12 13 14

Number of individuals

(b)

Sources: Annual accounts and Pillar 3 remuneration disclosures.

(a) Defined as the top five UK-headquartered banks by market capitalisation.(b) Introduction of the EBA’s regulatory technical standard on the identification of material

risk-takers.

Chart 3 Population of material risk-takers in the majorUK banks(a)

(1) Top five UK-headquartered banks by market capitalisation.(2) Economic profit is similar to accounting profit but also has the opportunity costs

deducted from revenues earned. Risk-weighted assets are a bank’s assets or offbalance sheet exposures, weighted according to risk.

(3) A measure of direct financial gain from the business undertaken.


brought in a new rule explicitly to require banks to use metricsbased on return or revenue only as part of a balanced,risk-adjusted scorecard when determining variableremuneration. This formalised the approach required underthe remuneration rules introduced after the crisis.

The performance metrics of executive directors are a usefulmeasure in assessing how incentives have changed over time.Executive directors are responsible for setting strategy (subjectto approval by their Boards) and their impact upon theUK banking sector has been highlighted by the SeniorManagers Regime (SMR), which introduces personalaccountability to their actions. Therefore it is insightful toconsider the personal incentives of executive directors, sincethese will underpin their actions and filter down through theincentives of staff working beneath them.

Chart 4 demonstrates that before and during the crisis,performance metrics were based purely on non risk adjustedreturn measures. Since 2009, the introduction of risk-adjustedreturn and non return based metrics has led to a reduction ofthe extent non risk adjusted return based metrics determineawards, from 100% to 30% for the average executivedirector.(1) This shift away from non risk adjusted returnmetrics, such as return on equity and total shareholder return,to a more balanced set of performance metrics, reflects thechange towards a new framework for setting senior managerincentives which are more in line with the long-term health ofthe firm.

European Banking Authority RegulatoryTechnical Standard and its link to the newdeferral rules and the Senior ManagersRegime

The Regulatory Technical Standard (RTS) sets out criteria forthe identification of categories of staff who have a materialimpact on an institution’s risk profile.

Firms are required to identify the material risk-takers in theirinstitution on an annual basis. To ensure harmonisation acrossthe European Union, there are two types of criteria that mustbe used by firms for identification:

• Qualitative criteria (‘Article 3’ in the RTS) which identifystaff by job roles and specific responsibilities; and

• Quantitative criteria (‘Article 4’ in the RTS) which are basedon total remuneration in absolute terms (staff earning morethan €500,000) and in relative terms (0.3% of staff with thehighest remuneration).

If a firm determines that individuals captured under thequantitative criteria do not have a material impact on its riskprofile, there are processes for their exclusion from theapplication of the RTS. Table 1 sets out requirements forexclusion under the quantitative criteria.

In June 2015, the Prudential Regulation Authority (PRA) andFinancial Conduct Authority (FCA) introduced new rules thatincrease the period of deferral for many material risk-takers(MRTs). The new rules distinguish between the levels ofresponsibility MRTs may have on the firm’s risk profile. ThePRA’s Senior Managers Regime (SMR) requires seniorindividuals responsible for the executive management oroversight of those areas of a firm which the PRA deemsrelevant to its safety and soundness objective to be heldindividually accountable.

As a result the new deferral rules split the MRT population intothree categories, using the SMR and RTS frameworks. Seniormanagers will be subject to seven-year deferral, risk managersto five-year deferral and all other MRTs to three to five-yeardeferral. The rules draw a distinction between risk managersand all other MRTs using the qualitative criteria of theEuropean Banking Authority RTS (Table 2). Any individualsolely caught by the quantitative criteria also falls into theother MRTs category.

Table 1 Exclusion criteria

Criterion Regulatory oversight requirement

Total remuneration €500,000–€750,000 Notification to the PRA and FCA

Total remuneration > €750,000 Prior approval of exclusion from PRA and FCA

Total remuneration > €1,000,000 PRA to inform EBA before approval

Source: European Banking Authority (EBA).

Table 2 New deferral requirements

MRT population Minimum deferral period

Senior managersAll individuals as defined by the SMR. This includes the Seven yearschief executive, chief finance officer and heads of key business areas.

Risk managersAll individuals in risk managing roles; derived from the Five yearsRTS. This includes members of the management body, heads of other material business units and managers of MRTs.

All other MRTsAll MRTs not captured by the categories above. Minimum CRD provision ofExamples include individuals exposing the firm to three to five yearstrading/market risk and individuals approving the introduction of new products.

Source: Prudential Regulation Authority.

(1) The chart analyses metrics in LTIP awards as they feature significantly in total variableremuneration for executive directors (approximately 70% in 2014).


DeferralDeferral of variable remuneration awards (both cash andnon-cash) is a key element of the way in which the rules seekto ensure that longer-term risk horizons are reflected. Bydeferring the payment (or ‘vesting’) of part of an award, thereis an opportunity to reassess the nature, scale and outcomesof the risks taken in order to assess the performance for whichvariable remuneration has been awarded. This is referred to asex-post risk adjustment or ‘malus’ (explained further below).

The current rules which came into force in 2011 stipulateminimum deferral rates that banks with assets of over£15 billion must apply in relation to variable remuneration fortheir MRT staff. These are:

• At least 40% of the total variable remuneration award forthe year to be deferred or 60% for senior executives or iftotal variable remuneration is £500,000 or higher.

• A minimum deferral period of three to five years.

• The payout or vesting of the deferred awards may not befaster than in equal tranches. This is known as ‘pro-ratavesting’.

It should also be noted that these minimum deferral periodsare required under the CRD. The PRA has encouraged banksoperating in the United Kingdom to apply higher deferral ratesto match their risk profiles. From 2016, more seniorrisk-takers will be subject to a minimum of five to seven years’deferral, with no vesting until year three for the most seniormanagers (see below).

The CRD provisions have increased the deferral periods banksapply to executive directors (Chart 5). Before the crisis,

deferral periods ranged from one to three years for executivedirectors. The remuneration rules shifted the range of deferralperiods upwards. Although the rules provided for a minimumof three to five years for deferral, most banks opted to use athree-year deferral period (as shown by the orange diamonds).

Ex-post risk adjustment: malusMalus is the reduction or cancellation of unvested or unpaidawards. Under the rules, firms must be able to make suchreductions, where justified, to take account of risks that havesubsequently crystallised, instances of individual misconductthat have been uncovered, failures in the management of risk oroversight of that part of the business, or a subsequent materialdownturn in the performance of the firm.(1) Malus has beenused increasingly by banks in relation to risk managementfailings that have come to light since 2010 (Chart 6).

0

10

20

30

40

50

60

70

80

90

100

2006 07 08 09 10 11 12 13 14

Non risk adjusted return

Risk-adjusted return

Prudential

Strategic

Conduct

Per cent

Sources: Annual accounts, annual remuneration reports and Bank calculations.

Chart 4 Average weighting of metrics used to determineexecutive directors’ long-term incentive plans in themajor UK banks

0

1

2

3

4

5

6

2007 08 09 10 11 12 13 14

Number of years

(b)Range of deferral applied

Most commonly applied deferral period

Sources: Annual accounts, annual remuneration reports and Bank calculations.

(a) Based on banks’ deferral policies for the annual bonus and LTIP schemes.(b) CRD deferral rules introduced.

Chart 5 Deferral applied to executive directors in themajor UK banks(a)

0

50

100

150

200

250

300

350

2010 11 12 13 14

£ millions

Sources: Pillar 3 remuneration disclosures and Bank calculations.

(a) Excluding figures for individuals who resigned or had their contracts terminated.

Chart 6 Total malus adjustments applied across thematerial risk-takers population within the majorUK banks(a)

(1) See Bank of England (2013).


Figure 2 provides an illustrative example of a three-yeardeferral period with and without the application of malus.In this example an individual working in a bank is awarded£500,000 of variable remuneration at the end of the year.Half of this will be paid in cash and half paid in shares of thebank. The maximum amount of the individual’s variable paythat can be received upfront at the end of the year is£200,000, 40% of the initial value. The remaining £300,000(60% of the initial value) will be paid out over the nextthree years in three equal amounts of £100,000. It is thesedeferred amounts that will be subject to reassessment untilthey are paid out.

Should no adjustments occur at the end of Year 3, theindividual would have received the entire £500,000 award, asindicated by the blue bars. However, should there be evidence

of misconduct, a risk management failing or materialdownturn in the firm’s performance in Year 2, say, then theremaining awards are potentially subject to reassessment andthe application of malus to all or part of them. Subject to theseverity, the firm could apply up to 100% malus on theremaining awards of Years 2 and 3, and the individual wouldonly have received £300,000 of his or her initial award at theend of Year 3, as indicated by the orange bars.

Rationale for longer deferral requirementsThe new extended deferral rules coming into force at the startof 2016 reflect the conclusions of the PRA and FCA and thePCBS that strengthening of the existing rules was desirable.Figure 3 shows how the deferred amounts change under thecurrent and new remuneration rules from three years toseven years. The figure also refers to ‘clawback’ which isexplained further on.

The PCBS took the view that short deferral periods wereinsufficient to take account of the timescales over whichmaterial business issues can come to light and concludedthere should be a presumption that all staff to whom the newrules apply should be subject to greater and longer deferralthan had been customary. Furthermore, the PCBS suggestedthat the regulators should have a power to require that asubstantial part of remuneration be deferred for up toten years where necessary for effective long-term riskmanagement.

Longer deferral periods better align the risk horizons of keyindividuals with the longer-term safety and soundness of thefirms for which they work. This is particularly important forsenior executives who help to determine or are responsible forimplementing the overall business strategy of a firm and areultimately responsible for risk management. For theseemployees, the risk horizon should reflect the timescales overwhich the risks associated with those strategic decisions are

0

50

100

150

200

250

Year 0(b) Year 1 Year 2 Year 3

Total upfront and deferred variable remuneration paid out

Total upfront and variable remuneration paid out prior to reassessment

Malus applied in Years 2 and 3 following a reassessment of performance

£ thousands

Source: Bank calculations.

(a) Based on current remuneration rules.(b) Year 0 represents the year in which the initial award of £500,000 is made and represents the

undeferred amount of 40% (with 60% deferred).

Figure 2 Comparison of a minimum three-year deferralperiod with and without malus applied(a)

0

20

40

60

80

100

120

Year 0(c) Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 Year 11 Year 12

Subject to clawback

Seven-year deferral (new rules)

Three-year deferral (old rules)

Per cent

Three-year clawback extension(b)

Figure 3 Illustration of deferral and clawback for senior managers under the old and incoming remuneration rules(a)

Source: PRA.

(a) Illustration based on 60% of the award being deferred.(b) Line includes representation of extension of clawback by a further three years where there is an investigation ongoing at Year 7 from the initial date of award.(c) Year 0 represents the year in which the initial award was awarded to the individual. Year 1 is the first anniversary of this date with the future years following suit.


likely to manifest themselves. Extended deferral periods alsoallow for more variable remuneration to be subject to malusover a longer time frame.

Recent examples of risk management failings demonstrate thelong-term risk horizons involved. The mis-selling of paymentprotection insurance is an example of a conduct failing whichtook many years to come to light. Libor rate manipulation isanother example where the initial practices took place beforethe crisis but the conclusion of initial enforcementinvestigations and subsequent fines came several years after.(1)

Deferral should also seek to align awards to the typicalbusiness cycles which banks operate in and underpin theperformance of the financial sector. In the November 2012Financial Stability Report the Financial Policy Committee(FPC)(2) noted that short deferral periods make it difficult toevaluate performance effectively. The report demonstratedthat the average LTIP length was much shorter (three tofive years) than the average business cycle (five and a halfyears) and the estimated length of the credit cycle (eight to30 years). However, in determining the most appropriate levelof deferral, consideration of the time-discounted value ofdeferred remuneration to the recipient is also required.

Ex-post risk adjustment: clawbackFrom 2015, the PRA and FCA extended the principle of ex-postrisk adjustment to awards that have already been paid orvested. This is called clawback.(3) Being able to makeemployees (or former employees) pay back the value ofawards already received is an important extension of theprinciple, not least because it applies to the whole award,including the undeferred portion. The capacity to applyclawback is required for seven years from the date of the initialaward for all MRTs. This may be extended by a furtherthree years for those covered by the SMR where there areongoing enquiries into a possible cause for clawback in theseventh year. Thus clawback increases both the amount andthe period over which an award is at risk (Figure 3).

The minimum grounds for which clawback may be applied arenarrower than malus with the scope excluding cases wherethere is a material downturn in firm performance. The PRAconcluded that clawback is most appropriate in cases wherethe individual has some responsibility or culpability forcircumstances giving rise to the grounds for action.

Aggregate payouts and capital levelsA firm’s profits, its remuneration payouts and its capitalposition are all interrelated. When a bank makes profit it canelect to use it in a number of ways. These include keeping it inthe firm, thereby increasing the bank’s capital base (and henceits capacity to absorb losses); paying dividends toshareholders; or paying it out, in the form of variableremuneration, to its employees.

The FSB P&S require that sufficiently large financialinstitutions should ensure that total variable remunerationdoes not limit their capacity to strengthen their capital bases.This has been an important element of the rules in recentyears as banks have built up their capital bases in theaftermath of the crisis in response to higher regulatory capitalrequirements. For the major UK banks in particular, the PRAneeds to be satisfied that any distributions through variableremuneration or distributions to shareholders are consistentwith a sustainable capital plan.

The bonus capCRD IV includes a ‘bonus cap’ whereby variable remunerationis limited to 100% (or 200% with shareholder approval) of thevalue of fixed pay, this is applicable to all MRTs. The premisebehind the bonus cap is that the expectation of large variableremuneration awards inherently creates adverse risk incentiveswhich cannot be solely mitigated by ex-ante and ex-post riskadjustments. Therefore a ceiling on variable pay should limitexcessive risk-taking as the upside gains are limited.

However a significant consequence of limiting variableremuneration is a shift to higher levels of fixed pay to offsetthe reduction in variable remuneration. The problem is thatthis reduces the amount of remuneration which is at risk. Thisis covered in more detail in the next section.

Looking ahead: the policy agenda

International discussionsRemuneration remains very much a live issue for regulatorsaround the world. Through the FSB, regulators continue towork to ensure the effective implementation of the reformsintroduced since 2009, and to focus on how to secure moreeffective alignment between risk-taking incentives andremuneration. The United Kingdom has been one of theleading jurisdictions in implementing the FSB P&S. However,implementation has been uneven around the world, with tenout of 24 jurisdictions having gaps remaining in their nationalimplementation of the FSB P&S as of November 2015.(4)

Following its March 2015 plenary meeting in Frankfurt, the FSBsaid it would review the impact of regulatory reformsincluding how remuneration structures had helped to reducemisconduct and whether any additional measures wereneeded.(5) Remuneration regimes have a role to play inincentivising good conduct in banks. This includes individualslosing rewards when a risk management failing comes to light.

(1) Libor stands for the London interbank offered rate and is the benchmark rate thatbanks charge each other for short-term loans.

(2) The FPC is charged with a primary objective of identifying, monitoring and takingaction to remove or reduce systemic risks with a view to protecting and enhancingthe resilience of the UK financial system as a whole.

(3) See Bank of England (2014).(4) See Financial Stability Board (2015a).(5) See Financial Stability Board (2015b).


(a) The current rules on deferral were introduced in 2010. The forthcoming rules (no less than seven-year deferral) will apply from January 2016 to senior managers in banks. See main text of the article for more details.

40%

20% 20% 20%

40%

12% 12% 12% 12% 12%

Pre-crisis

Post-crisis

Deferral, malus, and clawback arrangements for variable remuneration were not regulatory requirements.

Performance measures used to determine variable remuneration packages relied heavily on return metrics that were not adjustedfor the risks taken.

No restrictions on how much variable remuneration packages are paid out in cash versus non-cash instruments (such as shares).

No restrictions on the size of variable remuneration relative to fixed remuneration.

Variable

The bonus cap sets a maximum variable to fixed remuneration ratio of 2:1 with shareholder approval.

Balanced suite of performance metrics contribute to ex-ante adjustment of all variable remuneration.

All variable remuneration is subject to clawback. This is the process whereby banks are able to take back vested variable remuneration as a result of misconduct or risk management failings.

2:1ratio

50:50

Bonus cap

Cash/non-cash

Clawback

Performance metrics

All variable remuneration composed of 50% cash and 50% non-cash.

Comparison of deferral timelines under current (three years)and forthcoming (seven years) rules(a)

40% of all variable remuneration should be paid upfront with 60% deferred. All deferred remuneration, which has not yet vested, can be subjected to malus to take account of instances of misconduct, risk management failings, or downturn in financial performance.

Deferral and malus

Three-year rules

Seven-year rules

Risk-adjustedreturn

metrics

Strategicmetrics

Non riskadjusted

returnmetrics

Conductmetrics

Prudentialmetrics

Year 0 Y1 Y2 Y3 Y4 Y5 Y6 Year 7

Percentage of award paid out

The following rules have been introduced for the ‘material risk-takers’ in banks (see main text of the article for more details).

How have remuneration rules changed in the banking sector?


The rise in fixed remunerationThe proportion of fixed pay has increased since the rules wereintroduced in 2010, particularly in response to theintroduction of the EU bonus cap in 2014. This approach ofcapping bonuses has not gone unchallenged. Governor Carneyand Deputy Governor and CEO of the PRA, Andrew Bailey,have argued strongly in favour of variable remunerationconstituting a substantial portion of overall pay in order toalign incentives appropriately.(1) Indeed the bonus cap iscounterproductive: it reduces the scope for beneficial riskadjustment to affect incentives through malus and clawbackand it reduces flexibility in banks’ cost bases. A recentEuropean Systemic Risk Board report also raised concernsregarding the increase in fixed remuneration at the expense ofvariable remuneration.(2) The significance of this argumenthas been underscored by the clear shift to fixed pay from 28%to 54% across the MRT population from 2013 to 2014(Chart 7), during which period overall remuneration costs inthe major UK banks have remained broadly constant. Anincrease in the proportion of fixed pay has also been the caseelsewhere in the EU.(3)

The United Kingdom’s Fair and Effective Markets Review(FEMR) also noted the shift towards higher levels of fixedremuneration and cited regulatory concerns on aligningincentives.(4) To address this, FEMR said that the FSB shouldwork to ensure that an appropriate proportion of totalremuneration was variable.

Misconduct and new remuneration structuresFEMR also concluded that there was scope for firms toimprove further the link between conduct and remuneration.A series of misconduct failings in the global banking industry,including the Libor and foreign exchange manipulationincidents, evidence this further. The United Kingdom’s

introduction of extended deferral and clawback periods aim tostrengthen this link.

The issue of whether a greater proportion of totalremuneration should be put at risk of downward adjustment,potentially for ten years, was also a feature of proposals putforward by William Dudley, President and CEO of the FederalReserve Bank of New York. In a speech in October 2014 heproposed the concept of ‘performance bonds’ for seniorexecutives — funds built up over time from senior executives’remuneration and available to recapitalise the firm or to meetsubstantial regulatory fines.(5) This would require executivesto hold a meaningful component of long-term unsecured debt.Thus the performance bond would be a form of contingentliability for the firm to the employee that could be reduced incertain circumstances.

Governor Carney has supported further exploration of thisapproach as it could provide a solution to address the rise infixed pay and any future misconduct or prudential failings(Carney (2014)). It is important to look further at ways inwhich beneficial risk incentives are strengthened, includingways to ensure that a greater proportion of remunerationremains at risk for longer if misconduct or managementfailures come to light.

Conclusion

Poorly aligned incentives which encouraged excessiverisk-taking behaviours contributed to the financial crisis andled to the regulation of remuneration. However, closealignment between risk and reward, including the use ofvariable remuneration, can contribute to the safety andsoundness of firms and the stability of the financial system.The PRA, and other major international regulators, haveimplemented regulation on remuneration with the objectiveof ensuring employees’ incentives align with the longer-terminterests of the firm and society. The key tools are effectiverisk adjustment and the use of deferral, malus and clawback.However, challenges remain and the future regulatory focus islikely to centre on ensuring that a sufficient proportion of totalremuneration remains at risk, and on the role of compensationin addressing misconduct.

(1) See Carney (2014) and Bailey (2014).(2) See European Systemic Risk Board (2015).(3) See Figure 3.2, International Monetary Fund (2014).(4) See Bank of England, Financial Conduct Authority and HM Treasury (2015).(5) See Dudley (2014).

0

10

20

30

40

50

60

70

80

90

100

2010 11 12 13 14

Variable remuneration

Fixed remuneration

Per cent

Sources: Pillar 3 remuneration disclosures and Bank calculations.

Chart 7 Fixed versus variable remuneration as aproportion of total remuneration for the MRTs in themajor UK banks


References

Bailey, A (2014), ‘Andrew Bailey’s speech at the City Banquet, London’, available atwww.bankofengland.co.uk/publications/Documents/speeches/2014/speech763.pdf.

Bank of England (2013), ‘PRA expectations regarding the application of malus to variable remuneration’, PRA Supervisory Statement 2/13,October, available at www.bankofengland.co.uk/pra/Documents/publications/policy/2013/appofmalusss2-13.pdf.

Bank of England (2014), ‘Clawback’, PRA Policy Statement 7/14, July, available atwww.bankofengland.co.uk/pra/Documents/publications/ps/2014/ps714.pdf.

Bank of England and Financial Conduct Authority (2015), ‘Strengthening the alignment of risk and reward: new remuneration rules’,PRA Policy Statement 12/15, FCA Policy Statement 15/16, June, available atwww.bankofengland.co.uk/pra/Documents/publications/ps/2015/ps1215.pdf.

Bank of England, Financial Conduct Authority and HM Treasury (2015), ‘Fair and Effective Markets Review — Final Report’, available atwww.bankofengland.co.uk/markets/Documents/femrjun15.pdf.

Carney, M (2014), ‘The future of financial reform’, available atwww.bankofengland.co.uk/publications/Documents/speeches/2014/speech775.pdf.

Dudley, W C (2014), ‘Enhancing financial stability by improving culture in the financial services industry’, available atwww.newyorkfed.org/newsevents/speeches/2014/dud141020a.html.

European Systemic Risk Board (2015), ‘Report on misconduct risk in the banking sector’, available atwww.esrb.europa.eu/pub/pdf/other/150625_report_misconduct_risk.en.pdf.

Financial Stability Board (2009), ‘FSB principles for sound compensation practices’, available at www.financialstabilityboard.org/wp-content/uploads/r_090925c.pdf?page_moved=1.

Financial Stability Board (2015a), ‘Implementing the FSB principles for sound compensation practices and their implementation standards:fourth progress report’, available at www.financialstabilityboard.org/wp-content/uploads/FSB-Fourth-progress-report-on-compensation-practices.pdf.

Financial Stability Board (2015b), ‘Press release FSB Plenary Frankfurt 26 March’, available at www.financialstabilityboard.org/wp-content/uploads/Press-Release-FSB-Plenary-Frankfurt-final-26Mar15.pdf.

Financial Stability Forum (2009), ‘FSF principles for sound compensation practices’, available at www.financialstabilityboard.org/wp-content/uploads/r_0904b.pdf.

House of Commons Treasury Committee (2009), ‘Banking crisis: dealing with the failure of the UK banks’, available atwww.publications.parliament.uk/pa/cm200809/cmselect/cmtreasy/416/416.pdf.

International Monetary Fund (2014), ‘Risk taking by banks: the role of governance and executive pay’, available atwww.imf.org/external/pubs/ft/gfsr/2014/02/pdf/c3.pdf.

Parliamentary Commission on Banking Standards (2013), ‘Changing banking for good’, available atwww.parliament.uk/business/committees/committees-a-z/joint-select/professional-standards-in-the-banking-industry/news/changing-banking-for-good-report/.

www.parliament.uk/business/committees/committees-a-z/joint-select/professional-standards-in-the-banking-industry/news/changing-banking-for-good-report/

www.parliament.uk/business/committees/committees-a-z/joint-select/professional-standards-in-the-banking-industry/news/changing-banking-for-good-report/

www.financialstabilityboard.org/wp-content/uploads/Press-Release-FSB-Plenary-Frankfurt-final-26Mar15.pdf

www.financialstabilityboard.org/wp-content/uploads/Press-Release-FSB-Plenary-Frankfurt-final-26Mar15.pdf

www.financialstabilityboard.org/wp-content/uploads/FSB-Fourth-progress-report-on-compensation-practices.pdf

www.financialstabilityboard.org/wp-content/uploads/FSB-Fourth-progress-report-on-compensation-practices.pdf

www.financialstabilityboard.org/wp-content/uploads/r_090925c.pdf?page_moved=1

www.financialstabilityboard.org/wp-content/uploads/r_090925c.pdf?page_moved=1

www.financialstabilityboard.org/wp-content/uploads/r_0904b.pdf


• The Prudential Regulation Authority’s (PRA’s) secondary competition objective requires the PRA to act, where possible, in a way that facilitates effective competition when making policies toadvance its primary objectives of safety and soundness, and policyholder protection.

• This article explains the rationale for the objective, how the PRA interprets it, and what thismeans for the PRA’s regulation of banks and insurers.

The Prudential Regulation Authority’ssecondary competition objectiveBy Stephen Dickinson, David Humphry, Paolo Siciliani and Michael Straughan of the Bank’s Prudential PolicyDirectorate and Professor Paul Grout, PRA Senior Advisor on Competition.(1)

(1) The authors would like to thank Nicola Garbarino and Liam Kirwin for their help in producing this article.

Overview

On 1 March 2014 the PRA was given a secondary objective toact, so far as reasonably possible, in a way that facilitateseffective competition in the markets for PRA-authorisedfirms carrying out regulated activities. This key addition tothe PRA’s remit applies when it is making policies in pursuitof its primary objectives: the safety and soundness of banksand insurers and insurance policyholder protection.

The secondary competition objective (SCO) has at its centrethe notion of effective competition. This can be thought ofas a market in which suppliers offer a choice of products orservices on the most attractive, sustainable, terms tocustomers; where customers have the confidence to makeinformed decisions; and where firms enter, expand and exitfrom the market. Each of these characteristics provideimportant sources of competitive discipline. The PRA’sresponsibility for facilitating effective competition iscomplementary to, but distinct from, the role of theCompetition and Markets Authority, the Financial ConductAuthority, and the Payment Systems Regulator. The SCOdoes not mean that the PRA regulates competition infinancial services markets.

Acting to promote safety and soundness of banks andinsurers and insurance policyholder protection by addressingmarket failures is likely to enhance effective competition.Therefore, the PRA’s primary and secondary objectives areoften complementary. However, the effectiveness ofcompetition can be reduced if regulation has unintendedconsequences. So an important advantage of the SCO is thatit provides a useful check on whether prudential interventions

are being applied sensibly and proportionately by consideringprudential regulation through a competition lens.

Since it came into effect, the PRA’s SCO has helped informthe design of several important parts of the framework forprudential regulation. For example, the implementation ofthe new Pillar 2 capital regime for banks allows supervisorsto exercise judgement when assessing credit concentrationrisk for small firms where the methodology could overstaterisks. The SCO has also informed the implementation of aFinancial Policy Committee Recommendation to the PRA onlimiting the extent of high loan to income residentialmortgages, which was designed to take into account thedifferent business models of independent private banks andtheir ability to compete in the mortgage market. Lookingahead, the PRA will continue to ensure that the SCO informsthe design of new policies.

In addition to new policies, the PRA also takes a proactiveapproach to its secondary objective by considering changesto existing policies to facilitate effective competition.Examples here include measures the PRA has taken tofacilitate the entry of new banks. Finally, the SCO is alsoinforming positions taken in international policymakingforums. For instance, the Bank of England’s response to theEuropean Commission’s consultation on the possible impactof the Capital Requirement Regulation and Directive on bank financing of the economy emphasised that a moreproportionate approach to bank regulation could supportcompetition in the sector.

Click here for a short video that discusses some of the key topics from this article.

https://youtu.be/pB6r84Ggziw

Topical articles The PRA’s secondary competition objective 335

Introduction

On 1 April 2013 two new regulators, the Prudential RegulationAuthority (PRA) and the Financial Conduct Authority (FCA),came into existence, replacing the Financial Services Authority(FSA). The PRA is responsible for the prudential regulation ofbanks, building societies, and credit unions (referred to as‘banks’ in this article), insurers and major investment firms.The FCA is responsible for ensuring that relevant marketsfunction well, and for the conduct regulation of all financialservices firms. It is also responsible for the prudentialregulation of those financial services firms not supervised bythe PRA.

Parliament initially gave the PRA two statutory objectives.First, a general objective to promote the safety and soundnessof PRA-authorised firms. And second, an insurance objectiveto contribute to the securing of an appropriate degree ofprotection for those who are or may become policyholders.This article will refer to these two objectives as the PRA’s‘primary objectives’.

On 1 March 2014 the PRA was subsequently given a secondarycompetition objective (SCO). This article explains the SCO.The first section discusses the rationale for the objective. Byway of context, it also summarises the remit and powers ofthe competition regulators in relation to the financial servicessector. The second section discusses how to interpret theSCO, explaining the concept of ‘effective competition’ andhow the SCO fits with the PRA’s primary objectives. The thirdsection explains how the PRA advances the objective inpractice, citing some examples of how it has informeddecision-making to date. Paul Grout, PRA Senior Advisor onCompetition and Professor of Political Economy at theUniversity of Bristol, explains why the PRA has the SCO in ashort video.(1)(2)

Rationale for the PRA’s secondarycompetition objective

Setting the scene: the rationale for the PRA’s primaryobjectivesThe principle underlying the PRA’s primary objectives is that astable financial system, which is resilient in providing criticaleconomic functions, is necessary for a healthy and successfuleconomy.(3) Firms can adversely affect the stability of thefinancial system through the way in which they carry out theirbusiness and, in the extreme, by failing in a disorderly manner.It is not, however, the PRA’s role to ensure that no firm fails.

As set out in a previous Bulletin article,(4) the rationale forthese objectives stems from the risk of poor outcomes —termed ‘market failures’ — in the provision of criticaleconomic functions by banks and insurers to customers.

These include:

• the possibility of a large bank becoming subject to a ‘run’— whereby a large number of customers attempt towithdraw their deposits at the same time — even if thebank is solvent, leading to unnecessary costs;

• harm to the financial system caused by a loss ofconfidence following the failure of one or more firms, or bya wave of insolvencies among counterparties to financialcontracts triggered by the insolvency of other firms; and

• excessive risk-taking by banks and insurers caused byuncertain values of assets and liabilities, for instance, dueto differences in incentives between the managers andowners of these firms and the difficulty for policyholdersand shareholders to monitor risk-taking properly.

The benefits of competition and the potential forineffective regulation to harm effective competitionGenerally speaking, strong and fair competition in marketsgenerates greater choice, lower prices, and better-qualitygoods and services for consumers. For businesses, acompetitive environment encourages innovation andefficiency, both of which can help to drive productivity andgrowth in the economy as a whole.

Prudential regulation designed to address market failures canaffect the way that banks and insurers compete for customers’business. In most instances prudential regulation aimed ataddressing market failures will enhance effective competition.However, due to its often complex nature, there existchallenges in the design of prudential regulation which insome cases can create unintended consequences andundermine effective competition in financial services markets.For instance, regulation might create barriers to entry for newfirms whose small size and significance mean that they mightpose only a limited risk to the disruption of critical economicservices (for example, new retail banks).

The relationship between prudential interventions aimed ataddressing market failures and effective competition appliesalso in the opposite direction. Ineffective regulatory oversightthat fails to adequately tackle market failures might spurineffective competition which, ultimately, does not delivergood outcomes for customers. For instance, in the run-up to the recent financial crisis, the expectation of

(1) https://youtu.be/pB6r84Ggziw.(2) The previous page includes a correction to the printed version of the Bulletin, which

cited ‘aspects of the Solvency II regime’ — as well as the Pillar 2 capital regime — asallowing supervisors to exercise judgement when assessing credit concentration riskfor small firms.

(3) The PRA defines critical economic functions that firms provide to be: payment,settlement and clearing; retail banking; corporate banking; intra-financial systemborrowing and lending; investment banking; custody services; life insurance; andgeneral insurance. See Bank of England (2014a).

(4) See Bailey, Breeden and Stevens (2012).


Competition responsibilities in the regulationof UK financial services(1)

This box explains the relationship between the competitionresponsibilities of the concurrent competition regulators —the CMA, the FCA, the PSR — and their interaction with thePRA in relation to financial services. It distinguishes betweentheir objectives and powers, and the markets to which theirremits apply. The coverage of the different authorities issummarised in Figure A.

Competition and Markets Authority (CMA)The CMA assumed its role as the United Kingdom’s primarycompetition and consumer agency in April 2014. In doing so,the CMA brought together most of the powers andresponsibilities of the now-abolished CompetitionCommission and the Office of Fair Trading. The CMA has a UK economy-wide competition remit including financialmarkets (the magenta ellipse in Figure A), overseeing thepromotion of competition with the aim of making marketswork well for consumers, businesses and the economy.

Competition powers: the CMA is responsible for:

• investigating mergers between firms which could restrictcompetition;

• investigating where there may be breaches of UK or EU prohibitions of anti-competitive agreements and abusesof dominant positions;

• conducting market studies and investigations in marketswhere there may be competition and consumer problems(see for instance its current market inquiry into retailbanking)(2) that do not generally involve any firm(s) (orindividual(s)) breaching UK or EU competition law;

• bringing criminal proceedings against individuals whocommit cartel offences;

• enforcing consumer protection legislation to tacklepractices and market conditions that make it difficult forconsumers to exercise choice;

• co-operating with sector regulators, such as the FCA andPSR, and encouraging them to use their competitionpowers; and

• considering regulatory references and appeals.

Financial Conduct Authority (FCA)The FCA’s competition mandate is twofold, involving acompetition objective and a duty. The FCA is required to actin a way that is compatible with its strategic objective ofensuring that relevant markets for financial services functionwell, as well as advancing its operational objectives, one ofwhich is to promote effective competition in the interest ofconsumers in regulated financial services (the pale blueellipse in Figure A). This includes ‘regulated activities’, whichrefers to PRA and FCA-regulated activities. The FCA also has acompetition duty to promote effective competition whenaddressing its other operational objectives, in relation toconsumer protection and market integrity. This means that itmust look to achieve its desired outcomes using solutions thatpromote competition regardless of which objective it ispursuing, so far as compatible with acting in a way whichadvances the consumer protection or integrity objective. TheFCA can use its Financial Services and Markets Act powers topursue its competition mandate, including conducting marketstudies.

Competition powers: in April 2015 the FCA received powersin relation to the enforcement of UK competition law. It haspowers like those of the CMA to enforce against breaches ofthe prohibitions on anti-competitive behaviour. It also hasadditional powers to carry out market studies and powers tomake market investigation references to the CMA, in relationto the provision of all financial services (the green ellipse inFigure A), of which ‘regulated financial services’ are a subset.In other words, its concurrent competition powers extend to awider set of markets than its competition objective.

Payment Systems Regulator (PSR)The PSR, which is a subsidiary of the FCA, is an independentregulator of payment systems. It was established in April 2014and became fully operational in April 2015.

All sectors of theeconomy (CMA)

All financial services/payment systems

(FCA/PSR)

Regulated financial services/payment systems (FCA/PSR)

Services provided byPRA-authorised

persons in carrying on regulated activities

(PRA/FCA)

Figure A The scope of different UK bodies’ competitionresponsibilities

(1) Where the effect of anti-competitive behaviour extends beyond the United Kingdomto other Member States, EU competition law is applicable. The EuropeanCommission, through the Directorate General for Competition, is responsible for EU-wide enforcement of competition law. The CMA and FCA have powers to enforce EU competition law in the United Kingdom.

(2) The CMA is investigating the SME banking and personal current account markets: formore detail, see https://www.gov.uk/cma-cases/review-of-banking-for-small-and-medium-sized-businesses-smes-in-the-uk.

https://www.gov.uk/cma-cases/review-of-banking-for-small-and-medium-sized-businesses-smes-in-the-uk

https://www.gov.uk/cma-cases/review-of-banking-for-small-and-medium-sized-businesses-smes-in-the-uk


The PSR has an objective to promote effective competition inmarkets for payment services.(1)

Competition powers: like the FCA, it has powers to enforceUK competition law. The PSR’s competition powers relate toparticipation in payment systems.

PRAThe Financial Services and Markets Act 2000 as amendedplaced a duty on the PRA to ‘have regard’ to a number ofregulatory principles, in addition to its safety and soundness,and policyholder protection objectives. These regulatoryprinciples included ‘the need to minimise any adverse effecton competition in the relevant markets that may result fromthe manner in which the PRA discharges [its] functions’. Inessence, this regulatory principle sought to ensure thatcompetition considerations were at least a factor the PRAshould consider when taking actions to meet its primaryobjectives. Subsequent to the adoption of the FinancialServices Act 2012, a Parliamentary Commission on BankingStandards (PCBS) was established in 2012 to conduct aninquiry into professional standards and culture in the UK banking sector and to make recommendations forlegislative and other actions. One of its recommendationsargued for greater weight to be placed on competitionconsiderations by the PRA, with the addition of a secondarycompetition objective. In 2013, the Government agreed withthe PCBS’s recommendation and introduced the SCO.

Competition objective: the PRA’s secondary objective tofacilitate effective competition relates to markets for servicesprovided by PRA-authorised persons in carrying out regulatedactivities (the blue ellipse in the diagram). These are markets

in which PRA-authorised firms, including branches located inthe United Kingdom, supply regulated services. They may belocal, national, or international in nature. For instance, thereinsurance market tends to be considered international inscope, with UK-based reinsurers supplying customers in othercountries and vice versa, while life insurance tends to beconsidered a national market. Similarly, the market for theprovision of corporate and investment banking services tolarge corporations tend to be global in nature, whereas thegeographical scope of retail banking activities is typicallyaligned with national boundaries.

Competition powers: the PRA is not responsible for enforcingcompetition law, and therefore does not have associatedcompetition powers.

Interaction between the PRA and the CMA, FCA andPSRThe PRA consults with the FCA and PSR where itspolicymaking is expected to be of material interest to one orboth of them, and vice versa. This could include thecompetition implications of new policy.

The PRA also interacts with the CMA, FCA and PSR oncompetition matters in relation to their market studies andmarket investigations; where, for example, the PRA may berequested to supply information about the prudentialregulatory regime.

government-funded support for institutions that wereperceived to be ‘too big to fail’ meant that investors expectedto incur only limited losses in the event of these firms failing.This created artificially low funding costs for these firms,placing them at a competitive advantage relative to smallerrivals. This, in turn, enabled these firms to increase lendingand risk-taking (for example, corporate finance for mergersand acquisitions, or sub-prime lending).(1)

Similarly, there are examples where ineffective regulatoryoversight in the insurance industry has had adverse effects forcompetition. French, Minot and Vital (2015) discuss the caseof HIH Insurance Group, the Australian general insurancecompany which monopolised the market for mandatorybuilders’ warranty insurance by underpricing risk to such anextent that competitors exited the market and new entrantswere deterred from entering, which caused a disruption in theprovision of this service when the firm in question failed.

A secondary competition objective for the PRA andhow it fits with the competition objectives of otherregulators On 1 March 2014 the PRA was given a secondary competitionobjective (SCO). The objective states that, when dischargingits general functions in a way that advances its primaryobjectives, the PRA must, so far as is reasonably possible, actin a way which, as a secondary objective, facilitates effectivecompetition in the markets for services provided by PRA-authorised persons in carrying out regulated activities.No specific powers or toolkit have been given to the PRA inrelation to promoting competition — these lie appropriatelywith the primary concurrent competition regulators active inthe financial services sector, namely the FCA, the PaymentSystems Regulator (PSR), and the Competition and MarketsAuthority (CMA).

(1) The competition objective also extends to the markets for services provided bypayment systems, in the interests of those who use, or are likely to use, servicesprovided by payment systems.

(1) In addition, this caused moral hazard, contributing to excessive risk-taking since thesefirms did not bear the full economic cost of their actions. See Noss and Sowerbutts(2012).


Although it has an SCO, the PRA is not a ‘competitionregulator’. The SCO applies when the PRA is exercising itsgeneral functions (such as making rules or setting generalpolicies) to advance its primary objectives. The PRA’sobjective to facilitate effective competition is distinct from,but complementary to, the concurrent powers of thecompetition regulators relating to the financial services sector,the FCA, the PSR and the CMA. As explained in the box onpages 336–37, these bodies are responsible for promotingcompetition and have a wide-ranging toolkit and specificpowers to achieve their objectives.

Interpreting the PRA’s secondary competitionobjective

To understand how the PRA meets its competition objective inthe context of the design of prudential regulations it is usefulto consider: the secondary nature of the objective; the PRA’sduty to facilitate effective competition as far as is reasonablypossible; what is understood by ‘effective competition’; andhow the objective fits together with the PRA’s primaryobjectives. This section considers these in turn.

The secondary nature of the objectiveWhen the PRA is considering options for new regulation it willassess the scope to facilitate effective competition, choosingprudential regulation options that do so as far as is reasonablypracticable. The SCO does not require the PRA to act in amanner that is incompatible with its primary objectives. Inmany cases the PRA’s primary and secondary objectives will befully aligned: for example, reducing ‘too big to fail’ distortionsappear to have made both the financial system safer andcompetition more effective. Nevertheless, cases might existwhere, within the range of prudential regulation optionsavailable to the PRA, there may be some which would delivergreater benefits to competition and others which woulddeliver greater benefits to safety and soundness orpolicyholder protection. The existence of the SCO means thatthe PRA should consider — but is not necessarily required toadopt — those options which would deliver greater benefits tocompetition for a given objective of safety and soundness orpolicyholder protection.

The PRA should facilitate effective competition as faras is reasonably possibleThe PRA’s duty under the SCO extends ‘as far as is reasonablypossible’ recognising that the PRA must comply with somelaws where it may not have a choice of regulation options and,therefore, a choice of considering a range of outcomes forcompetition. For instance, there could be legal restrictions onthe implementation of regulation changes, such as where thePRA is bound by domestic legislation or European law.(1)

Indeed, a large proportion of prudential regulation applied bythe PRA is derived from ‘maximum harmonising’ Europeanlaw, where countries do not have discretion as to how theyimplement the rules.

Facilitating ‘effective competition’The legislation setting the PRA’s SCO does not define‘effective competition’.(2)

Effective competition is not the same as what economiststerm ‘perfect competition’. In a world of perfect competition,individual suppliers can increase supply but have no effect onprice, and can freely enter and exit the market. Products andservices in the same market are assumed to have the samespecifications and quality. Buyers are perfectly-informed andcan switch supplier without cost if prices rise above the marketprice. Market prices reflect the costs of supplying the productand the cost of any side-effects that fall on others in society.

In such a market there is no need for regulation as there are nomarket failures to be addressed.(3) These conditions, however,tend not to exist in markets for financial services.

Policymakers therefore need to recognise the challengesposed by the existence of market failures. In doing so, theycan seek to achieve a world of ‘effective competition’, whichcan be broadly understood as one in which market andregulation failures are either not significant or else have beenaddressed. The regulation framework, designed to overcomemarket failures, enables suppliers and customers to interact,with suppliers meeting customers’ needs and customers’demand stimulating rivalry between suppliers for customers’business. This framework for rivalry enables the benefits ofcompetition to materialise — lower prices, better quality, andgreater innovation — across markets in financial services,ranging from offering mortgages to dealing in investments,without striving to create the conditions for perfectcompetition, which could be an unrealistically difficult, orunnecessary, standard to achieve.

As such, the PRA view is that a market which consists of rivalproducts or services would tend to demonstrate effectivecompetition if the interaction between suppliers andcustomers, and the conditions for entry, expansion and exit,display the following characteristics:(4)

(1) For example, the Capital Requirements Directive and Regulation which, among otherrequirements, sets the level of capital requirements.

(2) It is worth noting that the objective refers to ‘competition’ rather than‘competitiveness’. Under the Financial Services and Markets Act 2000, the FSA,which was the predecessor to the PRA and FCA, was required to have regard to‘competition’ and ‘competitiveness’. Competitiveness is usually taken to mean afirm’s, or a group of firms’, or a nation’s ability to compete internationally on cost orquality grounds. Ability to compete is a part of competition, but competition isshaped by other factors, including the range of firms competing in the market;barriers to entry, expansion, or exit; and incentives to compete. So while the twoconcepts are related, they are clearly distinct.

(3) Another departure from the assumptions of perfect competition is that financialservices vary widely in the extent to which they are standardised. Products rangefrom commodity futures contracts traded on exchanges, with specification ofquantity, quality and delivery dates, to highly bespoke services, such as insuranceagainst damage to a pop star’s voice.

(4) Our definition of effective competition is consistent but distinct from the definitionof effective financial markets described under the Fair and Effective Markets Review,where the lack of transparency is seen as the major constraint to the development offair competition among financial market participants: see paragraphs 20–21 of Bank of England (2015).


(a) Suppliers compete to offer a choice of products orservices on the most attractive terms to customers —such as lower prices or better quality. At the same time,suppliers appropriately price in the risks associated withtheir businesses such that they have confidence in theirability to meet their service obligations.

(b) Customers have the confidence to make informedchoices. These choices are based on those qualityattributes that are easy to observe. Products and servicescan be obtained, and customers receive the products andservices they expect, at a price that allows suppliers toearn a return on their investment commensurate with thelevel of risk taken. Suppliers do not place unreasonable orunnecessary restrictions on products and services thatwould prevent customers from exercising choice.

(c) Effective entry, expansion and exit. It is possible forsuppliers, including those offering new products andservices, to enter the market and to expand; and suppliersoffering products or services on unattractive terms, orwhich are unable to meet their obligations, to exit themarket in an orderly fashion.

The relationship between the PRA’s primary objectivesand effective competitionEffective competition and regulation in pursuit of financialstability or policyholder protection are normally fullyconsistent. Regulation designed to improve financial stability(by addressing market failures) can facilitate effectivecompetition. For example, regulation can ensure that firms’and investors’ decision-making appropriately reflects thesocial cost of risk-taking, reducing incentives to compete in away that harms others in the financial system such as taking

unsustainable levels of risk only to expose others to lossesthrough failure and loss of confidence. A robust minimumprudential regulation standard can also provide customerswith greater confidence in the financial soundness of newbanks and insurers, enabling entrants to gain a foothold in themarket and to expand.

Similarly, regulation that creates minimum standards ofprotection for policyholders means customers can havegreater confidence that insurers will continue to be required tomeet their claims or payments of benefits, although these mayonly materialise many years into the future. This enablesinsurers to compete based on the quality and costs of theirproducts, responding to customer demand, taking prudentialstandards as a given.

Some examples of how the prudential regulation frameworksupports effective competition are considered in Table A. Theexamples refer to the three main features of effectivecompetition set out above and are not confined to regulationaddressing financial stability concerns. The box on page 340reviews some of the arguments put forward in the academicliterature on the relationship between competition andfinancial stability.

But not all forms of regulation that promote financial stabilityor policyholder protection will necessarily facilitate effectivecompetition. For example, due to its often complex nature,regulation may have unintended consequences that couldhave a negative impact on effective competition. Anadvantage of looking at prudential regulation through acompetition lens, therefore, is that it provides an additionalcheck on whether prudential interventions are being appliedsensibly and proportionately.

Aspect of effective competition Examples of how supported by regulation framework

Suppliers compete to offer achoice of products or services onthe most attractive terms tocustomers — such as lower pricesor better quality

The existence of banks perceived to be ‘too big to fail’ was a key factor in the financial crisis. Not only did such banks impose losses on taxpayerswhen they failed (rather than just those directly involved in the bank), their ‘too big to fail’ status encouraged them to take greater risks in goodtimes, increasing the likelihood that they would eventually fail. Stronger capital requirements for firms reduce excessive risk-taking by realigningthe incentives of systemically important banks with those of wider society, making competition more effective.

Similarly, consumers may find it difficult to judge the comparative financial soundness of insurers and whether insurers will need to take excessiverisks to meet policyholders’ claims.(a) There have been prominent examples of certain life insurers competing by offering savings products with highguarantees that subsequently many such insurers were unable to meet. In several cases, these episodes followed deregulation which was arguablytoo extensive.(b) The prudential framework established more recently under Solvency II creates a minimum set of standards around safety andsoundness and policyholder protection which means insurers should compete more effectively on the basis of the products they offer and how wellthese meet consumer demand.

Customers have the confidenceto make informed choices

Other aspects of the prudential regulation regime enable customers to exercise choice. Deposit guarantee schemes provide reassurance todepositors that they can safely leave their money, at least up to certain thresholds, with banks, whether these banks are incumbents or newentrants. This is important not only to avert ‘bank runs’, but also to empower consumers (depositors) to shop around with confidence. Consumerscan look more to those attributes of the retail banking service which they can verify such as cost and service. Insurance guarantee schemes, byprotecting policyholders from the effects of insurance company failure, play a similar role, enabling greater consumer choice in insurance markets.

Effective entry, expansion andexit

As discussed above, ‘too big to fail’ banks not only had increased incentives to take risk, but they were also able to maintain their share of themarket. This is because the ‘too big to fail’ status of some banks effectively reduced their funding costs, undermining the ability of smaller butefficient, well-run banks to gain market share, potentially forcing some to exit the market and others not to try to enter at all. A robust resolutionregime helps to remove the funding distortions caused by the ‘too big to fail’ perception, helping new firms to enter markets and to expand. Aresolution regime also aims to ensure that firms exit markets in an orderly manner if they do fail.(c)

The prudential regulation framework and deposit/insurance guarantee schemes also help build consumer trust in the safety and soundness of newentrants. This mitigates the barrier to entry arising from consumers’ tendency to favour established brands over unfamiliar ones.(d)

Table A Examples of how the prudential regulation framework supports effective competition

(a) See Debbage and Dickinson (2013).(b) See Das, Davies and Podpiera (2003). (c) See Bank of England (2014b).(d) Brand-building (or ‘persuasive’) advertising can be a source of what economists term a ‘sunk’ cost which might operate as a barrier to entry and, in particular, expansion and this could help to preserve an oligopolistic structure.

See, for instance, Geroski (1995).


There is no general consensus on the relationship betweencompetition and financial stability in the banking sector (thereis only a very limited literature available on this issue for theinsurance sector). It is helpful to distinguish between twoseparate, but related, issues. The first issue, discussed in themain text, concerns how and whether changes in regulation to

achieve the appropriate level of safety and soundness in thebanking sector help facilitate or hinder effective competition.

The second, discussed below, relates to how changes incompetition, with prudential regulation held constant orindeed absent, impact on the stability of the banking sector.

Is competition positive or negative for financial stability?

On the one hand…There are a number of hypotheses that argue forcompetition improving financial stability. The mechanismsby which competition improves stability include: entry,expansion and exit of firms; and reducing risk-taking.

On the other hand…There are also a number of hypotheses that focus on thepotential trade-off between financial stability andcompetition. These mechanisms include: increasing risk-taking and reducing banks’ ability to absorb losses.

Competition promoting stability through entry, expansionand exit: increased price rivalry among banks can benefitboth deposit holders — banks compete for deposits byraising deposit rates — and borrowers — banks lower interestrates on loans. Banks that are more efficient can thriveunder this sort of competition by offering the best depositand loan rates, taking business from less efficient, or lesswell-managed rivals over time. Efficiency can be in the formof better monitoring of customers and management ofloans. These banks are able to attract the least riskycustomers and investors are willing to fund their activitiesgiven they make fewer overall losses.(1) The stability of thebanking sector is improved over time as more efficient bankseither take over inefficient banks and/or inefficient banksleave the market.(2)

Competition reducing the level of risk-taking byborrowers: higher-risk projects tend to generate higherreturns. For an entrepreneur borrowing from a bank toinvest in a project, higher loan rates could make lower-riskprojects unviable. However, where competition reduces loanrates, more of these low-risk projects become viable. Awider spectrum of lower-risk projects can expect positivereturns on investment (net of interest payments) reducingthe overall level of risk in banks’ lending portfolios, andincreasing financial stability.(3) Separately, expected lowerloan rates (due to expected competitive rivalry amonglenders) could motivate entrepreneurs to apply more effortto making a project a success, since they would keep more ofthe profits than if loan rates were higher, helping banks avoidlosses.(4)

Competition reducing banks’ ability to absorb unexpectedlosses: increased price rivalry may compress banks’ marginsand reduce their ability to generate profits. This, in turn,reduces banks’ ability to accumulate additional capital inorder to absorb any losses that might occur, which couldworsen financial stability.(5)

Competition increasing risk-taking by banks: lower profitmargins from greater competition can push lenders to seekcreditors with higher, but less certain, returns — that is, theytake on higher-risk lending.(6) Shareholders are prepared totake the risk because they reap the benefits if things go wellbut their losses are limited if things go badly (their limitedliability means the most they can lose is the value of theirequity). This is commonly referred to as a loss of ‘charter/franchise value’ of the bank.

Competition reducing screening of borrowers: increasedprice rivalry can also reduce lenders’ incentives to screen andmonitor borrowers’ creditworthiness properly. Banks invest in collecting information about borrowers but wherecompetition is strong, banks cut costs wherever possible,which could include the costs of gathering information. Thiscould lead to banks taking greater risk than anticipated andsuffering higher losses.(7) However, as argued before, underthese circumstances there can be strong offsetting incentivesto improve the screening and monitoring of borrowers, inorder to outperform rivals which have underinvested in theseessential activities.

(1) See Perotti and Suarez (2002). (2) See Degryse and Ongena (2008).(3) See Boyd and De Nicolo (2005).(4) See Padilla and Pagano (1997).(5) See Marcus (1984).(6) See, for example, Dewatripont and Tirole (1994). (7) See, for example, Chan, Greenbaum and Thakor (1986). Similarly, in competing for borrowers by offering lower loan rates, banks might be exposed to the ‘winner’s curse’ problem, where the winning bank might have

overestimated the borrower’s creditworthiness: see Broecker (1990).

The relationship between competition and stability in the banking sector


How does the PRA apply the secondarycompetition objective?

The competition objective informs decisions about newregulation or supervision policies (so-called ‘generalfunctions’). Work to date to embed the objective acrossregulation and supervision activity within the PRA’s scope hasincluded updating internal decision-making procedures;developing training and guidance for staff; and recruitingspecialist competition resources, including the appointment ofa Senior Advisor on competition.

The PRA carries out competition analysis using establishedmethodologies, adapted to its own objectives, to ensure thesecondary objective informs the design of new policy.Applying its definition of effective competition, the PRAconsiders which policy options, among those that promote itsprimary objectives and are credible, also facilitate effectivecompetition. The PRA considers whether its policies couldhave unintended consequences for effective competition.Where this is possible, it identifies how a policy option couldchange the behaviour of suppliers and customers, or marketstructure, such as barriers to entry, in the markets affected bythe policy. For instance, the PRA sets capital and liquidityrequirements. The PRA assesses whether on balance thesechanges are positive or negative for the effectiveness ofcompetition and, where possible, chooses the action thatfacilitates effective competition while not undermining itsprimary objectives.

Important examples of competition analysis informing thePRA’s decision-making to date include:

• The PRA’s consultation on the ‘Pillar 2’ capital frameworkfor banks. The Pillar 2 framework deals with shortcomingsin the measures of risk-weighted assets.(1) A fundamentalprinciple underpinning this review was to make Pillar 2capital methodologies more risk-sensitive, including viahigher total capital requirements for systemically importantfirms and lower total requirements for smaller firms andnew entrants. Moreover, under the new regime, supervisorsmay exercise judgement when assessing creditconcentration risk — whether on a sectoral or ageographical basis — for small firms where they identifythat the credit concentration risk methodology couldoverstate risks, or could incentivise risk-taking behaviour.

• The FPC’s Recommendation in June 2014 that the PRA(and the FCA) take steps to ensure that mortgage lendersconstrain the proportion of new lending above a certainloan to income threshold.(2) This action was designed tolimit the size of any future economic downturn by limitingthe excess build-up of household debt with a high loan toincome ratio ahead of any downturn. From the perspective

of effective competition, the final PRA rules to implementthe FPC’s Recommendations were designed to take intoaccount the different business model of independentprivate banks (that is, those not part of a larger group) andtheir ability to compete in the mortgage market. For thesebanks, the minimum amount of lending required in orderfor a firm to fall within scope of the rules was set at ahigher level than for other types of firm, and stand-aloneprivate banks were exempt from the measures altogether.

In each case the PRA was able to advance its primaryobjectives while also facilitating effective competition.

The Bank undertakes research to help it improve its analysis ofcompetition issues relating to prudential regulation, which inturn benefits its development of new regulations. Work isunder way, for instance, to better understand the relationshipbetween safety and soundness and effective competition, aspart of the One Bank Research Agenda.(3)

The PRA takes a proactive approach by considering how itmight facilitate effective competition by making changes to itsexisting regulation framework, without undermining safetyand soundness and policyholder protection. For example, in2014 it conducted a review of the changes it made in 2013 tofacilitate entry and expansion in banking markets. Notingpositive developments,(4) it decided not to make furtherchanges at that stage.(5)

Finally, the PRA is taking a proactive approach to consideringfacilitating effective competition in its approach tointernational negotiations. It identifies internationalnegotiations on sets of rules (including existing ones underreview) where the outcome is likely to have a material effecton competition in the United Kingdom, and takes into accountthe scope to facilitate effective competition withoutcompromising the PRA’s primary objectives. One recentexample is the Bank’s response to the European Commission’sconsultation on the possible impact of the CapitalRequirements Directive and Regulation on bank financing ofthe economy, where the PRA has argued that a moreproportionate approach to bank regulation could supportcompetition in the sector.

Clearly, there is a need to be flexible in such negotiations,particularly when interacting with other regulators that havedifferent objectives. But by forming a strategy at an early

(1) For more information on the Pillar 2 framework, seewww.bankofengland.co.uk/pra/Documents/pillar2framework.pdf.

(2) See Financial Stability Report, June 2014, available atwww.bankofengland.co.uk/publications/Documents/fsr/2014/fsrfull1406.pdf.

(3) See Theme 2 of the agenda, available atwww.bankofengland.co.uk/publications/Pages/news/2015/032.aspx.

(4) Since the changes made by the PRA and FCA to the Authorisation process in 2013,thirteen new banks have been authorised, with around 20 more applications forauthorisation in the pipeline.

(5) See Bank of England and Financial Conduct Authority (2014).

www.bankofengland.co.uk/publications/Documents/fsr/2014/fsrfull1406.pdf

www.bankofengland.co.uk/pra/Documents/pillar2framework.pdf


stage it gives the PRA a better chance of influencing the rulesbefore they are made and before the PRA must implementthem. In practice, some of the PRA’s most importantinternational work to promote safety and soundness —removing ‘too big to fail’ distortions or the appropriatecalibration of capital requirements — is also its mostimportant work for facilitating effective competition.

Conclusion

A stable financial system, which is resilient in providing criticaleconomic functions, is necessary for a healthy and successfulUK economy. Prosperity can be reduced when there arematerial market failures, creating risks to the safety andsoundness of the financial system and to policyholderprotection. These failures provide the rationale for prudentialregulation. While prudential regulation improves prosperityand is likely to enhance effective competition, there can beinstances where regulation has unintended consequences. ThePRA’s secondary competition objective places an onus on it tofacilitate effective competition, whenever possible, to guardagainst the risks of such unintended consequences.

The PRA’s secondary competition objective does not, however,imply it should act as a competition regulator. Its role iscomplementary to, but distinct from, those of competitionregulators. Nevertheless, the focus on effective competitionin the PRA’s secondary objective makes clear that prudentialregulation and competition that improves prosperity arenormally aligned.

Prudential regulation requirements must be implemented in aproportionate way, in order not to stifle competition, forexample, from new and small firms. Nevertheless, trust inrobust but proportionate minimum prudential regulationstandards empowers competitive rivalry based on superiorefficiency, service quality and innovation. Effectivecompetition not only benefits customers, but also improvesover time the resilience of the financial system.

The practical implication of the secondary competitionobjective is that when the PRA considers new and existingpolicies, it analyses the effect of different policy options formitigating risks to safety and soundness and/or policyholderprotection, and considers to what extent these options wouldfacilitate effective competition. It will consider options tofacilitate effective competition where this would notundermine the objectives of safety and soundness andpolicyholder protection.

Since it came into effect, the secondary objective hasinformed the PRA’s approach. Examples highlighted in thisarticle include the capital framework for banks and insurers, aswell as the implementation of an FPC Recommendation inrespect of limiting the growth of high loan to incomemortgage lending.

The PRA has taken a proactive approach by considering theeffect of the existing regulation on the effectiveness ofcompetition, seeking to remove or alter regulation that isunnecessary for safety and soundness and/or policyholderprotection. Competition considerations can also inform ourapproach to international negotiations on both new rulesbeing designed or existing ones under review.


References

Bailey, A, Breeden, S and Stevens, S (2012), ‘The Prudential Regulation Authority’, Bank of England Quarterly Bulletin, Vol. 52, No. 4, pages 354–62, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/qb120405.pdf.

Bank of England (2014a), The Prudential Regulation Authority’s approach to banking supervision, available atwww.bankofengland.co.uk/publications/Documents/praapproach/bankingappr1406.pdf.

Bank of England (2014b), The Bank of England’s approach to resolution, available atwww.bankofengland.co.uk/financialstability/Documents/resolution/apr231014.pdf.

Bank of England and Financial Conduct Authority (2014), ‘A review of requirements for firms entering into or expanding in the banking sector:one year on’, available at www.bankofengland.co.uk/pra/Documents/publications/reports/2014/barriers2014.pdf.

Bank of England (2015), Open Forum: building real markets for the good of the people, available atwww.bankofengland.co.uk/markets/Documents/openforum.pdf.

Boyd, J H and De Nicolo, G (2005), ‘The theory of bank risk taking and competition revisited’, The Journal of Finance, Vol. 60(3), pages 1,329–43.

Broecker, T (1990), ‘Credit-worthiness tests and interbank competition’, Econometrica, Vol. 58(2), pages 429–52.

Chan, Y, Greenbaum, S J and Thakor, A V (1986), ‘Information reusability, competition and bank asset quality’, Journal of Banking and Finance,Vol. 10(2), pages 243–53.

Das, U, Davies, N and Podpiera, R (2003), ‘Insurance and issues in financial soundness’, IMF Working Paper No. 03/138, available atwww.imf.org/external/pubs/ft/wp/2003/wp03138.pdf.

Debbage, S and Dickinson, S (2013), ‘The rationale for the prudential regulation and supervision of insurers’, Bank of England QuarterlyBulletin, Vol. 53, No. 3, pages 216–22, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2013/qb130303.pdf.

Degryse, H and Ongena, S (2008), ‘Competition and regulation in the banking sector: a review of the empirical evidence on the sources ofbank rents’, in Thakor, A V and Boot, A W A (eds), Handbook of financial intermediation and banking, Elsevier, Amsterdam.

Dewatripont, M and Tirole, J (1994), ‘A theory of debt and equity: diversity of securities and manager-shareholder congruence’, Quarterly Journal of Economics, Vol. 109(4), pages 1,027–54.

French, A, Minot, D and Vital, M (2015), ‘Insurance and financial stability’, Bank of England Quarterly Bulletin, Vol. 55, No. 3, pages 242–57,available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q303.pdf.

Geroski, P A (1995), ‘What do we know about entry?’, International Journal of Industrial Organisation, Vol. 13(4), pages 421–40.

Marcus, A J (1984), ‘Deregulation and bank financial policy’, Journal of Banking and Finance, Vol. 8(4), pages 557–65.

Noss, J and Sowerbutts, R (2012), ‘The implicit subsidy of banks’, Bank of England Financial Stability Paper No. 15, available atwww.bankofengland.co.uk/financialstability/Documents/fpc/fspapers/fs_paper15.pdf.

Padilla, A and Pagano, M (1997), ‘Endogenous communication among lenders and entrepreneurial incentives’, The Review of Financial Studies,Vol. 10(1), pages 205–36.

Perotti, E and Suarez, J (2002), ‘Last bank standing: what do I gain if you fail?’, European Economic Review, Vol. 46(9), pages 1,599–622.

www.bankofengland.co.uk/financialstability/Documents/fpc/fspapers/fs_paper15.pdf

www.bankofengland.co.uk/pra/Documents/publications/reports/2014/barriers2014.pdf

www.bankofengland.co.uk/financialstability/Documents/resolution/apr231014.pdf

www.bankofengland.co.uk/publications/Documents/praapproach/bankingappr1406.pdf

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/qb120405.pdf


Trends in UK labour supplyBy Stuart Berry, Matthew Corder, Chris Duffy, Christopher Hackworth and Bradley Speigner of the Bank’sMonetary Analysis Directorate.

• The path of labour supply is a key consideration for the Monetary Policy Committee (MPC). Ithelps to determine the overall supply capacity of the economy and therefore the amount ofoutput that can be produced without generating excess inflationary pressure.

• Labour supply appears to have grown robustly since the financial crisis, as a greater number ofpeople have retired later and more part-time workers have sought full-time jobs.

• Nevertheless, the recovery in the demand for labour has led to a significant erosion of labourmarket slack in recent years, as unemployment, labour market participation and average hoursworked have all moved back towards their equilibrium levels.

Overview

Labour supply has an important bearing on monetary policyas it helps to determine how quickly the economy can growwithout generating inflationary pressure. Total hours workedfell sharply relative to the supply of labour following thefinancial crisis (see summary chart), leading to a substantialbuild-up of slack in the labour market as people becameunemployed or underemployed. The demand for labourrebounded subsequently, but labour supply also appears tohave grown robustly in recent years, suggesting that amargin of slack probably still remains in the labour market.

This article provides more detail on some of the analysis thathas fed in to the MPC’s recent assessments of the supply sideof the economy. It explains how, besides population growth,labour supply depends on the proportion of people wantingto work — the participation rate, the unemployment rate,and the average number of hours that people work.

Taking these elements of labour supply in turn, participationin the labour market has remained strong in recent years,despite the influence of an ageing population, as people haveworked longer, at least in part reflecting increased lifeexpectancy. Following the financial crisis, both younger andolder people appear to have become discouraged fromseeking work, but this cyclical effect seems to have sinceunwound.

Much of the substantial increase in unemployment followingthe crisis has since unwound as well. But the precise level atwhich the unemployment rate is likely to settle — the‘equilibrium’ unemployment rate — is unclear. On the onehand, given that the unemployment rate has not returned toits pre-crisis average despite the strong pickup in the demand

for labour, it could be argued that the equilibriumunemployment rate has increased slightly. On the otherhand, the persistent weakness of wage growth in 2013 and2014, despite the falls in unemployment, might point to alarger drag from slack, which could imply a lower equilibriumunemployment rate.

The number of hours people report that they want to workhas increased in recent years. Signs of underemploymentpersist, with an unusually high number of part-time workerswishing to move to a full-time job, for example.

Overall, this ‘bottom-up’ approach suggests that a modestamount of slack remains, concentrated in average hoursworked. Alternative, ‘top-down’, estimates of slack based onstatistical filtering techniques using information on the pathof output and other cyclical indicators also point to a smallmargin of slack remaining.

4

3

2

1

0

1

2

3

4

2000 02 04 06 08 10 12 14

Percentage change on a year earlier

+

–

Summary chart Four-quarter growth in total hoursworked

Topical articles Trends in UK labour supply 345

Introduction

The inflation outlook depends in part on the level of aggregatedemand relative to the economy’s capacity to produce goodsand services. The presence of spare capacity, or slack, is likelyto reduce price pressures whereas very intensive use ofproductive capacity is likely to increase them. The evolution ofthe supply capacity of the economy, therefore, determineshow quickly the economy can grow without generatinginflationary pressure.

The supply capacity of the economy depends on the resourcesavailable to produce goods and services, and how efficientlythose resources are used. One important component ofaggregate supply within an economy, therefore, is the amountof labour that is available. As the number of people availableto work increases, for example due to growth in the populationor changes in people’s desire to work, the economy is able toproduce more goods and services. The extent to which thatlabour is utilised will depend on the hiring plans of companies.When demand for a company’s products is strong they arelikely to employ more people, whereas in a period of weakdemand they are likely to reduce hiring or lay off workers, orcut their hours, leading to greater labour market slack.

The more slack there is in the labour market, the lower arelikely to be wage pressures, and therefore inflation. Theopposite will be true when the labour market is particularly‘tight’. So the balance between the demand for labour and itssupply is an important issue for monetary policy.

The May 2015 Inflation Report set out the views of theMonetary Policy Committee (MPC) on the outlook for laboursupply, and this assessment has underpinned its subsequentforecasts.(1) This article provides more detail on the recenttrends in labour supply that have informed the Committee’srecent assumptions on labour supply, and their implications forthe degree of slack in the labour market.

The first section of this article looks at the evolution of laboursupply in recent years and its role in facilitating increases inoutput. The second section sets out the different ways inwhich the supply of labour can vary. The third section looks atthe recent trends in each of the components of labour supplyin more detail, looking in particular at their evolution since thefinancial crisis. The fourth section briefly considers alternativeways of assessing the balance between demand and supply inthe economy, and the fifth concludes.

The role of labour in output growth since thefinancial crisis

In normal times, much of the growth in output in theUnited Kingdom has been facilitated by improvements in

productivity — the efficiency with which inputs can becombined to produce goods and services. Increases in laboursupply have typically played a smaller role. In the decadeleading up to the financial crisis, UK GDP growth averaged justunder 3% a year. On average, around three quarters of thatcould be accounted for by productivity growth (including theeffect of growth in the capital stock) measured as output perhour worked, with the rest coming from labour supply(Chart 1).

The picture has been quite different since the financial crisis.Output fell sharply in 2008 and 2009. Part of that wasaccounted for by a fall in the total number of hours worked, aspeople lost their jobs and unemployment rose, but there wasalso a substantial fall in productivity. Since then, outputgrowth has recovered, but that has been associated mostlywith an increase in the use of labour. GDP growth averagedaround 2¼% a year between 2010 and 2014, and aroundfour fifths of that was accounted for by increases in thenumber of hours worked (Chart 1).

The increased use of labour to deliver increases in output hasbeen associated with sharp falls in unemployment. And itscounterpart has been persistent weakness in productivitygrowth, whose causes have been the subject of a considerableamount of analytical work.(2) As labour market slack is usedup, it is likely that further sustained growth in output willrequire a recovery in productivity growth towards more normalrates.

(1) See the ‘Output and supply’ section of the May 2015 Inflation Report (pages 22–32);www.bankofengland.co.uk/publications/Documents/inflationreport/2015/may3.pdf.

(2) See Barnett, A, Batten, S, Chiu, A, Franklin, J and Sebastiá-Barriel, M (2014), ‘The UKproductivity puzzle’, Bank of England Quarterly Bulletin, Vol. 54, No. 2, pages 114–28;www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q201.pdf.

8

6

4

2

0

2

4

6

1997 2000 03 06 09 12 15

Hours worked

Productivity

Output (per cent) Percentage points

+

–


(a) The output series is based on the MPC’s best collective judgement about the final estimateof GDP, and the hours worked series is based on population data that have been adjusted forexpected revisions to ONS Labour Force Survey estimates to incorporate the latest data onnet migration.

Chart 1 Contributions to four-quarter UK GDP growth(a)

http://www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q201.pdf

http://www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q201.pdf

http://www.bankofengland.co.uk/publications/Documents/inflationreport/2015/may3.pdf


The outlook for both growth and inflation will depend in part,therefore, on trends in labour supply and on how much sparecapacity remains in the labour market. There are signs thatlabour supply has grown rapidly in recent years. That hasallowed the total number of hours worked to increase whilestill leaving some slack in the labour market. Decomposingmovements in the labour market into different componentscan help in assessing the likely future path of labour supplyand the amount of slack remaining in the labour market.

The components of labour supply

One standard measure of the overall amount of labouremployed in the economy is the total number of hours worked(which is the measure used in the section above). In anaccounting sense, there are four factors that determine howmany hours are worked across the economy in aggregate.First, the number of hours worked will depend on the size ofthe population. Second, only a subset of the population willwant to work. The proportion of the working-age populationthat wants to work (and is either employed or actively lookingfor work) is called the participation rate. Third, of those thatwant to work, there will always be some people who areunemployed at any given moment, for example as they movefrom one job to another. And finally, the number of hoursworked by those in employment can vary.

As Chart 2 highlights, in the pre-crisis period, populationgrowth accounted, on average, for all of the increase in totalhours worked, with a trend increase in participation offset by atrend decline in average hours worked. The number of peopleaged 16 and over rose fairly steadily at an average rate of 0.7%per year during the period 1997–2008, the same as theaverage growth of total hours worked over that period. As thefinancial crisis hit, the fall in total hours worked reflected a risein the unemployment rate as well as a decline in both theparticipation rate and the average hours worked by those inemployment. The recovery in the labour market has seen allof those effects unwind over time.

The implications of such movements in total hours workeddepend on the underlying trend for labour supply. Inflationarypressure depends in part on the degree of slack in the labourmarket, which can be thought of as reflecting the gap betweenthe current number of hours worked and their trend, orequilibrium, level. The latter refers to the amount of labourthat could be supplied without exerting excessive upward ordownward pressure on wage growth and inflation. It can beaffected by a wide range of factors, such as demographics,employment law and the tax and benefit system.

Changes in trend labour supply can come from any of itscomponents. For example, the population may grow morerapidly due to a sustained increase in inward migration, or anincrease in retirement ages could mean more people seeking

to work. Moreover, those people already in employment canvary the number of hours they want to work. If part-timeworkers would like to work full-time, for example, there isscope for total hours worked in the economy to increase.These decisions on different components of labour supply canalso be linked. For example, within a household of two ormore people, the decision to supply more labour could takethe form of either an additional person looking for work, whichwould affect the participation rate, or an existing workerlooking to move from a part-time job to a full-time one, whichwould affect average hours.

The degree of slack in the labour market, therefore, can bedecomposed into the gaps between the participation rate, theunemployment rate and average hours worked and theirrespective trends. Such a disaggregated approach can oftenreveal richer stories about developments in the labour market.

The next section looks at each of the components of thelabour market in turn, to assess how the underlying trends arelikely to have evolved and what that implies for the degree ofslack in the labour market.

Recent movements in trend labour supply

The trend or equilibrium for labour supply cannot be observeddirectly, making it difficult to judge the degree of slack. But itis possible to draw inferences based on other indicators,economic theory, and the relationship between labour marketquantities and other variables such as wages.

The main focus of this section is to assess recent trends inlabour supply, and what these imply for how much slack theremight still be in the labour market. But it also considers howtrends in labour supply might evolve in the future, as this willhelp to determine how quickly the economy can grow in thefuture without generating inflationary pressure. The evidence

5

4

3

2

1

0

1

2

3

4

1998 2000 02 04 06 08 10 12 14

Population

ParticipationUnemployment

Average hours

Hours worked (per cent)

Percentage points

+

–

Chart 2 Contributions to four-quarter growth in totalhours worked


across the different components is brought together at theend of this section to consider the likely evolution of overallslack in the labour market. Subsequent sections then look atalternative estimates of aggregate slack. In the November2015 Inflation Report, the Monetary Policy Committee’s bestcollective judgement was that spare capacity of around ½% ofGDP remained, but it also noted that there was a wide degreeof uncertainty around that estimate and a range of views onthe Committee.

PopulationPopulation growth is an important component of theevolution of labour supply and will help to determine howquickly output can grow without generating inflationarypressure. The population measure of most interest formonetary policy is the number of people aged 16 and over asthat represents the overall potential workforce. Growth in the16+ population tends to be relatively smooth, although it canbe affected by periods where the birth rate was unusually highor low: the ‘baby boom’ generation born following theSecond World War, for instance, boosted the workforcesignificantly once they entered adulthood.

Over the past decade, population growth has also beenboosted by an increase in net migration into theUnited Kingdom. While net inward migration eased followingthe financial crisis, it has picked up again over the past coupleof years, reaching 336,000 in the year to 2015 Q2 (Chart 3),equivalent to around ½% of the UK population. It is difficultto know how migration flows, and hence overall populationgrowth, are likely to evolve in future, but in recentInflation Report projections, the MPC has assumed thatmigration will gradually fall back towards its average over thepast decade.

The impact of migration on labour supply will depend on thecharacteristics of those arriving in the United Kingdom. Therecent pickup in net inward migration largely represented

people arriving having already secured a job or looking forwork, consistent with a cyclical response of net migration tochanges in relative job prospects in the United Kingdom andabroad. These migrants are likely, therefore, to boost thenumber of people flowing into the labour market and hencelabour supply.

Overall, the population of people aged 16 and over hasincreased by around 0.8% per year, on average, over the pastfive years. That growth is expected to slow to an average ofaround 0.7% per year over the next few years.

Participation in the labour marketThe proportion of the adult population willing and able towork also contributes to the overall supply of labour in theeconomy. Structural changes in the labour market can lead topersistent changes in the degree of participation. For example,the participation rate rose gradually between the mid-1990sand 2008, reflecting increases in female participation and inthe number of people retiring later (Chart 4). Butparticipation can also vary with economic conditions. Forexample, there may be less incentive for some people to lookfor work when unemployment is high and job prospects arepoor, so they may choose to ‘leave’ the labour market — thatis, cease to look actively for work — at least temporarily.Those people may nevertheless choose to return to work asjob prospects improve. These cyclical fluctuations do notaffect what we would think of as the ‘trend’ participation rate.Rather, the difference between actual and trend participationcan be thought of as the ‘participation gap’, which representsslack in the labour market. The challenge is to judge whichmovements in participation are long-lasting structural ones,and which are cyclical.

The cyclical response of participation to the financialcrisisBetween 2008 Q2 and 2010 Q1 the participation rate fell bynearly ¾ of a percentage point to a trough of 63%, asignificant fall equivalent to around 325,000 fewer peopleparticipating in the labour market. The rate has since

0

50

100

150

200

250

300

350

400

2005 07 09 11 13 15

Thousands over the previous year

(a) The data in the chart refer to flows over the preceding year. Prior to 2012 the data were onlycollected every six months or annually, rather than quarterly.

Chart 3 Net long-term inward migration(a)

61.5

62.0

62.5

63.0

63.5

64.0

1993 96 99 2002 05 08 11 14

Percentage of the 16+ population

0.0

Chart 4 The participation rate


recovered somewhat. Changes in participation rates fordifferent age groups suggest that those movements are likelyto have reflected cyclical factors. The initial decline mainlyreflected falls in participation among the young (18–24 yearsold) and older age groups (50–64 years old).

The fall in participation among the young would be consistentwith people staying in education for longer in the face of highunemployment. The number of students increased sharplyover that period. Participation within the 18–24 year old agegroup had been on a downward trend prior to the crisis, as theproportion of students in that group increased, but the fall in2008 and 2009 was much more marked (Chart 5). Inaddition, some older people may have retired early in responseto the deterioration in labour market conditions. There hadbeen an upward trend in participation among 50–64 year oldsprior to the crisis, but the participation rate flattened offsubsequently for a period.

More recently, these deviations of participation rates relativeto their pre-crisis trends have unwound for both age groups,consistent with the cyclical effects largely dissipating as labourmarket conditions improved. On the face of it, that suggeststhat there is relatively little slack remaining in the labourmarket as a result of cyclically low participation.

Another way to judge the participation gap is to look at thenumber of people that are not participating in the labourmarket but nevertheless say that they would like a job. Thesepeople can be seen as ‘discouraged’ workers given that theywant to work in principle, but are not actively seeking a job.The proportion of such people in the Labour Force Survey(LFS) picked up sharply in 2009, consistent with a cyclicalreduction in participation (Chart 6). Since then it hasgradually fallen back to close to its pre-crisis level, consistentwith a reduction in labour market slack.

The indicators discussed above suggest that the participationgap is currently quite small. Given that the actualparticipation rate is currently close to its pre-crisis level, thatwould also imply that the trend, or equilibrium, participationrate has been fairly flat since the crisis. So another way ofassessing whether the participation gap is small is to considerwhether a fairly flat trend over the past few years is plausible,which is discussed next.

Assessing the recent path of trend participationOther things being equal, demographic effects would probablyhave led to a fall in participation over the past few years.Older age groups tend to have lower participation rates asthey reach retirement age — the rate is just over 10% forthose aged 65+ and 72% for 50–64 year olds, compared with86% for the 35–49 age group. So it seemed likely that theproportion of the 16+ population working, or looking for work,would decline as the baby-boom generation started to reachretirement age. The proportion of the adult population aged50 and over has increased from 42% to 45% over the past fiveyears. But, in the event, that has been offset by the fact thatmany of them have retired later than previous generations. Asa result, the participation rates for those aged over 50 haveincreased substantially in recent years (Chart 7). That wouldpoint to a flatter trend for participation than the fall thatmight have been expected given the ageing of the population.

An attempt can be made to calibrate the effect of this andother factors driving the longer-term trend in the participationrate. Using disaggregated data on participation rates by ageand year of birth, it is possible to construct a ‘cohort model’that estimates the impact of factors such as life expectancy,demographics, health status, and education levels as well ascyclical factors such as unemployment and asset prices onparticipation rates for different generations, or cohorts, overtheir lifetimes.(1) By tracking those cohorts over time,

60

62

64

66

68

70

72

74

66

68

70

72

74

76

78

80

2000 03 06 09 12 15

Per cent Per cent

18–24 year olds (right-hand scale)

50–64 year olds (left-hand scale)

0 0


(a) The dashed lines indicate 2000–08 trends in the age-specific participation rates, extendedinto later years.

Chart 5 Participation rates for 18–24 and 50–64 yearolds(a)

4.0

4.5

5.0

5.5

6.0

6.5

7.0

1994 97 2000 03 06 09 12 15

Percentage of 16–64 population

0.0

Chart 6 ‘Discouraged’ workers: people that want a jobbut are inactive in the labour market

(1) The model uses LFS microdata and is similar to that used in Aaronson, D, Davis, J andHu, L (2012), ‘Explaining the decline in the U.S. labor force participation rate’, FederalReserve Bank of Chicago, Chicago Fed Letter No. 296, March.


alongside movements in the factors mentioned above, it ispossible to model the evolution of participation. And bystripping out the estimated effects of the cyclical factors it ispossible to get an estimate of the trend for participation.

The estimated effects of the various drivers of trendparticipation since 2007 are shown in Chart 8. Taking all ofthose effects together, the model would suggest a flat toslightly rising trend up to 2015, as shown by the orange line.There has indeed been a drag from the ageing population, butit has been offset by increases in life expectancy in particular.As life expectancy has increased, people have tended to workfor longer, perhaps encouraged by concerns over the adequacyof pension savings, and also by greater opportunities for lessphysical work. Greater educational attainment is typicallyassociated with higher participation rates, and so thecontinued shift towards a more highly educated population isalso estimated to have pushed up trend participation. Overall,a relatively flat trend over the past few years is consistent withthe other evidence presented in this section, suggesting thatthere is probably little slack remaining from the cyclicalreduction in the participation rate following the financial crisis.Over the next few years, a further drag from the ageingpopulation is likely to continue to be offset by increasing lifeexpectancy and rising education levels, suggesting a trend forthe participation rate that continues to be fairly flat. Thoseupward pressures on participation are likely to be reinforcedby structural changes such as increases in the state pensionage. Given the scale of the forces pushing in both directions,however, there is considerable uncertainty about the likelytrend for the participation rate.

Overall, this analysis suggests that the participation gap issmall. The flat trend for participation over the past few yearsthat this implies can be explained by the offsetting effects ofthe ageing population and increased life expectancy leading topeople working for longer.

UnemploymentThe unemployment rate is the most well-known indicator ofslack in the labour market. There will always be people thatare unemployed at a given point in time as it takes time forpeople to find the right job. This is sometimes referred to asfrictional unemployment, and represents the long-runequilibrium level of unemployment. But unemployment variesfor cyclical reasons as well, for example as weak labourdemand leads to an increase in lay-offs and less hiring. Thedifference between actual unemployment and estimates of itsequilibrium value is sometimes referred to as the‘unemployment gap’ and provides an indication of the degreeof slack implied by the current unemployment rate.

The LFS unemployment rate was relatively stable between2001 and the financial crisis at around 5% (Chart 9). It pickedup sharply in 2008 and 2009 as the recession hit. Over thepast two years it has fallen back again towards its pre-crisislevel as the demand for labour recovered, and there arereports of some skill shortages. The unemployment gapappears to have diminished substantially over that period,therefore, but estimating how big it is requires an estimate ofthe equilibrium rate.

A distinction can be made between the equilibrium rate ofunemployment in the medium and long run. The long-runequilibrium is the rate at which unemployment might beexpected to settle once all of the effects of shocks to theeconomy have faded away. The economy might only reachthis equilibrium after a long time. The medium-termequilibrium captures the idea that over the shorter time spansrelevant for monetary policy — of, say, a few years —persistently high unemployment can, for a period, itself affect

4

3

2

1

0

1

2

3

4

2007 0908 1110 12 14 16 1813 15 17

Demographics

Other cohort effectsResponsibility for children

Life expectancy

Health

Education levels

Trend

Percentage points5

+

–


(a) The chart shows the contribution of various structural factors to the change in the aggregateparticipation rate since 2007 based on the estimated impact of these factors on different agegroups, and changes in the age mix of the population over time.

Chart 8 Contributions to estimated trend participationsince 2007(a)

30

25

20

15

10

5

0

5

10

15

1993 96 99 2002 05 08 11 14

16–17 18–24 25–34

35–49 50–64 65+ Percentage points

+

–

Chart 7 Cumulative changes in participation rates since1993 for different age groups


the level of unemployment consistent with stable wagegrowth. That means that the unemployment gap that isrelevant for determining wage pressure is probably smallerthan that implied by the difference between actualunemployment and its long-run equilibrium value. These twoconcepts of equilibrium are considered in turn.

Long-run equilibriumThe long-run equilibrium rate of unemployment will dependon the structural characteristics of the labour market, such asthe degree to which the skills and experience of the pool ofpotential workers is a good match for the jobs that companieswant to fill, the incentives generated by the tax and benefitsystem for people to take jobs, and innovations in technologythat help workers find the right jobs and companies find theright people.

One simple metric for the long-run equilibrium is to look atthe pre-crisis period when the unemployment rate wasrelatively stable and wage pressures appeared contained,suggesting that there was little slack in the labour market.That would point to a long-run equilibrium rate of around 5%.

A more sophisticated approach makes use of economic theory.The ‘search and matching’ framework has become thestandard way of thinking about unemployment in amacroeconomic context. The intuition behind this frameworkis discussed in a box on page 351 but a key component of it isa relationship known as the Beveridge curve. There is typicallya negative relationship between the number of vacanciesadvertised by companies — which represent unfilled labourdemand — and unemployment. This can be seen in the bluediamonds in Chart 10. As the number of vacancies fell duringthe financial crisis, unemployment rose. But the chart alsoshows that the curve appears to have shifted out morerecently; the magenta diamonds are to the right of the bluediamonds.

One interpretation of this is that the unemployment rateconsistent with a given level of labour demand is higher than

before. That is, there could have been an increase in thelong-run equilibrium unemployment rate. But in practiceother factors can also shift the Beveridge curve temporarily.One such factor is a temporary rise in long-termunemployment, as it can be harder for people to find jobsquickly if they become more disconnected from the labourmarket (this is discussed in the section on the medium-termequilibrium rate below). Even taking into account the rise inlong-term unemployment, however, there appears to havebeen an outward shift in the Beveridge curve.

Combining that shift with another element of the search andmatching framework, known as the job creation curve (see thebox on page 351), might suggest that the long-run equilibriumhas risen from around 5% to 5½%. Alternatively, the pathtraced out by the Beveridge curve can also ‘loop’ during acyclical recovery as it takes time for the pickup in vacancies tofeed through to lower unemployment. The apparent outwardshift may prove to be temporary, therefore, so it is not clearthat the long-run equilibrium has shifted from its pre-crisislevel.

Moreover, there are a number of structural changes in thelabour market that may have pushed down on the long-runequilibrium rate, such as increased participation among olderage groups and improvements in educational attainment, bothof which are typically associated with lower unemploymentrates. Overall, the long-run equilibrium of around 5%assumed by the MPC seems to be a reasonable central case,but risks can be identified in both directions.

Medium-term equilibriumEven if the unemployment rate ultimately falls back to itspre-crisis average of around 5%, it is possible that the rate ofunemployment consistent with stable wage growth might, fora period after the financial crisis, have been higher than that.

0

2

4

6

8

10

12

14

1972 77 82 87 92 97 2002 07 12

Per cent

Chart 9 The unemployment rate

1.4

1.6

1.8

2.0

2.2

2.4

2.6

4 5 6 7 8 9

2002–12

2013–15

September 2015

Unemployment rate (per cent)

Vacancy rate (per cent)

0.0

(a) The vacancy rate is the stock of vacancies divided by the sum of the number of people inemployment and the stock of vacancies.

Chart 10 Beveridge curve(a)


A brief description of equilibriumunemployment theory

This box provides a brief outline of the ‘search and matching’model of equilibrium unemployment, which has become thestandard way of thinking about unemployment in theacademic literature.(1)

The theory is based on the idea that it takes time and effortfor job seekers to find the right job and for companies to findthe right employee — a process known as matching.Companies advertise vacancies and try to recruit people fromthe pool of available workers. The time taken to find the rightmatch means that there are always some people that areunemployed. But the greater the efficiency with which thematching process works, the more quickly new jobs will becreated and hence the lower the equilibrium unemploymentrate will be in normal times.

The search and matching framework tries to capture the flowof new jobs by looking at the relationship between vacancies— which represent unfilled labour demand — andunemployment. Cyclical variations in labour demand lead to anegative relationship between vacancies and unemployment,known as the Beveridge curve. A stylised example is shown inFigure A: when vacancies are unusually high, and companiesare trying to recruit lots of people, more job matches are likelyto be created and unemployment will tend to be low.Similarly, when labour demand and so vacancies are low,unemployment will tend to rise.(2)

On its own the Beveridge curve does not pin down theequilibrium unemployment rate. We need to know wherealong the Beveridge curve the labour market is likely to be innormal times. That will reflect the level of labour demand,and in particular companies’ decisions on how many vacanciesto advertise. Companies will post more vacancies if thepotential returns to filling a role are high. That is likely to bethe case when unemployment is high, because highunemployment depresses wages and recruitment costs arelikely to be low given the wider pool of available workers. Inthe search and matching framework, this leads to theupward-sloping ‘job creation curve’ in Figure A. Theintersection of the Beveridge curve and the job creation curvegives an estimate of the equilibrium unemployment rate. Atthis point, the flow of new jobs equals the flow of job lossessuch that the unemployment rate is stable.

Structural changes in the labour market can lead to a shift inthe Beveridge curve, and a change in the long-run equilibriumunemployment rate. For example, if the skills and experienceof the unemployed become less well suited to the jobs onoffer, it will become harder for companies to find the rightemployee and so job matches are likely to occur more slowly,even if their demand for labour remains unchanged. Such adeterioration in matching efficiency would shift the Beveridgecurve to the right in Figure A, and lead to an intersection ofthe curves at a higher equilibrium unemployment rate.

Vacancies

Unemployment

Equilibrium unemployment

Job creation curve

Beveridge curve

Figure A A stylised representation of equilibriumunemployment in the search and matching model

(1) See, for example, Pissarides, C (2000), Equilibrium unemployment theory, 2nd Edition,MIT Press, Cambridge MA.

(2) The actual location of the Beveridge curve can be estimated by looking at how therate at which jobs are created varies in response to cyclical shocks, alongside someassumptions about the rate at which jobs are lost.


This is because recessions are typically associated with a rise inlong-term unemployment. And people that have beenunemployed for a long period may not exert as muchdownward pressure on wages as the short-term unemployed,for example because they have become more disconnectedfrom the labour market. Indeed, as evidence of this, theproportion of long-term unemployed people moving intoemployment is much lower than that for the short-termunemployed.

One way to capture this effect is to calibrate a medium-termequilibrium unemployment rate that rises when long-termunemployment is unusually elevated (since more people mayhave become relatively more detached from the labourmarket). The relative likelihood of the long-term unemployedmoving into employment can be used to try to quantify thiseffect.

The persistent increase in unemployment following thefinancial crisis meant that, taken together, the unemploymentrates for those unemployed for six to twelve months and morethan twelve months picked up by around 2½ percentagepoints relative to their pre-crisis average (Chart 11). Based onthe relative likelihood of those groups moving intoemployment, the medium-term equilibrium might haveshifted up to around 6½% by the end of 2011, well above theestimates of the long-run equilibrium discussed earlier. Sincethen, much of that increase has unwound, suggesting that themedium-term equilibrium rate has fallen back towards itslong-run level.

An alternative way of estimating the medium-termequilibrium unemployment rate is to make use of therelationship between unemployment and wages. Whenunemployment is low relative to its equilibrium, pay pressuresare likely to be high as companies find it difficult to recruit the

right people and wages are bid up. This negative relationshipbetween wages and unemployment is often called thePhillips curve.

Statistical filtering techniques can be used to estimate a pathfor the equilibrium unemployment rate that best explains theobserved behaviour of wages, based on the Phillips curverelationship. The estimated equilibrium unemployment rate isallowed to vary over time in these models in order to capturepersistent changes in the equilibrium, for example due tochanges in the structural characteristics of the labour marketor ‘hysteresis’ effects from changes in long-termunemployment. But it is only allowed to vary in a smooth waythat avoids it being buffeted around by short-term volatility inthe data.

Different assumptions can be made about the shape of thePhillips curve in these models, and that can lead to differentestimates of the equilibrium unemployment rate. Forexample, the relationship can be assumed to be linear orcurved (specifically, convex).(1) It is also possible to add inextra information about the path of output that mightimprove the estimates of slack given that there is typically arelationship between output and unemployment.(2) Thepurple swathe in Chart 12 shows a range of estimates basedon different modelling assumptions. The latest estimatespoint to a medium-term equilibrium rate a little above 5%,close to the actual unemployment rate in 2015 Q3 of 5.3%.

Such estimates of the equilibrium unemployment rate aresubject to considerable uncertainty. This can be seen, for

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

2002 04 06 08 10 12 14

Up to six months

Six to twelve monthsOver twelve months

Per cent


(a) Dashed lines indicate 2002–07 averages.

Chart 11 Unemployment rates by duration(a)

0

2

4

6

8

10

12

1971 76 81 86 91 96 2001 06 11

Range of medium-term equilibrium unemployment estimates

Unemployment rate

Per cent


(a) The swathe includes three different estimates of the medium-term equilibriumunemployment rate: a baseline model; a model which assumes that the Phillips curve isconvex rather than linear and a model which incorporates additional information about thepath of output. The sample period is 1971 Q1 to 2015 Q2.

Chart 12 Estimates of the medium-term equilibriumunemployment rate from various statistical filteringmodels(a)

(1) A non-linear relationship would mean that a low unemployment rate pushes up onwages to a greater extent than a high unemployment rate pulls down on wages.

(2) This is known as Okun’s Law. When output is low, companies will tend to reduce theamount of labour they use leading to higher unemployment, and vice versa.


example, in earlier periods where the different estimatesdiverge more markedly. Moreover, the models have wideuncertainty bands of at least a percentage point on either side(based on 95% confidence intervals).

Business survey data and information from the Bank’s Agentscan provide an additional indication of the gap between actualand equilibrium unemployment. When companies report lowlevels of recruitment difficulties relative to normal, it meansthat they are finding it relatively easy to find the right peopleto fill their vacancies. That is likely to mean that there is slackin the labour market, with unemployment above itsequilibrium. A range of different survey measures suggest thatrecruitment difficulties were low following the financial crisis,but have since recovered, and are now at, or slightly above,their pre-crisis averages, consistent with there being relativelylittle slack in unemployment.

Overall, estimates of both the long-run equilibrium rate andthe medium-term equilibrium, which is more relevant forwage pressures, are close to the current unemployment rate,but it is important to stress the considerable uncertaintyaround any estimates of the equilibrium rate. Moreover, it ispossible that the econometric evidence presented here isbeing affected by the persistence of the cyclical downturn inrecent years. Given that unemployment has remainedelevated for a long period, the models are likely to attributesome of that to an increase in the equilibrium rate, but as thecyclical recovery continues, if unemployment were to fallfurther without generating excessive wage pressures suchestimates would fall back again.

Average hours workedPrior to the financial crisis, there had been a downward trendin average hours worked since at least the 1970s (Chart 13),and probably much longer. Estimates of average hours in the1850s were around 60 hours per week compared to around32 hours per week in the run-up to the financial crisis.(1) Therewas a sharper fall in average hours during the recession,consistent with reduced demand for labour, but since the endof 2009, average hours have increased by around 1¼% andare now at around their pre-crisis level.

Changes in actual average hours following the financialcrisisThere is a wide distribution in the number of hours peoplework and shifts in that distribution can have a significantimpact on the overall number of hours worked in theeconomy. One important feature of the recession was a shifttowards more part-time work. The share of part-timeemployment increased by around 2 percentage pointsbetween 2007 and 2010. Chart 14 shows the change inaverage hours worked relative to 2008 Q3, decomposed intochanges in full-time and part-time average hours and changesin the composition of full-time and part-time workers. The

shift in composition pulled down on overall average hours (theorange bars in Chart 14). The part-time share remainselevated, despite the increase in labour demand in recentyears.

Much of the increase in average hours over the past few yearscan be accounted for by people taking less holiday on averagethan they did previously. The LFS asks people about theirusual hours of work as well as their actual hours worked in aparticular period. In any given week, around 12% of those inemployment do not work any hours, and a further 25% reportthat they worked fewer hours than usual. In part that isbecause their hours or overtime tend to vary, but it alsoreflects other factors such as holidays, sickness or parentalleave.

As a result, reported average actual hours worked tend to bequite a bit lower than usual hours, and more volatile. Chart 15

(1) See the ‘three centuries’ data set available on the Bank’s website atwww.bankofengland.co.uk/research/Pages/onebank/threecenturies.aspx.

29

30

31

32

33

34

35

36

37

1972 77 82 87 92 97 2002 07 12

Number of hours per week

0

Chart 13 Average hours worked

Residual

Composition

Full-time men hours

Part-time women hours

Full-time women hours

Part-time men hours

Average hours

1.0

0.5

0.0

0.5

1.0

1.5

2.0

1992 96 2000 04 08 12

Average weekly hours

+

–

Chart 14 Contributions to changes in average hoursrelative to 2008 Q3


http://www.bankofengland.co.uk/research/Pages/onebank/threecenturies.aspx


shows that in recent years usual hours have remained lowerthan at the onset of the financial crisis, but this has been offsetby an increase in average hours due to fewer people sayingthey were on holiday. That could reflect changes inemployment arrangements that will persist, but it may also bedue a desire to make up for weak real income growth orpeople feeling less able to take holiday when labour marketconditions are poor.

UnderemploymentAn important question is whether movements in averagehours are structural, reflecting long-term shifts in behaviour,or temporary. Despite the recovery in average hours since therecession, there is evidence that they remain lower thanpeople would like — that is, some people appear to be‘underemployed’. The proportion of part-time workers whosay they would like a full-time job has remained elevatedrelative to its pre-crisis level, which might point to an ‘averagehours gap’ with hours below their trend, or equilibrium, level.

The LFS asks people about the number of hours they wouldlike to work, making it possible to get a more accurate gaugeof ‘desired hours’.(1) There are always people who want towork more hours, and always some that would like to workless. But in aggregate, the number of additional desired hourshas increased substantially since the financial crisis andremains elevated. If people were able to change their hours towhat they report to be their desired level in full, that would beequivalent to increasing overall average hours from theircurrent level of 31.9 to 32.3, an increase of around 1¼%(Chart 16). The increase in desired hours has mainly reflectedpart-time workers that would like a full-time job. In large part,therefore, underemployment appears to be concentrated inthis group.

The amount of slack in average hours will depend on whetherthis stated desire to work more hours translates into actualincreases in hours over time, and there are reasons why thatmight not happen in full. For example, some of the desire towork more hours may reflect the sharp fall in real incomesduring the recession or concerns about the job prospects ofother members of a household. As real incomes and labourmarket conditions recover, the incentive to work more hoursmay be reduced for some people.

There is also empirical evidence that not all of the reporteddesired increase in hours tends to feed through into actualhours. During 2013–14 those people that went from beingunderemployed to reporting that they were happy with theirhours did so by increasing their weekly hours, on average, byonly five, even though they had said they would prefer to worktwelve extra hours.(2) That would suggest that only aroundhalf of the increase in desired hours might feed through intothe trend for average hours.

The future evolution of trend average hoursOver time, one factor that might pull down on the trend foraverage hours is demographic effects. Older workers tend towork fewer hours than most other age groups, as the blue linein Chart 17 shows. As the share of the workforce in older agegroups increases, aggregate average hours might therefore beexpected to fall. Although since far fewer older people(especially those aged over 65) work at all, the downwardtrend in average hours from demographics is likely to be small.

(1) These calculations are based on Bell, D and Blanchflower, D (2013), ‘How to measureunderemployment?’, Peterson Institute for International Economics Working Paper No. 13–7.

(2) This is based on looking at how the desired and actual hours of individuals in theLabour Force Survey change over the course of a year. This analysis was originallyundertaken in Weale, M (2014), ‘Slack and the labour market’;www.bankofengland.co.uk/publications/Documents/speeches/2014/speech716.pdf.

1.0

0.5

0.0

0.5

1.0

1.5

2.0

1994 98 2002 06 10 14

Second jobWorking more than usual

Less than usual — other

Usual hours

Less than usual — leave or other holiday

Actual hours

Hours per week

+

–


(a) The data shown are four-quarter moving averages.

Chart 15 Contributions to changes in average hoursrelative to 2008 Q3(a)

30.5

31.0

31.5

32.0

32.5

33.0

2002 04 06 08 10 12 14

Desired hours

Actual hours

0.0

Number of hours


(a) Desired hours are the number of hours that the currently employed report that they wouldlike to work, on average per week, calculated from LFS microdata, which have beenseasonally adjusted by Bank staff.

Chart 16 Actual and desired average hours(a)

http://www.bankofengland.co.uk/publications/Documents/speeches/2014/speech716.pdf


One other factor that might have an effect on the trend is thefiscal consolidation. Average hours for public sector workersare slightly lower on average than for private sector workers,so as the mix of employment shifts towards the private sectorthat could push up aggregate average hours. But again, theeffect here is likely to be very small.

Overall, therefore, the evidence presented would suggest thatthere has been a flat to rising trend in average hours over thepast few years, in contrast to the downward trend seen inprevious decades. That would imply that there is a modestamount of slack remaining in this component of labour supply.A slight downward trend may be resumed in the coming yearsdue to demographics, and there is a risk that the trend couldfall more sharply if improving economic conditions reduce theincentive to work longer hours.

Overall slack in the labour marketThe evidence presented in this section suggests that thegrowth in the supply of labour in recent years has been robust.Population growth has been strong and, despite some dragfrom demographic effects, the trends in participation andaverage hours are likely to have been flat or rising.

Given the sharp fall in total hours worked following thefinancial crisis, those trends would imply that a margin of slackopened up in participation, unemployment and average hours.Taken together, the overall amount of slack in the labourmarket appears to have been substantial, and was likely tohave been providing a material drag on inflation. Since then,labour market conditions have improved substantially, andmuch of that slack appears to have eroded. Nevertheless, amodest amount of slack is likely to remain, concentrated inaverage hours.

Given the considerable uncertainty around any estimate of thetrends or equilibria in the labour market, it is difficult to put a

precise estimate on the remaining slack. In the face of thatuncertainty, it is useful to look at a wider range of evidencewhen coming to an overall view on the likely amount of slackin the economy.

Alternative ways of estimating slack

The analysis set out so far is only one way to assess theamount of slack in the economy. This section outlines brieflytwo other approaches that have also been important ininforming the views of the Monetary Policy Committee.

Using wages as a signal of labour market slackAn important channel through which slack in the labourmarket affects inflation is via wage pressures, so it is possibleto use pay developments as a signal of how much slack mightremain in the labour market. Information on wages was usedin the estimates of the medium-term equilibriumunemployment rate based on statistical filters discussedearlier in this article. But it is also possible to estimate somesimple equations that try to capture the effects of overall slackin the labour market on pay, alongside other factors such asproductivity and inflation. If pay growth deviates substantiallyfrom the path predicted by such equations, that might suggestthat slack is higher or lower than is being assumed in theequation.

Pay growth fell sharply in 2008 and 2009 as slack in thelabour market rose and productivity fell. It remained weak forseveral years thereafter, even as unemployment began to fallsharply in 2013 and 2014 and slack in the labour market wasbeing eroded (Chart 18). Wage growth was weaker thanwould be predicted by simple wage equations, and that couldimply that there was more slack than implied by thebottom-up analysis in the previous section (which was used tocalibrate the measure of slack in the wage equations). Thereare, of course, other reasons why wages may have been weakover that period and wage growth has picked up morerecently, but this may point to a risk that there is a moresubstantial degree of slack remaining than suggested by theevidence presented earlier in this article.

Top-down estimatesFocusing solely on the labour market when analysing theamount of slack in the economy does not take into accountthe intensity with which the resources within firms — theirlabour and capital inputs — are being used. If those resourcesare being used less intensively than usual, that could representan additional source of slack, as companies would be able toexpand their output without the need for additional workers.Evidence from business surveys can provide an indication ofhow intensively resources are being utilised relative to normal.Another approach is to try to decompose the path of outputinto ‘trend’ and ‘cyclical’ components, to give a measure of the

0

5

10

15

20

25

30

35

40

0

10

20

30

40

50

60

70

80

90

100

16 26 36 46 56 66 76

Participation rate (right-hand scale)

Average hours (left-hand scale)

Per cent Average weekly hours

Age


(a) Average participation rates and hours by age group over the period 1995 to 2015 Q3.

Chart 17 Participation rates and average hours by age(a)


output gap, which can be done using simple statistical filters.A problem with this approach, however, is that the persistentweakness in output since the crisis may mean that too muchof that weakness is interpreted as a lower trend.

One way to improve on such techniques is to augment thefilter with other information, such as wage and price inflation,which can help to identify the cyclical component. Chart 19shows the path for the output gap from one such model, witha negative number indicating a degree of slack in theeconomy.(1) It suggests that the amount of slack narrowedsubstantially between 2012 Q4 and 2015 Q2, and is consistentwith a relatively small amount of slack remaining. However,the uncertainty around such estimates is wide, as illustratedby the 90% confidence interval shown by the dashed lines inChart 19.

Conclusion

The path of labour supply is a major component of the overallsupply capacity of the economy, and will help to determinehow much slack there is in the labour market. That, in turn,has an important bearing on monetary policy, through itsimpact on wage pressures and hence inflation.

Bottom-up analysis of different components of the labourmarket, such as the participation rate, the unemployment rateand the average hours worked by those in employment canshed light on the underlying trends in those variables. Laboursupply growth appears to have been robust since the financialcrisis. In particular, participation among older age groups hasincreased, and in aggregate people in employment appear towant to work more hours than they currently do. A strongtrend for labour supply implies a substantial build-up of slackin the labour market between 2008 and 2011 as actual totalhours worked fell sharply. More recently, as the demand forlabour recovered, much of that slack appears to have beenused up, although a small amount is likely to remain.

There is substantial uncertainty around any estimate of slackin the economy, and it is prudent to draw on a range ofevidence. Different analytical approaches can give differentsteers: the persistent weakness of wages in 2013 and 2014highlighted the risk that there may be more slack thansuggested by the bottom-up evidence, for instance, whereastop-down estimates based on statistical filters point to arelatively small amount of slack remaining. In theNovember 2015 Inflation Report, the Monetary PolicyCommittee’s best collective judgement was that sparecapacity of around ½% of GDP remained, focused in thelabour market, but it also noted that there was a wide degreeof uncertainty around that estimate and a range of views onthe Committee.

(1) The model includes five variables: GDP growth, the unemployment rate, a measureof long-term unemployment, CPI inflation excluding food and energy and a financialconditions index. The Kalman filter is used to decompose the path of output intotrend and cyclical components.

0

1

2

3

4

5

6

7

8

9

4

2

0

2

4

6

8

2002 04 06 08 10 12 14

Percentage change on a year earlier

Unemployment rate (left-hand scale)

Whole-economy pay (right-hand scale)

+

–

Per cent

Chart 18 Unemployment and pay

6

4

2

0

2

4

6

8

1995 98 2001 04 07 10 13

Percentage of GDP

+

–

Sources: Bloomberg, ONS and Bank calculations.

(a) The dashed lines represent a 90% confidence interval around the central estimate,illustrating the uncertainty around such estimates. A negative number indicates slack in theeconomy.

Chart 19 Estimate of the output gap from a top-downfilter model(a)

Topical articles The impact of higher interest rates and fiscal measures 357

The potential impact of higher interest ratesand further fiscal consolidation onhouseholds: evidence from the 2015NMG Consulting surveyBy Philip Bunn and Lizzie Drapper of the Bank’s Structural Economic Analysis Division, Jeremy Rowe of the Bank’sMonetary Assessment and Strategy Division and Sagar Shah of the Bank’s Macro Financial Risk Division.

• The balance sheet positions of mortgagors have improved modestly over the past year.

• Households appear a little better placed to cope with a rise in interest rates than a year ago andsurvey responses do not imply that a rise in rates would have an unusually large impact onspending.

• The survey results suggest that the Government’s plans for fiscal consolidation are likely tocontinue to weigh on household spending.

Overview

The financial situation of households is a key determinant ofhow they respond to changes in the economy, and how theyadjust to such changes has important implications for bothmonetary policy and financial stability. Results from thelatest NMG survey of households point to a modestimprovement in mortgagors’ balance sheet positions overthe past year. The proportions of households with highmortgage debt to income and debt-servicing ratios havefallen slightly, while reported levels of financial distress arelow and have declined a little further for both mortgagor andrenter households. Using the survey responses, this articleassesses how households might respond to (i) a hypotheticaland immediate 2 percentage point rise in interest rates(although in practice any increase is likely to be more gradualand limited than that) and (ii) the Government’s plans forfiscal consolidation.

Higher interest rates would increase the financial pressureon some households, but households are in a slightly betterposition to cope with a rise in interest rates than they werea year ago. The survey results suggest that, if interest rateswere to rise, borrowers would cut spending by more foreach extra pound of interest payments than savers wouldraise spending for each extra pound of interest received.In other words, borrowers have higher marginal propensitiesto consume (summary chart). Those marginal propensitiesto consume are close to previous assumptions made byBank staff and do not imply that an increase in interest rates

would have an unusually large effect on aggregatehousehold spending.

Around a third of households reported that they have beennegatively affected by the fiscal consolidation over the pastfive years. The most common response has been to cutspending. Households continue to expect to be affected byfiscal consolidation in the future. And there are somehouseholds who may be vulnerable to higher interest ratesand who expect to be more heavily affected than average byfurther fiscal consolidation, although that group is relativelysmall.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Borrowers Savers

Average marginal propensities to consume

Sources: NMG Consulting survey and Bank calculations.

(a) See footnote to Chart 10 for more details on how this chart is constructed. The chart showsdata from the 2015 H2 survey.

Summary chart Estimated marginal propensity toconsume of different types of households(a)


Introduction

The financial situation of households is a key determinant ofhow they respond to changes in the economic and policyenvironment, and how they adjust to such changes may haveimplications for both monetary policy and financial stability.The health of households’ finances will affect the extent towhich they change their spending in response to thesedevelopments. It can also help to determine how likelyhouseholds are to suffer financial distress which, in turn, canaffect whether lenders incur losses on loans to thosehouseholds.

Aggregate data can provide only a limited assessment of thestate of households’ finances. There are large differences indebt positions between households. These differences willaffect their likelihood of facing financial distress and how theyrespond to shocks. It is therefore important to examinedisaggregated data at the household level to help understandthese differences. Household survey data can also provideuseful information on households’ views on a range of topics,including the extent to which they are prepared for policychanges.

Between 2 and 22 September 2015, NMG Consulting carriedout an online survey of around 6,000 UK households on behalfof the Bank of England. Each year since 2004, the Bank hascommissioned NMG Consulting to conduct a householdsurvey during September, with additional April surveys in 2014and 2015.(1) This article analyses the results from the latestsurvey. The box on page 359 provides more detail on thesurvey methodology.

The NMG survey includes questions on households’ balancesheet positions, income, financial distress and influences onspending. These questions are typically included in each waveof the survey. The results are available on a timelier basis thanthe results from similar questions included in other householdsurveys. In addition, each wave of the NMG survey typicallyalso includes more bespoke questions on areas of policyinterest.(2)

The recent financial crisis had a severe impact on the financialsituation of many households. In the early stages of therecovery, the NMG survey was used to explore factors thatmay have restrained household spending growth. Thesefactors included weak income growth, tight credit conditions,concerns about debt levels, the effects of fiscal consolidationand uncertainty about future income.(3) As the recovery hasprogressed, real income growth has strengthened, creditconditions have improved, and concerns about debt levelshave fallen. However, as set out in the 2015 AutumnStatement, the fiscal consolidation that has been under waysince 2010 is set to continue. Furthermore, at the time of theMonetary Policy Committee’s (MPC’s) meeting in

November 2015, financial markets expected that an increasein Bank Rate to 0.75% would occur by the fourth quarter of2016. And financial markets expected Bank Rate to reach1.5% in three years’ time.

This article uses responses to the latest NMG survey to assesshow households might respond to higher interest rates andfurther fiscal consolidation. The article begins by describingrecent developments in household balance sheets andfinancial distress. That sets the context for how householdsmay respond to future developments in the macroeconomicand policy environment, which is the focus of the second partof the article. Last year’s Bulletin article on the NMG survey(Anderson et al (2014)) included a detailed assessment of theimplications of higher interest rates. Some of that analysis isupdated in this article and is used to assess how preparedhouseholds are for an increase in interest rates and whetherthat has changed over the past year.

Developments in households’ balance sheets

The stock of household debt rose rapidly in relation to incomein the decade before the financial crisis, increasing fromaround 105% in 1999 to around 155% in 2008 (Chart 1). Ithas since fallen back a little, in part due to a reduction in theproportion of households with a mortgage, but remainedrelatively high at around 135% in 2015 Q2. And as reported in

(1) The data from the latest survey and all previous surveys are available atwww.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/nmgsurvey2015.xlsx.

(2) In the latest survey, questions were included on the implications of higher interestrates, the impact of fiscal consolidation, factors affecting housing market activity andthe buy-to-let market.

(3) For example, the 2012 Quarterly Bulletin article examines how those factors mayhave affected spending. See Bunn et al (2012).

6

7

8

9

10

11

12

13

14

90

100

110

120

130

140

150

160

1999 2003 07 11 15

Per cent

Total DTI ratio(a)

(right-hand scale)

Mortgage DSR(b)

(left-hand scale)

Per cent

00


(a) Household gross debt as a percentage of a four-quarter moving sum of household disposableincome. Includes all liabilities of the household sector except for the unfunded pensionliabilities and financial derivatives of the non-profit sector. The household disposableincome series is adjusted for financial intermediation services indirectly measured (FISIM).

(b) Household mortgage DSR is calculated as mortgage interest payments plus mortgageprincipal repayments as a percentage of household disposable income. The householddisposable income series is adjusted for FISIM.

Chart 1 Aggregate household debt to income (DTI) ratioand mortgage debt-servicing ratio (DSR)

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/nmgsurvey2015.xlsx

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/nmgsurvey2015.xlsx


Survey method

Introduction and methodologyThe latest NMG survey was carried out online between 2 and22 September, covering 6,007 households in Great Britain.The survey was conducted annually during Septemberbetween 2004 and 2013, with April and September surveys in2014 and 2015. The survey has been run online since 2012,following pilots in 2010 and 2011. Before that, the survey wasface-to-face.

Moving the NMG survey online facilitated the introduction ofa panel element where the same households were asked torespond to successive surveys. Around half of the respondentsto the latest survey had completed a previous survey. Unlessotherwise stated, this article reports results from thecross-sectional data.

The NMG survey has a number of advantages relative to otherhousehold surveys. It is more timely than many other surveys;it may be better at measuring financial distress by virtue ofbeing conducted online where respondents are more willing todisclose sensitive information about their finances; and itcontains questions on topical policy issues that are not oftenavailable in other surveys.

A drawback of the NMG survey is that there may be a greaterrisk of selection into the survey based on unobservablecharacteristics than is the case for some other householdsurveys. The survey is weighted to be representative of theage, gender, region, housing tenure and employment statusdistributions of Great Britain. But, because the sample isdrawn from the Research Now panel used by the surveyprovider rather than the population as a whole (which istypically the case for surveys conducted by the ONS) theremay be a risk that certain types of people are more likely torespond to online surveys and be part of that panel, whichcould bias the results. However, in any survey, even when theprobability of being invited to complete the survey is known,certain types of households may still be more or less likely torespond. The NMG survey data do follow broadly similartrends to the aggregate data and other surveys in manyrespects and so are still likely to be a useful source ofinformation on distributional issues given the advantagesdescribed above.

Different approaches to asking about householdincomeThe main methodological change introduced in the latestsurvey was around how households are asked to report theirincome. In previous years the NMG survey has always askedhouseholds to report only their total income. However,average household income in the NMG survey has typically

been lower than reported in other household surveys, whichtend to ask more detailed questions about the components ofhousehold income and then aggregate those up. It is possiblethat some respondents to the NMG survey forgot to reporteither some components of income or the income of somehousehold members.

Tests in the Spring 2015 NMG survey asked households toreport income using four different approaches that wererandomly allocated across respondents. Average income washigher when asking about the income of each householdmember separately, reflecting an increase in the number ofhouseholds in the upper part of the income distribution, whileasking about the components of income made less difference,but had a notable impact on response rates (Table 1). For theAutumn 2015 survey, only the previous approach of asking fortotal income and the one asking for the total income of eachhousehold member separately were retained, with the samplesplit in half. Again, similar differences were observed, withhigher average income when asking about each member(Table 1) suggesting that some households do not include theincome of all members when just asked for the total, althoughat the expense of a modest fall in response rates.

Changing the income question introduces a discontinuity inthe data for households presented with the new approach.Over the longer term this should help to improve the accuracyof the survey, and this new approach is likely to be extendedto all respondents in the next survey. But where year-on-yearcomparisons that involve income are made in this article, onlythe half of households that were asked the old incomequestion variant are used. That is to ensure that the reportedfigures reflect genuine changes and not methodologicaldifferences.

Table 1 Average income from experiments using different incomequestions

Income question 2015 H1 mean 2015 H2 mean 2015 H1 income (£) income (£) non-response rate (per cent)(a)

(1) Total household income only(b) 33,106 33,832 13

(2) Total household income for each household member(c) 39,140 38,511 22

(3) Income components for household only(d) 34,912 n.a. 33

(4) Income components for each household member(e) 38,365 n.a. 39


(a) Share of respondents that started the survey but either did not complete the survey or did not answer thehousehold income question.

(b) Question: ‘Please state the total annual income of the whole of your household, before anything isdeducted for tax, National Insurance, pension schemes etc’.

(c) Question: ‘Please state the total annual income of each adult in your household, before anything isdeducted for tax, National Insurance, pension schemes etc’.

(d) Question: ‘Please state the annual income of the whole of your household (before anything is deducted fortax, National Insurance, pension schemes etc) which comes from each of the following categories:[employment; self-employment; benefits; private pension; investments]’. Only asked in 2015 H1 survey.

(e) Question: ‘For each member of your household please state the annual income (before anything isdeducted for tax, National Insurance, pension schemes etc) which comes from each of the followingcategories: [employment; self-employment; benefits; private pension; investments]’. Only asked in2015 H1 survey.


the July 2015 Financial Stability Report, it remains highcompared with other developed economies.

Over the past year, aggregate total household debt has grownat a similar rate to the National Accounts measure ofhousehold income. Within that, the stock of unsecured debthas grown faster than mortgage debt, although mortgage debtstill accounts for around three quarters of the stock of totalhousehold debt. The overall proportion of household incomespent on servicing mortgage debt is currently low, given thelow level of interest rates, and has been little changed sincelate 2014 (Chart 1).

In the latest NMG survey, the average amount outstanding onunsecured loans was broadly unchanged from the 2014 survey,with a small increase in the amount outstanding on securedloans. In the 2015 survey, households reported that theaverage outstanding mortgage was around £85,000. Amongthose households with unsecured debt, the average amountoutstanding was £8,000.

There were changes to the income questions asked to somehouseholds in the latest survey, which are described in moredetail in the box on page 359. But for those asked the samequestion, average household income before tax was slightlyhigher in the latest survey than a year ago.

Measures of household debt sustainability

(i) MortgagorsTwo indicators that are commonly used to assess thesustainability of household balance sheet positions are thedebt to income (DTI) ratio and debt-servicing ratio (DSR),which is the proportion of income spent on debt repayments(including both capital and interest payments). With bothindicators, a higher ratio is typically associated with a higherrisk of financial distress, particularly if incomes were to fall orrepayment costs were to increase from their current levels.

The share of households with a very high mortgage DTI ratio(above five) has fallen back since 2012 (Chart 2). That share isnow back to levels seen in the 1990s. The proportion ofhouseholds with a moderately high mortgage DTI ratio(between three and five) has been more stable recently, and isnow at similar levels to those seen in the mid-2000s.

The share of households spending a high proportion of incomeon mortgage repayments has fallen a little over the past yearand, given lower interest rates, is a much smaller share than itwas in 2008. The proportion of households with mortgageDSRs between 30% and 40% has dropped over the past year(the left panel of Chart 3), although the share of householdswith a mortgage DSR ratio of 40% or above was broadlyunchanged at just over 1%.

In addition to mortgage debt, around 70% of households witha mortgage also held at least one form of unsecured debt inthe latest survey. Although average balances on unsecuredloans are usually small relative to mortgage debt, repaymentsfor these loans can be significant relative to income. This isbecause interest rates on unsecured loans tend to be high andbecause repayment periods are usually relatively short.

If both secured and unsecured debt repayments are taken intoaccount, the proportion of mortgagor households with DSRsabove 30% increases materially (comparison of left and rightpanels on Chart 3). On average, unsecured debt repayments

0

1

2

3

4

5

1992 97 2002 07 12

Percentages of households

Between three and four

More than five

Between four and five

(b)

Sources: Living Costs and Food (LCF) Survey, NMG Consulting survey and Bank calculations.

(a) Ratio of outstanding mortgage debt to pre-tax annual income.(b) Data up to 2013 are based on responses to the LCF survey. Data for 2014 and 2015 are

based on responses to the NMG Consulting survey and have been spliced onto the earlierLCF survey data series in 2013. 2014 and 2015 NMG data are from the H2 surveys only.

Chart 2 Distribution of mortgage DTI ratios(a)

0

2

4

6

8

10

12

20%–30% 30%–40% 40%+ 20%–30% 30%–40% 40%+

Percentages of households(b)

Mortgage-only DSR Total DSR(including unsecured)(c)

200820122013

20142015


(a) The mortgage DSR is calculated as total mortgage payments (including principalrepayments) as a percentage of pre-tax income. The total DSR is calculated as totalmortgage and unsecured debt payments (including principal repayments) as a percentage ofpre-tax income. Reported repayments may not account for endowment mortgage premia.

(b) The chart shows the number of mortgagors within a particular DSR band as percentage of allhouseholds (including non-mortgagor households). 2014 and 2015 NMG data are from theH2 surveys only.

(c) The total DSR of mortgagors who were unsure or preferred not to state if they hadunsecured debt is based only on their mortgage repayments. Households who were unsureor preferred not to state the value of their debt repayments, or who reported unsecured debtrepayments larger than their outstanding unsecured debt, were not used in calculating theDSR distribution for each housing tenure group.

Chart 3 Distribution of DSRs for households withmortgages(a)


accounted for around a quarter of the total debt-servicingcosts for mortgagors in the 2015 survey, rising to around athird for those with total DSR ratios of above 40%.

Once unsecured debt repayments are taken into account, theproportion of mortgagors with DSRs above 30% has still fallenover the past year (Chart 3). Consistent with thisimprovement in balance sheet positions, fewer mortgagorsreported concerns about their debt or problems paying fortheir accommodation (Chart 4).

(ii) All householdsDebt is held not just by households who own their homewith a mortgage. Renters and outright homeowners alsoneed to be considered too. Unsecured debt accounts for allof the debt-servicing costs of these households. Forty per centof outright owners and around 70% of renters reported thatthey held at least one form of unsecured debt in the2015 survey.

Including non-mortgagors, around 5% of all households had atotal DSR of 40% or above in the latest survey (Chart 5).A similar proportion had DSRs of between 30% and 40%. Theproportion of all households with a total DSR of 30% or abovehas fallen a little since the 2014 survey. However, the share ofnon-mortgagor households (renters and outright owners) withDSRs above 30% was little changed.

Large debt repayments relative to income may be a particularburden for renter households, who also have to pay rent out oftheir income. Nevertheless, the proportion of rentersreporting their unsecured debt repayments to be a heavyburden has fallen in recent years (Chart 6).(1)

Responses to developments in policy and themacroeconomic environment

The current state of households’ finances will be an importantfactor affecting how they respond to future changes in themacroeconomic environment. Two important changes thatare likely to affect households over the next few years areincreases in interest rates and the Government’s plans forfiscal consolidation. The impact of those changes on theeconomy will depend, in part, on how households react tothem. This section considers how households might respondbased on the latest NMG survey.

0

5

10

15

20

25

30

35

40

45

1991 93 95 97 99 2001 03 05 07 09 11 13 15

Percentages of mortgagors

Unsecured debt repayments a heavy burden(c)

Difficulty with accommodation payments(b)

Very concerned about debt(d)

(a)

Sources: British Household Panel Survey (BHPS), NMG Consulting survey and Bank calculations.

(a) Data from 2011 onwards are from the online NMG survey. Data for 2014 and 2015 are fromthe H2 surveys only. Data from 2005 to 2010 are from the face-to-face NMG survey. Datafrom 1991 to 2004 are from the BHPS. Data from the BHPS and face-to-face NMG surveyshave been spliced to match the online NMG survey results.

(b) Question: ‘Many people these days are finding it difficult to keep up with their housingpayments. In the past twelve months would you say you have had any difficulties paying foryour accommodation?’.

(c) Households with unsecured debt were asked ‘To what extent is the repayment of these loansand the interest a financial burden on your household?’.

(d) Question: ‘How concerned are you about your current level of debt?’.

Chart 4 Measures of financial distress for mortgagors

200820122013

20142015

0

5

10

15

20

25

30

35

>0–10 10–20 20–30 30–40 40+


Total DSR(b) (per cent)


(a) 2014 and 2015 NMG data are from the H2 surveys only.(b) See footnotes (a) and (c) of Chart 3 for a description of the calculation of total DSR.

Chart 5 Distribution of DSRs for all households(a)

0

5

10

15

20

25

30

35

40

1991 93 95 97 99 2001 03 05 07 09 11 13 15

Percentages of renters

Unsecured debt repayments a heavy burden(c)

Difficulty with accommodation payments(b)

Very concerned about debt(d)

(a)

Sources: BHPS, NMG Consulting Survey and Bank calculations.

(a) Data from 2011 onwards are from the online NMG survey. Data for 2014 and 2015 are fromthe H2 surveys only. Data from 2005 to 2010 are from the face-to-face NMG survey. Datafrom 1991 to 2004 are from the BHPS. Data from the BHPS and face-to-face NMG surveyshave been spliced to match the online NMG survey results.

(b) See Chart 4 footnote (b) for question asked.(c) See Chart 4 footnote (c) for question asked.(d) See Chart 4 footnote (d) for question asked.

Chart 6 Measures of financial distress for renters

(1) Fewer households reporting that their unsecured debts are a burden is also consistentwith recently released early indicators from the ONS Wealth and Assets Survey,which cover the period up to June 2015.


At the time of the MPC’s meeting in November 2015, anincrease in Bank Rate to 0.75% was fully priced into financialmarkets by the fourth quarter of 2016. The MPC alsoindicated at their November meeting that any increases inBank Rate were expected to be gradual and to a lower levelthan in previous cycles. The impact of a given rise in interestrates will, in part, depend on how much households changetheir spending in response and on the extent to which that risein rates leads to any increase in financial distress.

Since 2010, the Government has announced a succession ofmeasures in order to cut the United Kingdom’s budget deficit,some of which have already been implemented and some ofwhich will take effect over the coming years. In theAugust 2015 Inflation Report, the MPC noted that the overalldrag on economic activity from the consolidation isuncertain.(1) Measures such as changes to public sector pay ortaxes may take longer to feed through to economic activitythan other measures, such as lower public investment, and theeffects will depend on how households and companies reactto changes in their income.

Although the section below considers the implications ofhigher interest rates and further fiscal consolidation in turn,they should not be considered independently as they are likelyto affect some of the same households. The extent of thefiscal consolidation and its effect on consumer spending andinflation will be one of many factors that the MPC will takeinto consideration when assessing the appropriate stance ofmonetary policy.

Higher interest ratesHouseholds’ median expectations for Bank Rate were verysimilar to those based on financial market measures at thetime of the 2015 survey. Both expected gradual and limitedincreases in Bank Rate (see the magenta line and diamonds inChart 7). Survey respondents expected a lower path forBank Rate than at the time of the 2014 survey (those previousexpectations are shown by the blue diamonds on Chart 7). Inthe 2015 survey, households’ median expectation forBank Rate in September 2016 was 0.6% (compared to 1.4% atthe time of the 2014 survey) with 61% of householdsexpecting Bank Rate to be higher than its current level of 0.5%by that point. The interest rate expectations of householdswith high DSR ratios, who are potentially most vulnerable tohigher interest rates, were similar to those of otherhouseholds.

The analysis in this section is based on a purely hypotheticalscenario in which Bank Rate rises immediately by 2 percentagepoints to 2.5%. While this analysis can shed some light onhow households may respond to a rise in interest rates, it isimportant to interpret the results of the scenario analysiswith caution. Most obviously, both financial markets and theMPC expect Bank Rate to increase in a gradual and limited way

as the economy recovers. The actual path Bank Rate willfollow over the next few years will depend on economiccircumstances. This scenario was chosen for simplicity and toallow the results to be comparable to those from the2014 survey.

Other assumptions underpinning the scenarios also need to bementioned. Household income is assumed to remainunchanged — whereas in practice, increases in Bank Rate arelikely to occur in a period when incomes are rising. However,the approach taken here avoids the need to make assumptionsabout how other aspects of households’ financial situationsmay change over time. The scenarios also assume full andinstantaneous pass-through of higher Bank Rate into theinterest rates faced by households. This is likely tooverestimate the actual impact as many households holdloans and saving products whose interest rates are fixed for aperiod; although over a longer time horizon some of thosecontracts may need to be refinanced. In the 2015 NMGsurvey, 58% of mortgagors reported that they held afixed-rate mortgage, compared to 50% in the 2014 survey. Inaddition, and again for simplicity, we assume no change inrepayments of the principal on loans.(2)

(1) The box on page 16 of the August 2015 Inflation Report discussed the implications ofthe policy measures announced in the Summer Budget 2015. Seewww.bankofengland.co.uk/publications/Documents/inflationreport/2015/aug.pdf.

(2) Typically, when interest rates rise, principal repayments fall for those on a repaymentmortgage, and so the analysis presented here may overestimate the true extent ofany increase in debt repayments.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Sep. Sep. Sep. Sep. Sep. Sep.

2015 16 17 18 19 20

September 2015 NMG survey: Median(a)

September 2014 NMG survey: Median(a)

Financial market expectations at time of September 2015 NMG survey(b)

Financial market expectations at time of September 2014 NMG survey(b)

Per cent

Sources: Bloomberg, NMG Consulting survey and Bank calculations.

(a) Question: ‘The level of interest rates set by the Bank of England (Bank Rate) is currently0.5%. At what level do you expect that interest rate to be in each of the following timeperiods? One year from now/two years from now/five years from now?’.

(b) Forward curve estimated using overnight index swap rates over the period which the surveywas conducted. The September 2014 survey was conducted from 3 September to24 September 2014. The September 2015 survey was conducted from 2 September to22 September 2015. Forward curves constructed in this way are likely to reflect a measureclose to the mean expectation of financial market participants.

Chart 7 Expectations for Bank Rate


The results from the latest survey suggest that households arein a slightly better position to cope with a rise in interest ratesthan a year ago. When asked how much their monthlymortgage payments could increase for a sustained periodwithout them having to take action, such as cutting spending,working longer hours, or requesting a change to theirmortgage, households reported having more available incomethan was the case a year ago. An estimated 31% ofmortgagors would need to take action of the kind describedabove if interest rates rose by 2 percentage points while theirincome remained unchanged, down from 37% in 2014 and44% in 2013 (the magenta lines on Chart 8).

Far fewer households would have to take action in response tohigher mortgage payments if their income also rose by 10%.Assuming all of that extra income was used to meet additionalrepayments, the proportion of mortgagors that would need totake action in response to a 2 percentage point interest raterise would fall to around 2% (the solid blue line in Chart 8)which is lower than the proportion based on the 2013 or 2014responses.

An alternative approach to assessing the impact of higherinterest rates is to consider how the number of householdswith high DSRs might increase. This may be important as highDSRs are associated with a higher risk of financial distress. Ifinterest rates were to rise by 2 percentage points, but incomes

were to remain unchanged, the proportion of households witha total DSR of at least 40% would rise from around 5% to 7%(illustrated by the green bars in Chart 9). The share ofhouseholds who would fall into this category after a rate risewas slightly lower than when a similar experiment wasconducted in 2014. One reason for this is that there werefewer households in the latest survey with total DSRs of 30%to 40% that would be likely to move up into the over 40%group if rates were to increase. If the income of all householdswere to rise by 10%, there would be no increase in theproportion of households in the 40% DSR group after a2 percentage point rate rise (the orange bars in Chart 9).

These experiments illustrate that, unsurprisingly, the outlookfor household income is a key factor that will determine thevulnerability of households to a rise in interest rates. In theNovember 2015 Inflation Report, aggregate real post-taxhousehold income was projected to grow by 2¼% in eachyear from 2016 to 2018. However, the distribution of incomegrowth among households will also be important.

The survey also included some questions asking householdsdirectly how they would respond to a 2 percentage point risein interest rates. This was presented to them in the context ofhow much their own interest payments/receipts wouldchange, calculated using responses for the amount of debt anddeposits they have from earlier questions in the survey.Presenting figures in this way should have made it easier forhouseholds to respond accurately, by placing the impact ofhigher rates in the context of their own personal financialsituations.

0

10

20

30

40

50

60

0.0 1.0 2.0 3.0 Interest rate increase (percentage points)

Percentages of mortgagors

2013 NMG survey responses



Assuming 10% rise in income (2015 responses)



(c)


(a) Question asked to mortgagors with discounted, base rate tracker or standard variable-ratemortgages: ‘The interest payment on mortgages is often linked to the official interest rateset by the Bank of England. If the rate was to increase, your monthly payments would alsoincrease. About how much do you think your monthly mortgage payments could increasefor a sustained period without you having to take some kind of action to find the extramoney eg cut spending, work longer hours, or request a change to your mortgage?’.Households on fixed/capped-rate mortgages were asked the following question: ‘Althoughyour monthly mortgage payments are currently [fixed/capped] we would like to understandthe impact if your payments were to increase tomorrow. About how much do you thinkyour monthly mortgage payments could increase for a sustained period without you havingto take some kind of action to find extra money eg cut spending, work longer hours, orrequest a change to your mortgage?’. The answers were provided in pounds.

(b) Households are defined as having to take action if the additional mortgage payments fromhigher interest rates (calculated using information on the size of the current outstandingmortgage) exceed the income available to meet higher mortgage payments. The incomegrowth scenario line uses the same calculation but assumes that monthly householddisposable incomes are increased in line with a 10% increase in annual gross householdincome.

(c) Denotes a 2 percentage point increase in interest rates.

Chart 8 Proportion of mortgagors that would need torespond to a rise in mortgage rates(a)(b)

0

5

10

15

20

25

30

35

>0–10 10–20 20–30 30–40 40+


Total DSR(b) (per cent)

Total DSR based on 2015 NMG responses

Scenario 1: 2 percentage point rise in rates; 10% increase in income

Scenario 2: 2 percentage point rise in rates; income unchanged


(a) The 2015 distribution replicates that shown in Chart 5. Scenarios 1 and 2 assume fullpass-through of a 2 percentage point interest rate rise to the debt-service costs of bothmortgage debt and unsecured debt. Scenario 1 also assumes a 10% increase in gross incomefor all households.

(b) See footnotes (a) and (c) of Chart 3 for a description of the calculation of total DSR.

Chart 9 Sensitivity of the distribution of total DSRs tohigher interest rates(a)


Over half of borrowers — both mortgagors and householdswith only unsecured debt — reported that they would cutspending in response to a 2 percentage point rise in interestrates (Table A). The expected responses to higher rates weresimilar to those reported a year ago, and the expectedresponses of mortgagors with high mortgage DSRs (the middlecolumn in Table A) were not very different to those ofborrowers as a whole.

The share of savers who reported that they would raisespending in response to an interest rate rise was considerablylower than the proportion of borrowers who reported thatthey would spend less. Only 10% of savers reported that theywould respond by increasing spending (Table B). Around halfof savers said that they would allow the extra income toremain in their account, and just over a third indicated theywould put aside additional savings. The additional savingsincome earned by these households would be available to bespent at a later date.

The survey also asked households to quantify how much theywould change their spending. A summary statistic forassessing the relationship between changes in income andconsumption is the marginal propensity to consume (MPC). Inthis case, the MPC for borrowers is a measure of how muchspending would be cut for each extra pound of income that isdiverted to higher interest payments. For savers it representshow much spending would increase for each extra pound ofinterest receipts received.

Borrowers reported an average MPC of 0.48 in response tohigher loan repayments (Chart 10). In calculating this figure,more weight is given to borrowers with a larger amount ofoutstanding debt in order to give a sense of how much

aggregate spending would change for a given change ininterest payments. The reported MPC of borrowers wasmarginally lower than a year ago. The average MPC of saverswas unchanged at 0.09.

The survey results do not imply that an increase in interestrates would have an unusually large effect on householdspending. Based on these MPCs, a 1 percentage point rise ininterest rates is estimated to reduce aggregate spending

Table A Borrowers’ responses to a hypothetical 2 percentagepoint rise in interest rates(a)(b)

Percentages of households All Mortgagors with Unsecured(change relative to mortgagors DSR ≥ 30%(c) only borrowers2014 NMG survey)

Cut spending 55 (-2) 51 (2) 58 (-4)

Save less 33 (-3) 21 (0) 24 (-2)

Work more/take a second job 18 (-1) 26 (3) 20 (-3)

Take up employment myself 1 (-1) 3 (-2) 2 (-1)

Someone else in household will take up employment 2 (-1) 2 (-4) 2 (1)

Get financial help 5 (-1) 11 (0) 10 (1)

Request change to loan 22 (0) 30 (9) 14 (-1)

Move somewhere cheaper 7 (-2) 13 (4) n.a.

Move and rent 4 (-2) 9 (-3) n.a.

Other 7 (1) 7 (2) 7 (2)


(a) Question: ‘If your monthly mortgage/unsecured loan payments were to increase for a sustained period by£x [which is calculated automatically from software as the payment increase under a 2 percentage pointincrease in interest rates], how do you think you would respond? Please assume your income would not beany higher unless you take action to increase it’. Households were allowed to select up to three options.

(b) The table only records the responses of households with net debts. Unsecured borrowers were only askedthe question if they had more than £4,999 of unsecured debt. When calculating percentage shares,households who preferred not to respond to this question are included in the denominator.

(c) See footnote (a) of Chart 3 for a description of the calculation of mortgage DSR.

Table B Savers’ responses to a hypothetical 2 percentage pointrise in interest rates(a)(b)

Percentages of households (changerelative to 2014 NMG survey) All savers

Increase spending 10 (0)

Do nothing (let interest accumulate) 50 (2)

Put more money into savings accounts 36 (-2)

Work fewer hours 2 (0)

Other 2 (0)


(a) Question: ‘If the monthly interest you receive on your savings were to increase for a sustained period by £x[which is calculated automatically from software as the payment increase under a 2 percentage pointincrease in interest rates], how do you think you would respond? Please assume your other sources ofincome would not change’. Households were allowed to select any of the options. When calculatingpercentage shares, households who preferred not to respond to this question are included in thedenominator.

(b) The table only records the responses of households with positive net savings. Savers were only asked thequestion if they had more than £4,999 of savings. Households with a mortgage were not asked thisquestion, regardless of their level of savings.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Borrowers Savers

2014

2015Average MPC


(a) Questions: ‘If your monthly loan payments were to increase for a sustained period by £x[which is calculated automatically from software as the payment increase under a2 percentage point increase in interest rates], how do you think you would respond? Pleaseassume your income would not be any higher unless you take action to increase it’.Households were allowed to select up to three options. Respondents who reported theywould cut spending were then asked ‘How much would you reduce your monthly spendingby in this situation?’. The marginal propensity to consume is calculated as the reportedaggregated change in spending within a type of household as a share of the aggregatedchange in interest payments within a type of household. Respondents who reported thatthey would not change spending were given an MPC of zero. Unsecured borrowers andsavers with debt/deposits of less than £5,000 were not asked how they would respond tohigher interest rates and are therefore assumed to have an MPC of zero. For mortgagors,only their MPC out of changes in mortgage interest payments is considered. Borrowers whoreported that they would cut spending but did not respond to the question about by howmuch were assumed to have an MPC of 1 (the median response of mortgagors who did sayhow much they would change spending). Savers who reported that they would increasespending but did not respond to the question about by how much were assumed to have anMPC of 0.75 (the median response of savers who did say how much they would changespending). Mortgagors are defined as households with a mortgage who have positive netdebt, unsecured borrowers are non-mortgagors with positive net debt and savers arehouseholds with net savings. Data for 2014 and 2015 are from the H2 surveys only.

Chart 10 Estimated marginal propensity to consume fordifferent types of households(a)


directly by around 0.5% as a result of redistributing incomefrom borrowers to savers, although this is only one of thechannels through which changes in interest rates can affectaggregate spending.(1) These estimates of marginalpropensities to consume are close to previous assumptionsmade by Bank staff (which are based on the academicliterature).(2)

Fiscal consolidationSince 2010, the Government has announced a number ofmeasures to reduce the country’s budget deficit. Estimatesfrom the Institute of Fiscal Studies suggest that, as of fiscalyear 2014/15, around half of the planned consolidationrelative to the March 2008 Budget had taken place. The latestNMG survey, as in previous waves, for example, 2012, includeda set of questions asking households how they had beenaffected by the fiscal measures, how they had responded andhow they expected be affected in the future. The survey wascarried out prior to policy announcements in theAutumn Statement.

Again, there are some caveats to make before drawing stronginferences from this analysis. To start with, it is hard to assesswhat would have happened in the economy if a fiscalconsolidation had not taken place. In addition, the questionsasked do not allow for an assessment of whether householdsonly focus on how they have been/expect to be directlyaffected by the fiscal consolidation or whether they also takeinto account the implications it has on them via the aggregateimpact of the consolidation on GDP and employment.

Around 30% of all households reported that they had beenadversely affected by the fiscal consolidation since 2010(Chart 11). That is lower than in the 2011 and 2012 surveys,the last time the survey included fiscal questions, althoughthis difference may partly be due to a change in the way thequestion was asked.(3) Fifty per cent of households in thelatest survey reported that fiscal measures introduced since2010 had not affected them, with a very small proportionsaying they had been positively affected and the remainderunsure.

Survey responses indicate that the impact of the fiscalconsolidation has varied by employment status, with thoseemployed in the public sector being most heavily affected,followed by non-retirees that are not employed (Chart 11).

When asked in 2015 about which aspects of the fiscalconsolidation had negatively affected their household, themost frequently chosen option was a reduction in spending onpublic services. This differs from previous surveys in 2011 and2012 when higher taxes were cited more often (Table C). Thatcould reflect the closer proximity of the previous surveys tothe VAT increase that came into effect in January 2011. It maybe the case that households in 2015 are not remembering as

far back as 2011 and are only recalling how they have beenaffected over the more recent past.

As in previous surveys, the proportion of households thatexpected to be affected by further fiscal consolidation waslarger than the proportion who reported that they havealready been affected. In 2015, 34% of households reportedthat they had been affected since 2010, but 67% expectedthere to be an effect still to come.

Higher taxes and cutbacks in public services are the two mostcommonly cited ways in which households expected to beaffected by the fiscal consolidation in the future. A high levelof concern about higher taxes (although a little less than in2012) was despite the fact that the remaining consolidation isplanned to be achieved primarily through lower governmentspending rather than higher tax revenue. The proportionconcerned about services being cut back had increased a littlesince the 2012 survey, while concerns about loss of benefits orjob had fallen slightly, although they remained important for anumber of households (Table D).

0

10

20

30

40

50

60

70

Positively Negatively No impact Don't know

All households

Retired

Employed in public sector

Employed in private sector

Not employed(b)



(a) Questions: ‘Since 2010, the Government has announced a succession of measures in orderto cut the country’s budget deficit. How have these measures affected your household overthe past five years?’. ‘Does your household gain more than half of its income from work forthe public sector (by this we mean working directly for the public sector or for industries orservices that mainly depend on contracts with the government for their business activity)?’.Percentages are calculated excluding those respondents who reported ‘prefer not to state’ tothe effect of fiscal measures on their household. Results are shown from 2015 H2NMG survey only.

(b) Includes full-time students/those still at school; those unemployed and seeking work; thosenot in paid work because of long-term illness or disability; those not in paid work for otherreasons.

Chart 11 Impact of fiscal measures since 2010 onhouseholds by employment status(a)

(1) For example, higher interest rates will also encourage consumption to be postponedbecause greater returns on saving increase the amount of future consumption thatcan be achieved by sacrificing a given amount of spending today.

(2) Previous internal work by Bank staff has assumed MPCs of 0.5% for borrowers and0.2% for savers.

(3) In 2011 and 2012 the survey asked how respondents had been affected by the pastfiscal measures with the option to choose ‘I do not think I have been heavily affected’.In 2015, the survey first asked whether respondents had been positively or negativelyaffected by the fiscal consolidation before asking what the specific impact had beenon those who had been negatively affected.


In response to the fiscal measures implemented so far, themost common actions by households have been spending lessand saving less (blue bars in Chart 12).(1) In the 2015 survey,around 20% of households reported having cut spending inresponse to fiscal measures, which is broadly similar to thepercentage of households who reported cutting spending dueto concerns regarding credit availability or their level of debt.The results from the survey suggest that the fiscalconsolidation may have been a significant factor weighing onhousehold spending since the financial crisis, although it isimpossible to know how household spending would haveevolved in the absence of a fiscal consolidation.

Cutting spending was also the most common expectedresponse to fiscal consolidation in the future. Relative to theiractions over the past, an even greater proportion — around30% — of households expected to cut spending in response tofiscal measures in the future (magenta bars in Chart 12). Thesurvey results suggest that the consolidation might weigh onaggregate household spending for some time to come,although some households were concerned about highertaxes, which are not currently a large component of theplanned consolidation.

While the fiscal consolidation will affect different householdsin different ways, the implications for monetary and financialstability may be greater if the most affected households alsoface financial pressure from other sources. As alreadydiscussed, households with high debt-servicing costs may bemost vulnerable to a rise in interest rates. On average, highDSR mortgagors expected to be more affected by further fiscalconsolidation than low DSR mortgagors and the averagehousehold (the orange lines in Chart 13).

High DSR mortgagors had a higher level of concern around theloss of income benefits than low DSR mortgagors, on average(the blue line in Chart 13). However, the survey was carriedout prior to policy announcements in the Autumn Statement,which included the decision to reverse the main cuts to taxcredits announced in the Summer Budget 2015. These

Table C Impact of fiscal measures since 2010 on households(a)


2011(b) 2012 2015

Affected 51 48 34

Negatively n.a. n.a. 31

Loss of income benefits 11 14 11

Loss of job 6 6 3

Cut back on services 16 17 14

Lower income(c) 21 n.a. n.a.

Lower pre-tax income(c) n.a. 6 5

Higher taxes 21 22 7

Positively n.a. n.a. 3

No impact 39 41 50

Don’t know 10 11 16


(a) Questions: In 2015, respondents were asked ‘Since 2010, the Government has announced a succession ofmeasures in order to cut the country’s budget deficit. How have these measures affected your householdover the past five years?’. Respondents who answered ‘Negatively’ were then asked ‘In what way has yourhousehold been affected by these changes?’ and were allowed to choose up to three options so impactsmay not sum to totals. Some households responded ‘don't know’ or ‘prefer not to state’ to the questionasking them to specify the way they had been affected negatively. Percentages are calculated excludingthose respondents who reported ‘prefer not to state’ to the effect of fiscal measures on their household.See footnote 3 on page 365 for changes in survey questions from 2011 and 2012.

(b) Includes only responses completed online.(c) The possible responses to this question changed marginally between 2011 and 2012 surveys. 2012 and 2015

surveys refer to ‘Lower pre-tax employment income’ while the 2011 survey refers to ‘Lower income’.

Table D Expected future impact of fiscal measures onhouseholds(a)


2011(b) 2012 2015

Affected 76 70 67

Loss of income benefits 18 21 17

Loss of job 22 22 18

Cut back on services 26 27 30

Lower income(c) 33 n.a. n.a.

Lower pre-tax income(c) n.a. 9 7

Higher taxes 39 40 33

No impact 15 19 19

Don’t know 10 10 14


(a) Question: ‘Some of the Government’s measures will come into effect over coming years. Which of thefollowing are you most concerned about for the future?’. Respondents were allowed to choose up tothree options so impacts may not sum to totals. Percentages are calculated excluding those respondentswho reported ‘prefer not to state’ to the future effect of fiscal measures on their household.

(b) Includes only responses completed online.(c) The possible responses to this question changed marginally between 2011 and 2012 surveys. 2012 and 2015

surveys refer to ‘Lower pre-tax employment income’ while the 2011 survey refers to ‘Lower income’.

(1) These options were only added to the latest survey, which makes a comparison ofresponses over time difficult. The proportion of households reporting that they havetaken no action has reduced to 6% from around 20% in 2011 and 2012 which isperhaps a consequence of the change of options rather than an indication ofhousehold reactions.

0

5

10

15

20

25

30

35

Action takenAction planned


Savemore

Saveless

Spendless

Newjob

Other Noaction

Longerhours(b)

Spendmore


(a) Respondents that have been/expect to be positively or negatively affected by the fiscalmeasures are asked ‘Which, if any, of the following actions have you taken/will you take inresponse to those measures?’ and can choose up to three options. Some householdsresponded ‘don’t know’ or ‘prefer not to state’, not shown on the chart. Percentages arecalculated excluding those respondents who reported ‘prefer not to state’ to the effect offiscal measures on their household. Results are shown from 2015 H2 NMG survey only.

(b) Work longer hours/take a second or better paid job.

Chart 12 Actions taken or planned by households inresponse to fiscal consolidation(a)


reductions in tax credits may have been one of the measuresthat households were most concerned about in the survey,and so it is possible that those concerns may now be lowerthan when the survey was conducted, although, relative to thesize of the economy, the Office for Budget Responsibility stillexpects welfare spending to fall during the parliament.Mortgagors with higher DSRs did not expect to bedisproportionately affected by most other fiscal measures.

The monetary and financial stability implications of tighterfiscal policy may not just be limited to mortgagors. Unsecuredborrowers who are already struggling with their debts mayfind it harder to make their repayments if their income isaffected. Non-mortgagors who reported their unsecured debtto be a heavy burden were more likely than those not findingtheir debt a burden to expect to be affected by future fiscalmeasures. These groups also had different average levels ofconcern about a loss of benefit income (Chart 14).

While there are some households who may be adverselyaffected by both monetary and fiscal policy simultaneously,that group is relatively small. Only a quarter of all mortgagorhouseholds with a total DSR in excess of 40% were concernedabout loss of benefit income, which corresponds to less than1% of all households. And non-mortgagors finding theirunsecured debts to be a heavy burden and who were worriedabout a loss of benefit income only accounted for around 4%of all unsecured debt.

Conclusion

In aggregate, household debt remains high relative to income,but the cost of servicing that debt is historically low. From adistributional perspective, data from the latest NMG surveysuggest that there has been a modest further improvement inthe balance sheet positions of mortgagors over the past year.The proportion of households with high DTI ratios and highDSRs has fallen slightly, while reported levels of financialdistress are low and have declined a little further.

Those modest improvements in balance sheet positions implythat households are in a slightly better position to cope withan increase in interest rates than they were a year ago.Households reported that they had more income available tomeet any increase in mortgage repayments. The surveyresults do not imply that an increase in interest rates wouldhave an unusually large effect on household spending.

The survey results suggest that the fiscal consolidation is animportant factor that has weighed on household spending,and it is likely to continue to do so. There are also somehouseholds who may be more vulnerable to both higherinterest rates and who expect to be more heavily affectedthan average by further fiscal consolidation, although thatgroup is relatively small. Developments in income will be animportant determinant of how households’ financial positionsevolve, but will be of particular importance for morevulnerable households.

>0–10 10–20 20–30 30–40 40+


Total DSR (per cent)


55

50

60

65

70

75

80

85

0

5

10

15

20

25

35

30

Averages across all households

Concerned about loss of income benefits (right-hand scale)

Expect to be affected by fiscal consolidation (left-hand scale)


(a) Question: ‘Some of the Government’s measures will come into effect over coming years.Which of the following are you most concerned about for the future?’. Respondents wereallowed to choose up to three options. Percentages are calculated excluding thoserespondents who reported ‘prefer not to state’ to the future effect of fiscal measures on theirhousehold. Results are shown from 2015 H2 NMG survey only.

(b) See footnotes (a) and (c) of Chart 3 for a description of the calculation of total DSR.

Chart 13 Mortgagors’ expectations about the futureimpact of fiscal measures by total DSR(a)(b)

Percentages of householdsPercentages of households

55

50

60

65

70

75

80

85

0

5

10

15

20

25

35

30

Averages across all households

Concerned about loss of income benefits (right-hand scale)

Expect to be affected by fiscal consolidation (left-hand scale)

No debt Has debt butnot a burden

Somewhat ofa burden

A heavyburden

Unsecured debt repayments


(a) Question: ‘Some of the Government’s measures will come into effect over coming years.Which of the following are you most concerned about for the future?’. Respondents wereallowed to choose up to three options. Percentages are calculated excluding thoserespondents who reported ‘prefer not to state’ to the future effect of fiscal measures on theirhousehold. Results are shown from 2015 H2 NMG survey only.

(b) Question: Households with unsecured debt were asked ‘To what extent is the repayment ofthese loans and the interest a financial burden on your household?’.

Chart 14 Non-mortgagors’ expectations about thefuture impact of fiscal measures by reported burden ofdebt repayments(a)(b)


References

Anderson, G, Bunn, P, Pugh, A and Uluc, A (2014), ‘The potential impact of higher interest rates on the household sector: evidence from the2014 NMG Consulting survey’, Bank of England Quarterly Bulletin, Vol. 54, No. 4, pages 419–33, available atwww.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q405.pdf.

Bunn, P, Johnson, R, Le Roux, J and McLeay, M (2012), ‘Influences on household spending: evidence from the 2012 NMG Consulting survey’,Bank of England Quarterly Bulletin, Vol. 52, No. 4, pages 332–42, available atwww.bankofengland.co.uk/publications/Documents/quarterlybulletin/qb120403.pdf.

Recent economic andfinancial developments

Quarterly Bulletin Recent economic and financial developments 369


• UK short-term interest rates declined over the review period as a whole, with much of that moveoccurring following the publication of the November Inflation Report.

• Long-term interest rates in the United Kingdom declined a little, with rather larger falls observedin the euro area. The component for inflation compensation rose somewhat in the United Statesand euro area, reflecting some abatement of concerns about the prospects for global growth.

• The sterling ERI ended the period up by 1.8%, reflecting an appreciation versus the euro, whichmore than offset a fall against the dollar.

• Worries surrounding global growth weighed on developed-economy equities initially, butimproving sentiment later in the review period more than offset the earlier losses.

Overview

The past quarter has been characterised broadly bytwo distinct episodes. The early part of the review periodwas essentially a continuation of the volatile conditionsobserved during much of the summer. And existing worriesabout slowing global growth were somewhat heightened bythe decision of the Federal Open Market Committee (FOMC)in the United States not to tighten policy at its Septembermeeting. Short-term interest rates in both the United Statesand United Kingdom fell following the September meeting.

But, during the latter part of the review period, sentimentimproved materially, and US and UK short-term interestrates rose. In large part, the improvement in confidence wasdue to a reduction in the emphasis on internationaldevelopments at the October meeting of the FOMC, as wellas strong US labour market data. Commentary from theEuropean Central Bank (ECB) — to highlight that theGoverning Council was considering options for furthermonetary policy easing ahead of its December meeting —gave an additional boost to sentiment. Shortly after the datacut-off the ECB announced several measures to loosenpolicy.

In terms of specific UK monetary policy expectations, therewas a decline in short-term interest rates following therelease of the Inflation Report. Overall, UK short-terminterest rates were down slightly over the review period asa whole. Longer-term government bond yields were fairlystable, but there was an increase in the component of yieldsthat compensates for long-term inflation expectations.

Meanwhile, there were some large declines in swapspreads — the difference between interest rate swap andgovernment bond yields of equivalent maturity — and in thecross-currency basis swap market. In both cases, thesewere partly the result of short-term, temporary factors. Butcontacts reported that structural changes, particularly as aresult of an increase in the capital intensity of securedlending and borrowing activity among banks, had also playeda role, and some of those factors might be expected topersist.

Currency moves were broadly consistent with the directionof changes in relative interest rates in the United Kingdom,United States and euro area. Thus, there was broadappreciation of the US dollar, while the euro fell. Given therelatively large weight of the euro in the sterling exchangerate index, the rise in sterling versus the euro more thanoffset the decline versus the dollar, leading to anappreciation of sterling overall. There was a modest pickupin sterling-dollar implied volatility, having been quite steadyfor much of the year. It was unclear whether the changereflected a shift in expectations about the relative paths ofmonetary policy in the two countries, or some UK-specificrisk factor.

Early in the review period, most developed equity marketsdeclined slightly, driven by worries around possible spilloversfrom a slowdown in emerging markets. But improving risksentiment subsequently resulted in a reversal of those earlierdeclines, with equities broadly higher overall.

Markets and operations

Recent economic and financial developments Markets and operations 371

In discharging its responsibilities to ensure monetary andfinancial stability, the Bank gathers market intelligence fromcontacts across a range of financial markets. Regular dialoguewith market contacts provides valuable insights into howmarkets function, and provides context for the formulation ofpolicy, including the design and evaluation of the Bank’s ownmarket operations. The first section of this article reviewsdevelopments in financial markets between the 2015 Q3Quarterly Bulletin and 27 November 2015. The second sectiongoes on to describe the Bank’s own operations within theSterling Monetary Framework.

Monetary policy and interest ratesShort-term UK market interest rates fell during the quarter(Chart 1), with one-year, one-year forward overnight indexswap (OIS) rates declining by around 12 basis points overall.As a result, the expected timing of the first increase inBank Rate, as implied by market interest rates, was pushed outto the first quarter of 2017, and the expected pace ofsubsequent Bank Rate rises slowed. But the quarter could besplit into two distinct episodes. The early part of the reviewperiod was driven largely by worries about global growth, withexpectations for policy tightening in the United Kingdom andUnited States being pushed out as a result. Later on, as thoseconcerns abated, the focus of market participants shifted backto the possibility of an increase in US policy rates in 2015, aswell as the likelihood of further loosening by the EuropeanCentral Bank (ECB).

In September, UK short-term interest rates declined, alongsidethose in the United States, after the Federal Open MarketCommittee (FOMC) opted to leave the target range for thefederal funds rate unchanged (Chart 2). Contacts suggestedthat markets had placed a material weight on the chances ofan increase, prompting a fall in US interest rates when thisfailed to materialise. Contacts also pointed to theaccompanying FOMC communication, which noted that‘global risks’ had been a factor in their decision. It was widely

held by market participants that the United Kingdom, as asmaller and more open economy than the United States, mustalso be vulnerable to such risks. That helped to account forthe corresponding shift down in UK short-term interest ratesthat also occurred following the FOMC decision.

References to ‘global risks’ were subsequently removed fromthe FOMC’s October statement, helping to push up bothUS and UK short-term interest rates (Chart 2). Subsequentstrong labour market data and commentary from variousFOMC members led to a growing expectation among marketcontacts that the FOMC would vote to increase the targetrange for the federal funds rate at its 16 December policymeeting. Overall, the US one-year, one-year forward rate roseby 14 basis points during the review period. By the end of thereview period, the US yield curve was pricing in around 75% ofa rate rise by the time of the December meeting.

From around the start of November, UK short-term interestrates began to drift lower (Chart 2). The start of this declinecoincided roughly with the publication of the NovemberInflation Report. US short-term interest rates, meanwhile,remained fairly stable. It was possible that the deviation ofthe two reflected the growing weight placed upon thelikelihood of tightening by the FOMC in the near future, incontrast to the United Kingdom, where there remainedrelatively greater uncertainty about the precise timing oflift-off. It was also noted by some contacts that there mayhave been downward pressure on UK short-term interest ratesdue to widely anticipated easing by the ECB. Such a movemight be expected to result in a material rise in the sterlingexchange rate index. In turn, contacts thought that thisimplied that returning inflation to target would requirecomparatively less tightening via Bank Rate than previouslyanticipated.

0.5

0.0

0.5

1.0

1.5

2.0

2.5

2015 16 17 18 19 20

Per cent

Dashed lines: 3 September 2015Solid lines: 27 November 2015

US dollar

Sterling

Euro

+

–

Sources: Bloomberg and Bank calculations.

(a) Instantaneous forward rates derived from the Bank’s OIS curves.

Chart 1 Instantaneous forward interest rates derivedfrom OIS contracts(a)

0.4

0.3

0.2

0.1

0.0

0.1

0.2

0.3

0.4

Sep. Oct. Nov.

Per cent

United Kingdom

United States

Euro area

November Inflation Report

FOMC policy meeting

ECB policy meeting

FOMC policy meeting

2015

+

–


(a) Forward rates derived from the Bank’s OIS curves.

Chart 2 Cumulative change in one-year OIS rates,one year forward since 3 September(a)


Developments in swap spreads

A ‘swap spread’ is the difference between interest ratesderived from government securities and the fixed rate onequivalent-maturity interest rate swaps. It tends to be thecase that the yields on interest rate swaps are higher thanthose of the equivalent government bond maturity, such thatthe spread is positive. In large part, that is because it isrelatively less risky to lend to governments than to banks. ButUS and, to a lesser degree, UK swap spreads declinedmaterially during the review period (Chart A). In fact, the fallwas sufficiently large as to cause US swap spreads at longmaturities to fall to record lows.

Contacts thought that the deviation of swap spreads frommore ‘normal’ levels had been driven initially by temporaryfactors. The start of the recent falls broadly coincided with aperiod during the summer when emerging market centralbanks were thought to have been selling large amounts ofUS government bonds. This put upward pressure onUS government bond yields, relative to interest rate swaps.Given the close comovement of government bond yieldsacross developed economies, there was a similar decline inUK swap spreads as well.

Another potential short-term driver of falling swap spreadssometimes cited by contacts, was the relatively robust pace ofbond issuance by US corporates. There were two channels viawhich this effect might be felt. First, corporate bond issuersoften transform the fixed interest rate on their liabilities to afloating rate, via swap agreements. The increased demand forreceiving the fixed leg of the swap causes the swap rate to fall,reducing the swap spread. Second, there is likely to have beensome crowding out of demand for government debt, asinvestors absorbed the increased supply of corporate bonds.

Alongside these short-term factors, some contacts thoughtthat part of the fall in swap spreads could be explained by amaterial change in the approach of banks toward pricing of‘repo’ activity for clients. A repo is a form of securedfinancing, in which the borrower agrees to repurchase thecollateral used to secure the loan at a later date. Andgovernment bond yields can be thought of as the expectedaverage short-term government repo rate over the life of thebond.

In a repo agreement, a bank may take either side of thetransaction, according to a client’s requirements. In the past,banks ran large repo books, but netted off the secured loansand borrowings against each other, holding capital onlyagainst the net of the two sides of the balance sheet. But thecapital intensity of that business has risen as a result of newrules on bank leverage, leading to a rise in the cost of suchactivity — versus other capital-intensive services — and adecline in the proportion of dealer balance sheets devoted torepo (Chart B).

Until recently, banks had refrained from passing on thosehigher costs to clients. But contacts report that there hasbeen a marked change in pricing behaviour over the past fewmonths, with banks passing on rather more of the increase incosts. This, in turn, has pushed up on the cost associated withholding government bonds as a form of collateral, causing arise in yields. And while the short-term factors thoughtresponsible for the initial fall in swap spreads might reasonablybe expected to diminish — as those temporary supply factorsabate — systematic repricing of repo is more likely to persist.

20

15

10

5

0

5

10

15

20

25

30

May June July Aug. Sep. Oct. Nov.

Ten-year US dollar swap spread

Ten-year sterling swap spread

Basis points

2015

+

–

Source: Bloomberg.

(a) The marked jumps in the sterling swap spread are due to the roll of the underlying ten-yearbenchmark gilt. When rolling to a longer maturity — and so, higher-yielding — bond, theswap spread falls, as the swap rate has a constant maturity.

Chart A Ten-year UK and US swap spreads(a)

0

2

4

6

8

10

12

14

16

2006 07 08 09 10 11 12 13 14 15

Fed funds sold and reverse repos

Fed funds purchased and repos

Per cent

Sources: Federal Reserve Bank of New York and Bank calculations.

(a) Board of Governors of the Federal Reserve System: ‘Federal funds (fed funds) are reservesheld in a bank’s Federal Reserve Bank account. If a bank holds more fed funds than isrequired to cover its Regulation D reserve requirement, those excess reserves may be lent toanother financial institution with an account at a Federal Reserve Bank. To the borrowinginstitution, these funds are fed funds purchased. To the lending institution, they are fedfunds sold.’

Chart B Share of bank balance sheet allocated to repoand reverse repo(a)


The Monetary Policy Committee (MPC) also used the Reportto clarify that it would defer sales of assets held in the AssetPurchase Facility (APF) until Bank Rate had reached a levelfrom which it could be cut materially and that that level wascurrently judged to be around 2%. Some market contactssuggested that this news represented a relative loosening inpolicy, as it implied that the stock of assets would be heldconstant for longer than had previously been anticipated.But other contacts noted that this, in turn, implied thatrelatively more tightening would have to be achieved throughBank Rate to hit the inflation target, compared with what waspriced in to the yield curve. Most simply viewed the news as aclarification of existing statements made by the MPC.Long-term UK government bond yields fell marginally onthe day.

In contrast to the expected monetary policy tightening of theFOMC, the ECB loosened policy in early December.Expectations for such a move had been building steadily, withthe ECB stating at its October meeting that it was consideringoptions for further monetary policy easing — reflectingpersistent concerns around growth and low inflation. Thisexpectation led to a large decline in euro-area interest rates,as markets placed increasing weight on the probability of botha reduction in the ECB’s main policy rate and an expansion ofits asset purchase programme. In the event, after the datacut-off, the ECB loosened policy, through a combination of acut in the deposit rate, and an extension of the asset purchaseprogramme.

Over the review period, longer-term interest rates werelargely unchanged. But this masked some notable moves inlong-term inflation compensation in the United Kingdom,United States and euro area, with five-year inflation swaprates, five years forward increasing towards the end of thereview period (Chart 3). Contacts cited a range of factors forthis. In particular, they pointed to the October FOMCstatement as a potential explanation, as this had helped tolessen concerns regarding the impact on inflation from slowingactivity in the rest of the world. There had also been a tickupin US and euro-area core inflation.

Alongside these developments, the review period was markedby sizable movements in spreads between long-term interestrates derived from government-issued securities and the fixedrate on equivalent-maturity interest rate swaps — or ‘swapspreads’. These have fallen substantially in the United States,

and to a lesser extent, the United Kingdom since the summer.Contacts attributed this to a combination of temporary andstructural factors. The box on pages 372–73 sets out the issuein greater detail.

Foreign exchangeExchange rates were influenced heavily by monetarypolicy expectations in the United States and the euro areaduring the quarter. Reflecting the approach of lift-off in theUnited States, the US dollar exchange rate index rose 1.7%over the review period, back up to around its post-crisis highs.Meanwhile, the euro ERI declined by 3.7%. Broadly reflectingmoves in relative interest rate differentials, sterling declined alittle against the US dollar (-1.4%), but appreciatedsignificantly against the euro (+3.3%). On balance, thesterling exchange rate rose by 1.7% (Chart 4).

Interestingly, towards the end of the review period there wasa rise in sterling-dollar implied volatility (Chart 5). This hadbeen fairly steady in recent months, unlike implied volatilityfor many of the other majors versus the dollar, which havebeen rising. In the past, contacts have often attributed thatrelative stability of sterling-dollar implied volatility to theview held by some that the monetary policy cycles of theUnited Kingdom and United States are closely correlated.Some contacts suggested that the recent rise in sterling-dollarimplied volatility might point to a weakening of thatperception.

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Jan. Apr. July Oct. Jan. Apr. July Oct.

United Kingdom (RPI)

United States (CPI)

Euro area (HICP)

Previous Bulletin

Per cent

2014 15

0.0


(a) Swap rates derived from the Bank's inflation swap curves.

Chart 3 Selected five-year inflation swap rates,five years forward(a)

A secondary consequence of the increase in the price cost ofrepo financing is that it has increased the cost of borrowing forleveraged investors that seek to exploit small arbitrageopportunities and deviations from perceived fair value. This, inturn, requires prices to deviate further from fair value before itbecomes profitable for such investors to enter the market toexploit mispricing. Again, this effect might result in a

permanent deviation in the swap spread from what wouldpreviously have been considered ‘normal’.

There is also likely to have been additional volatility in swapspreads as a result of the usual balance sheet ‘windowdressing’ by banks ahead of key reporting periods, at whichtime there is an even greater premium on balance sheet space.


But it is worth noting that the rise in sterling-dollar impliedvolatility is not particularly evident at short maturities. Ifthere had been a material shift in views about the respectivepositions in the business cycles of the United States andUnited Kingdom, one might reasonably expect to see thismanifest itself in expectations for volatility in the short term.The fact that the rise becomes most evident from around theone-year point on the volatility curve might, then, beindicative of more medium-term risks to the currency.

There has also been a modest pickup in the price of putoptions — which provide protection against depreciation ofthe currency — versus call options. That has caused a slightfall in the sterling-dollar risk reversal (the difference betweenimplied volatilities on call and put options that are anequivalent distance from the prevailing spot price). For furtherdiscussion of implied volatility and risk reversals see the boxon page 375.

Elsewhere, perhaps the most widely anticipated event was thedecision by the International Monetary Fund to includeChinese yuan in the Special Drawing Rights currency basket.This followed a steady programme of reform by the Chineseauthorities over recent years to liberalise the currency. Thedecision was largely expected by contacts and promptedrelatively little market reaction. Contacts thought that thePeople’s Bank of China had continued to intervene in theforeign currency market to support the value of the yuan inrecent months.

Last, the review period has seen some unusual moves incross-currency basis swaps — instruments which allowinvestors to swap principal and interest paymentsdenominated in different currencies. Generally, investorsare willing to pay a small premium to hold dollars versusother currencies. But that premium has increased to levelsnot usually observed outside of periods of market stress,with some major bouts of volatility in recent months. Someof the drivers were thought to be similar to those operatingin swap spreads mentioned above, with both temporary andstructural factors at work. For further discussion, see thebox on page 377.

Corporate capital marketsFollowing references in the September FOMC statementabout risks arising from a potential slowing in global growth,there was a broad-based decline in developed-economy equitymarkets. But those losses were more than offset bysubsequent increases over the remainder of the quarter(Chart 6). The FTSE All-Share ended the review period up by2.5%, while the S&P 500 and Euro Stoxx increased by 7.1%and 6.9% respectively. After a turbulent few months duringthe summer, Chinese equities also rose strongly, with theShanghai Composite index up around 11.5% over the reviewperiod (Chart 7). At the time of the data cut-off the indexstood at 3436, down one third compared with the earlierpeak.

Spreads on developed-economy investment-grade corporatebonds were broadly unchanged over the quarter as a whole,abstracting from the period of volatility during the first half ofthe review period (Chart 8). US corporate bond issuancecontinued apace, however, in both dollars and euros (Chart 9),rising to record levels for the year to date. It was thought thatmuch of the issuance was motivated by US mergers andacquisitions activity, as well as a desire to lock in lowborrowing costs ahead of prospective policy tightening by theFOMC. Spreads on US high-yield corporate bonds alsowidened significantly, which contacts attributed to the highconcentration of companies operating in the energy sector inthe high-yield basket. Renewed weakness in the price of oilwas expected to reduce the earnings of those companies,leading to an increase in the probability of default.

65

70

75

80

85

90

95

100

105

70

75

80

85

90

95

100

105

2008 09 10 11 12 13 14 15

Indices: January 2005 = 100

Sterling ERI (right-hand scale)

US dollar index (left-hand side)

0.5

0.0

0.5

1.0

1.5

2.0

2.5

+

–

Per cent

Other

Emerging markets(a) Japanese yen

Euro

US dollar

Sterling ERI

Sources: Bloomberg, Thomson Reuters Datastream and Bank calculations.

(a) The emerging market currencies in the narrow sterling ERI are: Chinese renminbi, Czech koruna,Indian rupee, Polish zloty, Russian rouble, South African rand and Turkish lira.

Chart 4 Sterling and US dollar exchange rate indices(ERIs) and contributions to the change in the sterling ERIsince the start of the review period

6.00.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

One

wee

k

Thre

ew

eeks

Two

mon

ths

Thre

em

onth

s

Four

mon

ths

Six

mon

ths

Nin

em

onth

s

One

year

Eigh

teen

mon

ths

Two

year

s

Thre

eye

ars

Implied volatility of sterling-dollar options on 27 November 2015

Implied volatility of sterling-dollar options on 3 September 2015

Per cent

Source: Bloomberg.

Chart 5 Term structure of sterling-dollar impliedvolatility


Risk reversals

A risk reversal is the implied volatility on a call option minusthe implied volatility on an equivalent put option. One canthink of the risk reversal as a steer on expected volatility,conditional on an appreciation versus a depreciation. Becauseoptions can have different strike prices, the risk reversal isgiven in terms of a put and a call option with the same delta,where delta is a measure of the sensitivity of the value of theoption to changes in the underlying spot exchange rate. Forexample, the value of a ‘25-delta’ call option will rise (or fall)by a quarter of the change in the underlying exchange rate.

Contacts report that, generally speaking, views about ‘event’risks are often more evident in option markets than in spot, asoptions provide a means of protecting against the possibilityof unlikely, but impactful, outcomes. The referendum onScottish independence in September 2014 provides a goodexample of such a source of event risk. The fall in thesterling-dollar risk reversal in the run-up to the vote (Chart A),indicated that the market as a whole was more willing to payfor protection against a large depreciation of sterling, than anequivalent appreciation. A similar pattern was observedaround the time of the UK general election, when contactsreported that market participants were concerned about thedownside implications for sterling in the event of uncertainelectoral outcomes. At that time, the perceived downside riskto sterling was reflected in prices rather than more graduallythan it was prior to the Scotland poll.

It is important to note, however, that there is a distinction tobe made between the balance of risks around the futuredirection of moves in a currency — which one might think of

the risk reversal as capturing — and the central expectation forthe currency, which may differ in sign to the risk reversal. And,indeed, it is instructive to note that the broad path of sterlingover the periods mentioned above, has been upward — despitethese sporadic event risks which were thought at the time tohave significant downside implications for the currency, hadthey materialised.

That said, some market participants certainly do use optionsto express views about the most probable path of exchangerates. Contacts suggest that the extent to which one can inferbeliefs about the likely direction of future moves in a currencywill depend, at least in part, on positioning in the foreignexchange market as a whole.

Thus, when conviction in a particular position is low (andimplied volatility is likely to be low), investors may choose touse options to express directional views. If a trend thenbecomes established (and volatility rises), one might then seeincreasing amounts of positioning in spot instruments.Eventually, if the trend becomes stretched — or the tradebecomes ‘crowded’ — hedging via options might begin to rise.Thus, depending on the extent of positioning in the marketoverall, one might sometimes expect to observe the spotexchange rate move in the same direction as indicated by therisk reversal, while at other times the opposite might apply.

For some currencies, however, there may be structural reasonsto expect risk reversals to move in the opposite direction toexpectations for the spot exchange rate — particularly in thecase of currency pairs used for ‘carry trades’. Carry tradeinvestors sell the funding currency to buy the carry currency,putting downward pressure on the value of the fundingcurrency. At the same time, the resulting currency risk willoften be hedged using call options on the funding currency,causing the risk reversal to rise. As a result, carry fundingcurrencies often exhibit positive risk reversals.

–

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0


Three-months 25-delta risk reversalOne-year 25-delta risk reversal

2014 15

Per cent

(a) The delta of an option refers to the sensitivity of the value of the option to changes in theprice of the underlying asset. For a 25-delta option the value of the instrument will changeby a quarter of any change in the price of the underlying spot exchange rate.

Chart A Sterling-dollar 25-delta(a) risk reversals


Bank funding marketsUK banks’ senior unsecured bank bond spreads were broadlyunchanged over the quarter, and bank bond issuance wasfairly subdued. In November, the Financial Stability Boardpublished the final total loss-absorbing capacity (TLAC)standard to which all globally systemically important banksare subject — including some UK institutions. UK bankscontinued to issue holding-company senior debt, which canbe bailed-in under the new rules. The bonds carried apremium of around 30 basis points to 80 basis points overequivalent debt issued at operating-company level. Somecontacts suggested that this premium would fall over time,perhaps settling at around 15 basis points to 20 basis points,citing the experience of US banks which already have largeamounts of holding-company debt outstanding.

The results of the 2015 annual stress test released on1 December 2015 indicated that the banking system wouldhave the capacity to maintain its core functions in thestress-test scenario. The results also showed that UK bankcapital was adequate to cope with stressed projections formisconduct costs and fines, over and above those paid orprovisioned for by end-2014. Equity prices of UK banks rosefollowing the news, while there was limited reaction in bondspreads.


High-yield corporates (US dollar) (right-hand scale) High-yield corporates (euro) (right-hand scale) High-yield corporates (sterling) (right-hand scale) Investment-grade (US dollar) (left-hand scale) Investment-grade (euro) (left-hand scale) Investment-grade (sterling) (left-hand scale)

Previous Bulletin

80

120

160

200

240

280

320

0

100

200

300

400

500

600

700Basis points Basis points

2014 15

Source: BofA Merrill Lynch Global Research.

Chart 8 International corporate bond option-adjustedspreads

0

100

200

300

400

500

600

700

0

10

20

30

40

50

60

70

Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec.

Euro issuance in US$ billions

2013 2014

2015(a)

US dollar issuance in US$ billions

Euro-denominated in solid lines (right-hand scale)

US dollar-denominated in dashed lines (left-hand scale)

Sources: Dealogic and Bank calculations.

(a) Data to 27 November 2015.

Chart 9 Cumulative gross bond issuance by US privatenon-financial corporations

Previous Bulletin

0

1000

2000

3000

4000

5000

6000

June Sep. Dec. Mar. June Sep.

Index: 19 December 1990 = 100

2014 15

Source: Bloomberg.

(a) The index is quoted in domestic currency terms.

Chart 7 Shanghai Stock Exchange Composite Index(a)

60

70

80

90

100

110

120

130

140

Jan. Mar. May July Sep. Nov. Jan. Mar. May July Sep. Nov.

Euro Stoxx FTSE All-Share

S&P 500

Topix MSCI Emerging Markets index

2014

Previous Bulletin

Indices: 6 January 2014 = 100

15


(a) Indices are quoted in domestic currency terms, except for the MSCI Emerging Markets index,which is quoted in US dollar terms. The MSCI Emerging Markets index is a free-floatweighted index that monitors the performance of stocks in global emerging markets.

Chart 6 International equity indices(a)


Cross-currency basis swaps

A cross-currency basis swap allows investors to swap principaland interest payments denominated in two differentcurrencies. As a convention, each swap is usually quotedagainst the US dollar. Therefore one counterparty to a tradewill be lending US dollars (and borrowing a non-dollarcurrency), while the other will be borrowing dollars (lendingnon-dollar currency). Historically, investors have beenprepared to pay a small premium to hold dollars versus othercurrencies. As a result, the difference between the lending legof the agreement and the borrowing leg — the ‘basis’ — istypically negative.

The most notable moves in cross-currency basis markets havebeen in the dollar-yen basis which has become much morenegative — reflecting greater demand for US dollars againstthe yen — and reached levels last observed during theeuro-area sovereign debt crisis in 2011 (Chart A). Contactssuggest that the recent falls in the basis reflect a combinationof short-term, temporary, drivers and long-term structuralfactors.

In large part, contacts pointed to growing expectations thatthe Federal Reserve would move to tighten US monetarypolicy in 2015. And, indeed, there was a marked widening ofthe basis following strong non-farm payrolls data inNovember, which contacts thought increased the likelihood ofUS lift-off in December. As a result, there was an upsurge indemand for long-term US dollar funding ahead of the nextFOMC meeting.

In addition, the attractiveness of the euro as a fundingcurrency has led to a large amount of bond issuance in eurosby non euro-area issuers, which then swap the proceeds backinto domestic currency, adding to downward pressure on theeuro basis.

Among the other cyclical factors, as in the case of swapspreads, contacts noted that there had been a sharp increasein demand for US dollars for the purposes of intervention inthe foreign exchange market by emerging market centralbanks — raising US dollars to buy back domestic currency.There was also the usual seasonal increase in financialinstitutions’ demand for dollar funding over year-end, thoughtto be a particularly important driver of the dollar-yen basis.

But — again, as in the case of swap spreads — contacts placeda significant weight on structural developments in financialmarkets to explain the widening of the basis. In particular,contacts pointed an increase in the cost of repo which meantthat arbitrageurs — often financed via repos — now requirepricing anomalies to be that much greater before entering themarket. Contacts also thought that there has been a materialdecline in the amount of capital available to deploy for thepurposes of arbitraging the basis, especially in light ofdisappointing year-to-date returns for many leveragedinvestors typically involved in such activity.

120

100

80

60

40

20

0

20

2010 11 12 13 14 15

Basis points

¥/US$

€/US$

£/US$ +

–

Source: Bloomberg.

Chart A Five-year cross-currency basis swaps

Operations

Operations within the Sterling Monetary Frameworkand other market operationsThis section provides an update of the Bank’s operationswithin the Sterling Monetary Framework (SMF) over thereview period, as well as its other market operations.Collectively, these operations help implement the Bank’smonetary policy stance and provide liquidity insurance toinstitutions when deemed necessary.

The aggregate level of central bank reserves is closelymonitored by the Bank, as it affects monetary conditions inthe UK economy. The level of central bank reserves is affectedby (i) the stock of assets purchased via the Asset PurchaseFacility (APF); (ii) the level of reserves supplied by operationsunder the SMF; and (iii) the net impact of other sterling flowsacross the Bank’s balance sheet. Over the review period,aggregate reserves remained around £315 billion, but hadfluctuated due to the redemption and subsequentreinvestment of a gilt held in the APF (discussed below).

Operational Standing FacilitiesSince 5 March 2009, the rate paid on the OperationalStanding Deposit Facility has been zero, while all reservesaccount balances have been remunerated at Bank Rate. As aconsequence, there is little incentive for reserves accountholders to use the deposit facility. Reflecting this, the averageuse of the deposit facility was £0 million in the three monthsto 4 November 2015.(1)

The rate charged on the Operational Standing Lending Facilityremained at 25 basis points above Bank Rate. However, giventhe large aggregate supply of reserves, there was no demandfrom market participants to use the lending facility. Theaverage use of the lending facility was also £0 million over thequarter to 4 November 2015.

Indexed Long-Term Repo operationsThe Bank conducts regular Indexed Long-Term Repo (ILTR)operations as part of its provision of liquidity insurance tobanks, building societies and broker-dealers. During thereview period, the Bank offered a minimum of £5 billion viasix-month repos in each of its ILTR operations on8 September, 6 October and 10 November 2015 (Table A).

Participation in, and usage of, ILTR operations has continuedto remain higher than during the same period last year.Nonetheless, the total amount allocated in each operationremained below the minimum £5 billion on offer (Chart 10).This continued to reflect usage of the ILTR by someparticipants as a source of term repo liquidity. Over thereview period, a total of £10.5 billion of ILTRs matured and£7.4 billion of new ILTRs were allocated, resulting in a netreduction of central bank reserves of around £3.1 billion.

Contingent Term Repo FacilityThe Contingent Term Repo Facility (CTRF) is a contingentliquidity facility that the Bank can activate in response toactual or prospective market-wide stress of an exceptionalnature. The Bank reserves the right to activate the facility as itdeems appropriate. In light of market conditions throughoutthe review period, the Bank judged that CTRF auctions werenot required.


Table A Indexed Long-Term Repo operations(a)

Total Collateral set summary

Level A Level B Level C

8 September 2015 (six-month maturity)

Minimum on offer (£ millions) 5,000

Total bids received (£ millions) 1,195 745 10 440

Amount allocated (£ millions) 1,195 745 10 440

Clearing spread (basis points) 0 5 15

6 October 2015 (six-month maturity)


Total bids received (£ millions) 2,290 2,170 0 120

Amount allocated (£ millions) 2,290 2,170 0 120

Clearing spread (basis points) 0 n.a. 15

10 November 2015 (six-month maturity)


Total bids received (£ millions) 3,920 3,365 0 555

Amount allocated (£ millions) 3,920 3,365 0 555

Clearing spread (basis points) 0 n.a. 15

(a) The minimum amount on offer is the size of the operation that the Bank is willing to allocate, in aggregate,across all collateral sets at the minimum clearing spreads.

Three-month Level A allocated (left-hand scale)

Three-month Level B allocated (left-hand scale)

Six-month Level A allocated (left-hand scale)

Six-month Level B allocated (left-hand scale)

Six-month Level C allocated (left-hand scale)

Level A clearing spread (right-hand scale)

Level B clearing spread (right-hand scale)

Level C clearing spread (right-hand scale)

Amount allocated (£ millions) Clearing spread (basis points)

0

1,000

2,000

3,000

4,000

5,000

6,000

0

10

20

30

40

50

60

70

Jan.Apr.

JulyOct.

Jan.Apr.

JulyOct.

Jan.Apr.

2011 12 13 14 15

JulyOct.

Jan.Apr.

July JulyOct. Oct.

Jan.Apr.

(a) Where there has not been any allocation to a collateral set, no clearing spread is marked.

Chart 10 ILTR reserves allocation and clearing spreads(a)

(1) Operational Standing Facility usage data are released with a lag.


Discount Window FacilityThe Discount Window Facility (DWF) is a bilateral on-demandfacility provided to institutions experiencing a firm-specific ormarket-wide liquidity shock. It allows participants to borrowhighly liquid assets in return for less liquid collateral inpotentially large size and for a variable term. The Bankpublishes quarterly data of DWF usage with a lag. The averagedaily amount outstanding in the DWF in the three months to30 June 2014 was £0 million.

Other operationsFunding for Lending SchemeThe Funding for Lending Scheme (FLS) was launched bythe Bank and HM Treasury on 13 July 2012. The initialdrawdown period for the FLS ran from 1 August 2012 until31 January 2014. The drawdown period for the FLS extensionopened on 3 February 2014 and will run until 31 January 2018,as announced on 30 November 2015.(1)

The quantity current participants can borrow in the FLS islinked to their lending to the UK real economy from 2013 Q2to 2015 Q4, with the incentives currently skewed towardssupporting lending to small and medium-sized businesses.From 1 February 2016, participants will initially retain fullaccess to draw against their borrowing allowance, but theallowance will reduce by 25% after six months and by thesame amount every six months thereafter, phasing the schemeout gradually by 31 January 2018.

US dollar repo operationsOn 23 April 2014, in co-ordination with other central banksand in view of the improvement in US dollar fundingconditions, the Bank ceased the monthly 84-day US dollarliquidity-providing operations. The seven-day US dollaroperations will continue until further notice. The network ofbilateral central bank liquidity swap arrangements provides aframework for the reintroduction of further US liquidityoperations if warranted by market conditions. There was nouse of the Bank’s US dollar facilities throughout the reviewperiod.

Bank of England balance sheet: capital portfolioThe Bank holds an investment portfolio that is approximatelythe same size as its capital and reserves (net of equityholdings, for example in the Bank for InternationalSettlements, and the Bank’s physical assets) and aggregatecash ratio deposits. The portfolio consists ofsterling-denominated securities. Securities purchased by theBank for this portfolio are normally held to maturity, thoughsales may be made from time to time, reflecting, for example,risk or liquidity management needs or changes in investmentpolicy. The portfolio currently includes around £5.6 billion ofgilts and £0.2 billion of other debt securities.

Asset purchasesIn the publication of the Inflation Report on 5 November 2015,the Monetary Policy Committee announced that it expects tomaintain the stock of purchased assets at £375 billion,including reinvesting the cash flows associated with allmaturing gilts held in the APF, at least until Bank Rate hasreached a level from which it can be cut materially.

The 4.75% September 2015 gilt held by the APF maturedduring the review period. A total of £16.9 billion of cash flowsassociated with the purchase of the maturing gilt wassuccessfully reinvested in gilts across the curve throughoutSeptember.

The total stock of gilts outstanding in the APF, measured asproceeds paid to sellers, remains at £375 billion. The stock ofgilts comprised of £66.6 billion of purchases in the 3–7 yearsresidual maturity range, £145.1 billion in the 7–15 yearsresidual maturity range and £163.2 billion with a residualmaturity of greater than 15 years (Chart 11).

Gilt lending facilityThe Bank continued to offer to lend gilts held in the APF viathe Debt Management Office in return for otherUK government collateral. In the three months to30 September 2015, the daily average value of gilts lent, aspart of the gilt lending facility, was £152 million. The averagedaily lending in the previous quarter was higher at£330 million.

(1) For more details, see www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx.

0255075

100125150175200225250275300325350375400

Feb. Feb. Feb. Feb. Feb. Feb. Feb.

15+ years

7–15 years

0–7 years£ billions

14131211102009 15

(a) Proceeds paid to counterparties on a settled basis.(b) Residual maturity as at the date of purchase.

Chart 11 Cumulative gilt purchases by maturity(a)(b)

www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx

www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx


Corporate bondsThere were no purchases of corporate bonds during the reviewperiod. Future purchase or sale operations through thescheme will be dependent on market demand, which the Bankwill keep under review in consultation with its counterparties.Reflecting the recent lack of activity, the scheme currentlyholds no bonds.

Secured commercial paper facilityThe Bank continued to offer to purchase secured commercialpaper backed by underlying assets that are short term andprovide credit to companies or consumers that supporteconomic activity in the United Kingdom. No purchases weremade during the review period.

Report

Quarterly Bulletin Report 381

During 16–17 June 2015, the Bank of England (BoE), theHong Kong Monetary Authority (HKMA) and the InternationalMonetary Fund (IMF) held the first joint conference onmonetary, financial and prudential policy interactions in thepost-crisis world. The conference provided a forum for leadingacademics and senior policymakers from across the world todiscuss challenges that central banks and other policymakersface in choosing the optimal mix of monetary,macroprudential and microprudential policies.(1)

This article summarises the main issues discussed byparticipants during the two-day conference. The programmeand presentation slides for the sessions held on 16 June areavailable on the Bank’s website.(2) The roundtable discussionon 17 June was conducted under ‘Chatham House Rules’.(3)

How has the financial system changed sincethe global financial crisis?

Participants noted two key developments since the financialcrisis. First, many bank holding companies and non-banks hadturned into ‘hybrid intermediaries’ which provided specificservices in the process of credit intermediation — for example,in asset securitisation and securities lending. It was not yetclear whether this represented efficient evolution of theindustry, or more worryingly, its attempt to shift risks tospheres that were less visible to regulators. This could suggestthat financial activities, rather than entities, should beregulated, thus requiring greater co-ordination amongregulatory agencies, both domestically and internationally.Enhancing the information available on non-bank financialinstitutions would make it easier for regulators to monitortheir activities more effectively.

Second, it was noted that global credit growth since the crisishas been driven by corporate bond issuance rather thancross-border bank lending; and that asset managers now helda significant proportion of these corporate bonds — most ofwhich were denominated in US dollars. Even though assetmanagers were not leveraged, they typically benchmarkedtheir performance against broad indices, held only small

amounts of cash to meet redemptions, and tended to useValue-at-Risk type position limits on specific currencyexposures. So a rise in policy rates in advanced economies,particularly in the United States, could reduce the value ofthese bonds and potentially trigger a sell-off by assetmanagers.

Some participants were concerned that a large-scale sell-offby asset managers could potentially amplify the impact of rateincreases across the financial system. But others remarkedthat it was not yet clear how quantitatively significant thistransmission mechanism was.

End of ‘too big to fail’?

Some participants thought that the ‘too big to fail’ problemfor banks would be largely solved once the internationalreforms to facilitate recovery and resolution, including thedesigns of total loss-absorbing capacity and cross-borderresolution, were completed. These reforms could restoremarket discipline on big banks by creating a clear mechanismfor imposing losses on private claimholders through equitywrite-downs and debt holder ‘bail-in’. But a number of othersurged governments and central banks to have a contingencyplan, given that ‘bail-in’ was untested. Some maintained thathaving an orderly and transparent mechanism for publiccapital injection remained important and worked well in somecountries, for example in Japan.

Participants also noted that ‘too big to fail’ remained aproblem outside the banking sector, particularly for centralcounterparties (CCPs), which had grown more systemic withthe increased central clearing of derivatives. Someparticipants argued that, in some countries, CCPs were atpresent not adequately supervised and regulated, and effectiveresolution mechanisms were yet to be developed.


BoE-HKMA-IMF conference onmonetary, financial and prudentialpolicy interactions in the post-crisisworld

(1) This report was prepared by Misa Tanaka (BoE), Gaston Gelos (IMF), Julia Giese (BoE),Daryl Ho (HKMA), Sujit Kapadia (BoE), Laura Kodres (IMF) and James Talbot (BoE).

(2) www.bankofengland.co.uk/research/Pages/conferences/160615prog.aspx.(3) This summary does not represent the views of the BoE, the Monetary Policy

Committee, the Financial Policy Committee, the HKMA or the IMF.

Report BoE-HKMA-IMF conference 383

Is the current set of macroprudential policytools adequate?

Participants noted that many macroprudential tools wereavailable for banks, but some asked whether prudential rulesneeded to evolve in light of the growth of the assetmanagement industry and other non-bank financialinstitutions. Others, however, cautioned against expandingthe set of macroprudential policy tools, arguing that the focusshould instead be on maintaining market discipline onnon-banks by keeping them small, so that they could failwithout causing system-wide disruptions.

There was a broad consensus that stress tests were a usefultool for testing the resilience of the banking system, but alsoother parts of the financial system. Some participants notedthat stress tests had ensured that banks were now able toreport their exposures to a particular sector. Some alsoargued that, in the future, stress tests could be used to set thecountercyclical capital buffer (CCB) such that a desired degreeof systemic resilience is achieved at any point in the cycle.Variability in stress-testing scenario design might helpregulators stay ahead of regulatory arbitrage by firms.

Participants discussed which authority should be responsiblefor operating macroprudential policy. Some highlighted thatthe fragmented regulatory structure in the United Statesimpeded the effective operation of macroprudential policytools, and thought that the concentration of regulatorypowers at central banks was appropriate. Central banks thathad powers over macroprudential policy were also less likelyto be compelled to use monetary policy for financial stabilitypurposes. Others, however, argued against this view, pointingout that central banks already had too much power, and thatthey might not be able to assess financial stability risksobjectively, given that monetary policy might be acontributing factor to those risks.

It was also noted that misconduct problems, for example infixed income, currency and commodities (FICC) markets, couldhave macroeconomic consequences. Mechanisms to ensuregreater accountability among individuals and firms operatingin FICC markets were therefore needed.

Finally, participants noted that little progress had been madein reducing tax incentives to take on excessive debt in theform of tax deductibility for interest payments. Although thiswas a fiscal issue, some thought that central banks shouldcollectively voice concerns over tax incentives to issue debtrather than equity, arguing that excessive debt was the rootcause of financial instability.

How should monetary policy andmacroprudential policy be co-ordinated?

Several participants noted that monetary policy should not beused as the primary tool for achieving financial stability. Oneexample cited was the Riksbank’s attempt to use monetarypolicy to ‘lean against the wind’ (LATW) in the face of risinghouse prices and household debt: some argued that itultimately generated below-target inflation, higherunemployment and probably even higher real debt. Hence,cost-benefit analysis was needed before monetary policy isused for LATW.

Others, however, were sceptical that this principle could beapplied, in particular to emerging market economies (EMEs),which typically had less developed financial systems, wereprone to rapid credit growth, and had a shorter history ofmonetary and fiscal credibility. Pointing to the vulnerabilitiesthat such structural weaknesses create, some argued thatEMEs should use all policy levers — including monetary policy— to maintain macroeconomic and financial stability, in linewith the so-called ‘precautionary principle’. Some participantsalso argued that any quantitative cost-benefit analysis toassess the appropriateness of using monetary policy to LATWwas unlikely to be credible, given the high degree ofuncertainty over the monetary policy strategies of advancedeconomies and the spillover mechanisms from theirconventional and unconventional monetary policies.

Participants also noted that the effectiveness ofmacroprudential policy tools was, as yet, largely untested. Forexample, the CCB could enhance the resilience of banks toshocks, but it was not yet clear how effective it would be intaming the credit cycle. Others questioned whethercountercyclical macroprudential policies would be used in thefirst place: for example, the institutional set-up in theUnited States might prevent such policies and therefore leadto more pressure on monetary policy to fill the gap.

What has been the aggregate impact ofpost-crisis regulatory reforms on the globalfinancial system?

Several participants noted that the new regulatory frameworkwas highly complex, and that the aggregate impact of all thepolicy reforms could be assessed properly only after sometime. In particular, there was a high degree of uncertaintyover how the Basel III leverage ratio and the risk-weightedcapital ratio regulations would jointly affect banks’ incentives.Some expressed concerns that the leverage ratio couldencourage banks to take greater risks in order to maintain ahigh return on equity, while others emphasised its financialstability benefits.

It was argued by some that, while the core of the financialsystem — the banking sector — was likely to have becomemore resilient, the combination of new prudentialrequirements on dealers and structural changes in marketsmight have reduced market depth and increased asset pricevolatility. Policymakers needed to be alert to thesedevelopments, including their consequences for investmentfunds that offer daily liquidity while investing in securities thatmay not turn out to be liquid during periods of market stress.

Several participants noted that risks were already migratinginto the non-bank sector in many countries in response totighter bank regulations. But views were mixed as to whetherthat was a good thing. For some participants, aspects ofnon-banks’ activities needed to be regulated better, as theywere responsible for a large part of credit provision in someeconomies. But others thought that it would be better if riskyactivities moved outside the banking sector, and the bankingsector focused on its core business of supplying credit to thereal economy. For example, some expressed support for theregulatory regime that discouraged banks from holdingsecuritised debt through higher risk weights, while allowingnon-banks to take these risks onto their balance sheets.

An open question going forward was how central banks shouldsupport liquidity in key financial markets. Some argued thatsafety nets should be extended to some non-banks inexchange for closer supervision and regulation, while othersargued that the aim should be to allow non-banks to fail safelywithout causing contagion to banks.

Is there a need for international policyco-ordination to mitigate global risks?

Participants noted that monetary policy in major advancedeconomies and global risk appetite were the two main, andinterrelated, drivers of global capital flows in general, and thecorrelation of bond markets in particular. EMEs were therefore

particularly exposed to the vagaries of global policy cycles andinvestor behaviour.

For some, this called for increased availability of internationalcontingent credit lines and better risk-sharing mechanisms. Intheir absence, EMEs would simply self-insure, with competitivepolicy easing going hand in hand with competitive reserveaccumulation. But several participants noted thatinternational monetary policy co-ordination had previouslybackfired and that there was no clear need for co-ordinatingmonetary policies now.

On macroprudential policies, there was disagreement as towhether more co-ordination was a priority, given differingviews on the size of policy spillovers. However, someparticipants argued that recent changes in financial regulationdid have an effect on market liquidity in EMEs. Moreover, theeffects of capital controls by countries on each other wereidentifiable in the data. It was also noted that there was aclear need for international co-operation in making CCPs safer.

Participants also noted that EMEs had several tools at theirdisposal to defend themselves against external risks. Forexample, they could increase the resilience of their financialsystem and economy against external shocks by strengtheningfinancial regulation and deploying macroprudential policies;and by monitoring foreign currency exposure of financial andnon-financial corporates more closely and encouraginghedging. Some participants also emphasised the need to usecapital flow management measures at times, in part becausemacroprudential policy may turn out to be insufficientlypowerful to deal with risks associated with capital flows. Itwas noted, however, that the evidence on their effectivenesswas mixed. Some participants therefore concluded thatestablishing sound fiscal policies and good governance, andimplementing structural reforms, could potentially be themost reliable ways for EMEs to achieve resilience againstexternal shocks.


Quarterly Bulletin Appendices 385

Appendices


The articles that have been published recently in theQuarterly Bulletin are listed below. Articles fromDecember 1960 to Winter 2004 are available on theBank’s website at:

www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx.

Articles from Spring 2005 onwards are available at:

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx.

Articles

2012 Q2– How has the risk to inflation from inflation expectations evolved?– Public attitudes to monetary policy and satisfaction with the Bank– Using changes in auction maturity sectors to help identify the impact of QE on gilt yields– UK labour productivity since the onset of the crisis — an international and historical perspective– Considering the continuity of payments for customers in a bank’s recovery or resolution– A review of the work of the London Foreign Exchange Joint Standing Committee in 2011

2012 Q3– RAMSI: a top-down stress-testing model developed at the Bank of England– What accounts for the fall in UK ten-year government bond yields?– Option-implied probability distributions for future inflation– The Bank of England’s Real-Time Gross Settlement infrastructure– The distributional effects of asset purchases– Monetary Policy Roundtable

2012 Q4– The Funding for Lending Scheme– What can the money data tell us about the impact of QE?– Influences on household spending: evidence from the 2012 NMG Consulting survey– The role of designated market makers in the new trading landscape– The Prudential Regulation Authority

2013 Q1– Changes to the Bank of England– The profile of cash transfers between the Asset Purchase Facility and Her Majesty’s Treasury– Private equity and financial stability– Commercial property and financial stability– The Agents’ company visit scores– The Bank of England Bank Liabilities Survey– Monetary Policy Roundtable

2013 Q2– Macroeconomic uncertainty: what is it, how can we measure it and why does it matter?– Do inflation expectations currently pose a risk to the economy? – Public attitudes to monetary policy– Cross-border bank credit and global financial stability– The Old Lady of Threadneedle Street– Central counterparties: what are they, why do they matter and how does the Bank supervise them?– A review of the work of the London Foreign Exchange Joint Standing Committee in 2012

2013 Q3– Macroprudential policy at the Bank of England– Bank capital and liquidity– The rationale for the prudential regulation and supervision of insurers– Recent developments in the sterling overnight money market– Nowcasting world GDP and trade using global indicators– The Natural Rate Hypothesis: an idea past its sell-by date– Monetary Policy Roundtable

2013 Q4– SME forbearance and its implications for monetary and financial stability– Bringing down the Great Wall? Global implications of capital account liberalisation in China– Banknotes, local currencies and central bank objectives– Banks’ disclosure and financial stability– Understanding the MPC’s forecast performance since mid-2010– The financial position of British households: evidence from the 2013 NMG Consulting survey– What can company data tell us about financing and investment decisions?– Tiering in CHAPS

Contents of recent Quarterly Bulletins

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx


www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx

www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx


– The foreign exchange and over-the-counter interest rate derivatives market in the United Kingdom– Qualitative easing: a new tool for the stabilisation of financial markets

2014 Q1– Money in the modern economy: an introduction– Money creation in the modern economy– The Court of the Bank of England– Dealing with a banking crisis: what lessons can be learned from Japan’s experience?– The role of business model analysis in the supervision of insurers– Nowcasting UK GDP growth– Curiosities from the vaults: a Bank miscellany– Monetary Policy Roundtable

2014 Q2– The UK productivity puzzle– The Bank of England as a bank– Credit spreads: capturing credit conditions facing households and firms– Assessing the risk to inflation from inflation expectations– Public attitudes to monetary policy– How have world shocks affected the UK economy?– How has the Liquidity Saving Mechanism reduced banks’ intraday liquidity costs in CHAPS?– Risk managing loan collateral at the Bank of England– Sterling Monetary Framework Annual Report 2013–14– A review of the work of the London Foreign Exchange Joint Standing Committee in 2013

2014 Q3– Innovations in payment technologies and the emergence of digital currencies– The economics of digital currencies– How might macroprudential capital policy affect credit conditions?– Household debt and spending– Enhancing the resilience of the Bank of England’s Real-Time Gross Settlement infrastructure– Conference on Monetary and Financial Law– Monetary Policy Roundtable– Changes to the Bank’s weekly reporting regime

2014 Q4– Bank funding costs: what are they, what determines them and why do they matter?– Why is the UK banking system so big and is that a problem?– The interaction of the FPC and the MPC– The Bank of England’s approach to resolving failed institutions

– The potential impact of higher interest rates on the household sector: evidence from the 2014 NMG Consulting survey

2015 Q1– Investment banking: linkages to the real economy and the financial system– Desperate adventurers and men of straw: the failure of City of Glasgow Bank and its enduring impact on the UK banking system– Capital in the 21st century– The Agencies and ‘One Bank’– Self-employment: what can we learn from recent developments?– Flora and fauna at the Bank of England– Big data and central banks

2015 Q2– Mapping the UK financial system– Banking sector interconnectedness: what is it, how can we measure it and why does it matter?– The prudential regulation of insurers under Solvency II– A bank within a bank: how a commercial bank’s treasury function affects the interest rates set for loans and deposits– Do inflation expectations currently pose a risk to inflation?– Innovations in the Bank’s provision of liquidity insurance via Indexed Long-Term Repo (ILTR) operations– A review of the work of the London Foreign Exchange Joint Standing Committee in 2014

2015 Q3– How has cash usage evolved in recent decades? What might drive demand in the future?– Bank failure and bail-in: an introduction– Insurance and financial stability– How much do UK market interest rates respond to macroeconomic data news?– Estimating market expectations of changes in Bank Rate– Over-the-counter (OTC) derivatives, central clearing and financial stability– Monetary Policy Roundtable

2015 Q4– Bonus regulation: aligning reward with risk in the banking sector– The Prudential Regulation Authority’s secondary competition objective– Trends in UK labour supply– The potential impact of higher interest rates and further fiscal consolidation on households: evidence from the 2015 NMG Consulting survey– BoE-HKMA-IMF conference on monetary, financial and prudential policy interactions in the post-crisis world


The Bank of England publishes information on all aspects of its work in many formats. Listed below are some of themain Bank of England publications. For a full list, please referto our website:

www.bankofengland.co.uk/publications/Pages/default.aspx.

Staff working papers

An up-to-date list of staff working papers is maintained on the Bank of England’s website at:

www.bankofengland.co.uk/research/Pages/workingpapers/default.aspx

where abstracts of all papers may be found. Papers publishedsince January 1997 are available in full, in portable documentformat (PDF).

No. 556 A sectoral framework for analysing money, credit andunconventional monetary policy (October 2015)James Cloyne, Ryland Thomas, Alex Tuckett and Samuel Wills

No. 557 The banks that said no: banking relationships, creditsupply and productivity in the United Kingdom (October 2015)Jeremy Franklin, May Rostom and Gregory Thwaites

No. 558 Bankers’ pay and excessive risk (October 2015)John Thanassoulis and Misa Tanaka

No. 559 Stabilising house prices: the role of housing futurestrading (October 2015)Arzu Uluc

No. 560 Mortgage debt and entrepreneurship (October 2015)Philippe Bracke, Christian Hilber and Olmo Silva

No. 561 Threshold-based forward guidance: hedging the zerobound (October 2015)Lena Boneva, Richard Harrison and Matt Waldron

No. 562 International banking and liquidity risk transmission:lessons from the United Kingdom (November 2015)Robert Hills, John Hooley, Yevgeniya Korniyenko andTomasz Wieladek

No. 563 Extreme risk interdependence (November 2015)Arnold Polanski and Evarist Stoja

No. 564 Why are real interest rates so low? Secularstagnation and the relative price of investment goods(November 2015)Gregory Thwaites

No. 565 Ambiguity, monetary policy and trend inflation(November 2015)Riccardo M Masolo and Francesca Monti

No. 566 The Great Recession and the UK labour market(November 2015)Stephen Millard

No. 567 A new approach to multi-step forecasting usingdynamic stochastic general equilibrium models(November 2015)George Kapetanios, Simon Price and Konstantinos Theodoridis

External MPC Unit discussion papers

The MPC Unit discussion paper series reports on researchcarried out by, or under supervision of, the external membersof the Monetary Policy Committee. Papers are available fromthe Bank’s website at:

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx.

The following papers have been published recently:

No. 42 What are the macroeconomic effects of assetpurchases? (April 2014)Martin Weale and Tomasz Wieladek

No. 43 The shocks matter: improving our estimates ofexchange rate pass-through (November 2015)Kristin Forbes, Ida Hjortsoe and Tsvetelina Nenova

Monetary and Financial Statistics

Monetary and Financial Statistics (Bankstats) contains detailed information on money and lending, monetary andfinancial institutions’ balance sheets, banks’ income andexpenditure, analyses of bank deposits and lending, externalbusiness of banks, public sector debt, money markets, issues of securities, financial derivatives, interest and exchange rates, explanatory notes to tables and occasional relatedarticles.

Bankstats is published on a monthly basis, free of charge, onthe Bank’s website at:

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx.

Further details are available from the Statistics and RegulatoryData Division, Bank of England: telephone 020 7601 5432;email [email protected].

Bank of England publications

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx




Articles that have been published in recent issues of Monetary and Financial Statistics can also be found on theBank’s website at:

www.bankofengland.co.uk/statistics/Pages/ms/articles.aspx.

Financial Stability Report

The Financial Stability Report is published twice a year underthe guidance of the Financial Policy Committee (FPC). Itcovers the Committee’s assessment of the outlook for thestability and resilience of the financial sector at the time ofpreparation of the Report, and the policy actions it advises toreduce and mitigate risks to stability. The Bank of Englandintends this publication to be read by those who areresponsible for, or have interest in, maintaining and promotingfinancial stability at a national or international level. It is ofespecial interest to policymakers in the United Kingdom andabroad; international financial institutions; academics;journalists; market infrastructure providers; and financialmarket participants. The Financial Stability Report is availableat:

www.bankofengland.co.uk/publications/Pages/fsr/default.aspx.

Handbooks in central banking

The series of Handbooks in central banking provide concise,balanced and accessible overviews of key central bankingtopics. The Handbooks have been developed from studymaterials, research and training carried out by the Bank’sCentre for Central Banking Studies (CCBS). The Handbooksare therefore targeted primarily at central bankers, but arelikely to be of interest to all those interested in the varioustechnical and analytical aspects of central banking. TheHandbook series also includes ‘Technical Handbooks’ which areaimed more at specialist readers and often contain moremethodological material than the Handbooks, incorporatingthe experiences and expertise of the author(s) on topics thataddress the problems encountered by central bankers in theirday-to-day work. All the Handbooks are available via theBank’s website at:

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx.

The Bank of England’s Sterling MonetaryFramework (the ‘Red Book’)

The ‘Red Book’ describes the Bank of England’s framework forits operations in the sterling money markets, which is designedto implement the interest rate decisions of the Monetary

Policy Committee while meeting the liquidity needs, and socontributing to the stability of, the banking system as a whole.It also sets out the Bank’s specific objectives for theframework, and how it delivers those objectives. Theframework was introduced in May 2006. The ‘Red Book’ isavailable at:

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf.

Cost-benefit analysis of monetary andfinancial statistics

The handbook describes a cost-benefit analysis (CBA)framework that has been developed within the Bank to ensurea fair balance between the benefits derived from good-qualitystatistics and the costs that are borne by reporting banks.Although CBA is a well-established approach in othercontexts, it has not often been applied to statistical provision,so techniques have had to be adapted for application to the Bank’s monetary and financial statistics. The handbook alsodiscusses how the application of CBA has enabled cuts in boththe amount and the complexity of information that is requiredfrom reporting banks.

www.bankofengland.co.uk/statistics/Pages/about/cba.aspx.

Credit Conditions Survey

Developments in credit conditions are of key interest to theBank of England in its assessment of economic conditions.This quarterly survey of bank and building society lenders is aninput to this assessment. The survey covers secured andunsecured lending to households and small businesses; andlending to non-financial corporations, and to non-bankfinancial firms. Copies are available on the Bank’s website at:

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx.

Bank Liabilities Survey

Developments in lenders’ balance sheets are of key interest tothe Bank of England in its assessment of economic conditions.The aim of this quarterly survey of banks and building societylenders is to improve understanding of the role of lenders’liabilities and capital in driving credit and monetary conditions.The survey covers developments in the volume and price ofbank funding; developments in the loss-absorbing capacity ofbanks as determined by their capital positions; anddevelopments in the internal price charged to business unitswithin individual banks to fund the flow of new loans,

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx


sometimes referred to as the ‘transfer price’. Copies areavailable on the Bank’s website at:

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx.

Credit Conditions Review

This quarterly publication presents the Bank of England’sassessment of the latest developments in bank fundingconditions and household and corporate credit. It drawsmainly on long-established official data sources, such as theexisting monetary and other financial statistics collected bythe Bank, and other data sources such as surveys of businessesand data from other organisations. The analysis also draws onthe results of the Bank of England’s Bank Liabilities Survey andCredit Conditions Survey. Copies are available on the Bank’swebsite at:

www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx.

Quarterly Bulletin

The Quarterly Bulletin explores topical issues relating to theBank’s core purposes of monetary and financial stability.Some articles present analysis on current economic andfinancial issues, and policy implications. Other articlesenhance the Bank’s public accountability by explaining theinstitutional structure of the Bank and the various policyinstruments that are used to meet its objectives. TheQuarterly Bulletin is available at:

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx.

Inflation Report

The Bank’s quarterly Inflation Report was first published in1993. The Report sets out the detailed economic analysis and inflation projections on which the Bank’s Monetary Policy Committee bases its interest rate decisions, andpresents an assessment of the prospects for UK inflation. TheInflation Report is available at:

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx.

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx



www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx

www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx