capital flows tto eng 170614.ppt - trinidad and tobago dollar capital_flows_tto...capital flows in...
TRANSCRIPT
Drivers of Capital Flows and the Role of the Investor Base in Latin America
Port of Spain – June 14, 2017
Carlos Caceres(with Carlos Gonçalves, Gen Lindow, and Galen Sher)
International Monetary Fund
Capital Flows in Emerging Markets
2
• Questions addressed here: What are the main drivers of capital flows in emerging
markets? Are there any factors that can help mitigate or amplify the
sensitivity of capital flows to external shocks?
• IMF-WHD work on capital flows: Spring 2017 Regional Economic Outlook (Chapter 4) Caceres et al (2017 A; forthcoming): “Understanding the Cyclical
and Structural Drivers of Capital Flows in Emerging Markets” Caceres et al (2017 B; forthcoming): “The Role of the Investor
Base in the Sensitivity of Capital Flows to External Shocks”
OUTLINE
3
Part I: Stylized Facts of Capital Flows- Recent developments- Cyclical behavior of capital flows- Structural component of capital flows
Part II: Analyzing the Drivers of Capital Flows- Global ‘push’ factors- Country-specific ‘pull’ factors
Part III: Sensitivity of Capital Flows to External Shocks- The role of the “Investor Base”
Part IV: Conclusions
I. Stylized Facts
Gross Inflows to Latin America
5
Capital flows to LAC have halved since the end of the “commodity super-cycle”
LA7: Gross Inflows(Billions of U.S. dollars; 4-quarter moving averages)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
-50
0
50
100
150
200
250
300
350
400
450
2003:Q1 05:Q1 07:Q1 09:Q1 11:Q1 13:Q1 15:Q1
FDIPortfolioOther investmentFinancial derivativesTotal
360bn
175bn
Gross Inflows to TTO
6
…though TTO’s story is quite different; reflecting the country’s large CA surpluses over this period
TTO: Gross Inflows(Billions of U.S. dollars; 4-quarter moving averages)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.
-8
-6
-4
-2
0
2
4
2002 2004 2006 2008 2010 2012 2014 2016
FDI Portfol ioOther investment Total
Gross Inflows to Latin America
7
“Low frequency” of capital inflow cycles in LAC
-2
0
2
4
6
8
10
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
LA7: Gross Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
Gross Inflows and Outflows in LAC
8
Gross outflows only partially offsetting gross inflows…
-2
0
2
4
6
8
10
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
Gross inflows Gross outflows
LA7: Gross Inflows and Outflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
Gross and Net Inflows in LAC
9
Overall, gross inflows tend to drive net inflows in LAC
-2
0
2
4
6
8
10
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
Gross inflows Net inflows
LA7: Gross and Net Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
Gross and Net Inflows in Other EMs
10
The predominant role of gross inflows over gross outflows is also observed in other emerging markets…
Other Emerging Markets: Capital Flows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.
-2
0
2
4
6
8
10
12
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
Gross inflowsNet inflowsGross outflows
Gross and Net Inflows in TTO
11
…and this is also the case in TTO
TTO: Capital Flows(Percent of trend GDP)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.
-7-6-5-4-3-2-10123
2002 2004 2006 2008 2010 2012 2014 2016
Thou
sand
s
Gross inflows Net inflows
Gross outflows
Cyclical Behavior of Capital Flows
12
Capital flows to LAC exhibit a common cyclical componentLA7: Gross Inflows(Percent of trend GDP)
Sources: Haver Analytics; IMF, Balance of Payment Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.
Argentina Brazil Chile
MexicoColombia Peru
-12
-8
-4
0
4
8
12
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1-2
0
2
4
6
8
10
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1-4
0
4
8
12
16
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
-202468
1012
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q10
2
4
6
8
10
12
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q10
3
6
9
12
15
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
Total gross inflows Average (2000–16) LA7 median
Cyclical Behavior of Capital Flows
13
…which is mirrored by the high synchronicity observed in other macroeconomic and financial variables
First Principal Component(Percent share explained)
Sources: IMF, Balance of Payment Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.Note: For the exchange rate, all euro area countries are treated as a single entity.
0
20
40
60
80
100
LA5 LA7 Advanced OEM
Gross inflows Real GDP growth Exchange rate
Structural Behavior of Capital Flows
14
…but average levels of capital inflows differ substantially across countries
Gross Inflows Average 2000-16(Percent of trend GDP)
Haver Analytics; IMF, Balance of Payments Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.
Haver Analytics; IMF, Balance of Payments Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.
Net Inflows Average 2000-16(Percent of trend GDP)
-8
-4
0
4
8
12
16
VN
MKA
ZA
LBBG
RG
HA
HU
NH
RV CH
LR
OU
UR
YPO
LD
OM
TUR
PER
CO
LM
EXM
YS
CH
NIN
DBR
ASA
UR
US
ZAF
MA
RPH
LTH
AID
NEG
YPR
YA
RG
TTO
Latin AmericaOther emerging markets
VNM = 29.1
-10
-6
-2
2
6
10
14
VN
MA
LBBG
RG
HA
HRV
RO
UD
OM
TUR
POL
PER
HU
NU
RY
IND
MEX ZA
FC
OL
KAZ
BRA
CH
NPR
YM
AR
EGY
CH
LPH
LID
NTH
AA
RG
SAU
RU
SM
YS
TTO
Latin AmericaOther emerging markets
VNM = 19.5
II. Drivers of Capital Flows
Gross inflows in LAC and other EMs
16
Quantitative analysis focused on the period 2000-2016
-2
0
2
4
6
8
10
2000 2002 2004 2006 2008 2010 2012 2014 2016
LA7: Gross Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: OEM = Other emerging markets.
-2
0
2
4
6
8
10
12
2000 2002 2004 2006 2008 2010 2012 2014 2016
OEM: Gross Inflows(Percent of trend GDP; median)
Capital Inflows & Global Growth
17
High correlation of gross inflows and global GDP growth
-8
-6
-4
-2
0
2
4
6
-2
0
2
4
6
8
10
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows G7 real GDP growth (right scale)
LA7: Gross Inflows and G7 Real GDP Growth(Percent of trend GDP; percent change; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
OEM: Gross Inflows and G7 Real GDP Growth(Percent of trend GDP; percent change; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: OEM = Other emerging markets.
-8
-6
-4
-2
0
2
4
6
-2
0
2
4
6
8
10
12
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows G7 real GDP growth (right scale)
Capital Inflows & Domestic Growth
18
High correlation of gross inflows and real GDP growth
-4
-2
0
2
4
6
8
10
-2
0
2
4
6
8
10
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows Real GDP growth (right scale)
LA7: Gross Inflows and Real GDP Growth(Percent of trend GDP; percent change; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
OEM: Gross Inflows and Real GDP Growth(Percent of trend GDP; percent change; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: OEM = Other emerging markets.
-2
-1
0
1
2
3
4
5
6
7
8
-2
0
2
4
6
8
10
12
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows Real GDP growth (right scale)
Capital Inflows & Commodity Prices
19
Gross inflows track the “commodity super cycle” well
3.8
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
-2
0
2
4
6
8
10
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows Commodity prices (right scale)
LA7: Gross Inflows and Commodity Prices(Percent of trend GDP; logs; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA7 = Argentina, Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
OEM: Gross Inflows and Commodity Prices(Percent of trend GDP; logs; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: OEM = Other emerging markets.
3.8
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
5.6
-2
0
2
4
6
8
10
12
2000 2002 2004 2006 2008 2010 2012 2014 2016
Gross inflows Commodity prices (right scale)
Capital Inflows & Structural Factors
20
Average level of capital inflows correlated with institutional and other structural factors
Gross Inflows Vs Regulatory Quality(Average 2000-16)
Sources: Haver Analytics; IMF, Balance of Payment Statistics Yearbook database; World Governance Indicators; and IMF staff calculations.Note: “Regulatory quality" and "rule of law" are composite indicators from the World Bank's Worldwide Governance Indicators (WGI); "political risk index" is computed within the International Country Risk Guide (ICRG) database."
ARG
BRACHL
COL
MEX
PER URY
0
4
8
12
16
-1 0 1 2
Gro
ss In
flow
s (p
erce
nt o
f tre
nd G
DP)
Regulatory Quality
Other emerging marketsLA7
ARG
BRA
CHL
COL
MEX
PERURY
0
4
8
12
16
-1 0 1 2
Gro
ss In
flow
s (p
erce
nt o
f tre
nd G
DP)
Rule of Law
Other emerging marketsLA7
Gross Inflows Vs Rule of Law(Average 2000-16)
ARG
BRA
CHL
COL
MEX
PERURY
0
4
8
12
16
50 60 70 80
Gro
ss In
flow
s (p
erce
nt o
f tre
nd G
DP)
Political Risk Index (higher = less risk)
Other emerging marketsLA7
Gross Inflows Vs Political Risk(Average 2000-16)
‘Pull’ and ‘Push’ Factors
21
• Panel of 29 emerging markets countries (incl. LA7)
• Period: 2000Q1-2016Q3
• Dependent variable(s): gross and net capital flows
• Cyclical behavior of capital flows:– Global ‘push’ factors– Country-specific ‘pull’ factors
• Structural behavior of capital flows:– Country-specific ‘pull’ factors (low time variability; high
cross-sectional variability)
Cyclical Drivers of Capital Inflows
22
Predominant role of commodity prices in driving the cyclical component of capital flow cycles…
Sources: IMF staff calculations; Caceres et al (2017; forthcoming A). Notes: green denotes a positive coefficient estimate; red denotes a negative coefficient estimate. *, **, *** denote statistical significance at the 10, 5, and 1 percent confidence level.
Individual effect
Joint effect
Individual effect
Joint effect
Individual effect
Joint effect
Global variables
VIX ** - - - - -
Global GDP growth ** ** - - - -
U.S. interest rates - - - - - *
Global commodity prices *** *** *** *** *** ***
Country-specific variables
GDP growth differential ** - *** *** *** ***
Interest rate differential *** - * - *** -
Latin America Other Emerging Markets All Emerging Markets
“We Just Ran A Million Regressions”
23
• Objective: minimize ‘model selection’ bias
• In the spirit of Sala-i-Martin (1997, AER)
• Idea: to avoid selecting a particular (non-random) model, can consider N potential explanatory variables, then regress dependent variable on all (2N-1) possible combinations… This yields a histogram of estimated coefficients
• The ‘most robust’ variables would tend to have their distribution mass to the right/left of the “zero” line
Robustness of Global Factors
24
Commodity prices robust to alternative model specifications
Sources: IMF staff calculations; Caceres et al (2017; forthcoming A). Note: LA5 = Brazil, Chile, Colombia, Mexico, Peru.
0
2,000
4,000
6,000
8,000
10,000
-7 -5 -2 0 2 4 7
LA5Other emerging markets
Commodity prices, log(Frequency)
0
2,000
4,000
6,000
8,000
10,000
-9 -6 -4 -1 1 3 6 8
LA5Other emerging markets
VIX, log(Frequency)
G7 Real GDP growth(Frequency)
0
2,000
4,000
6,000
8,000
10,000
12,000
-1.1 -0.6 0.0 0.5 1.1
LA5Other emerging markets
0
2,000
4,000
6,000
8,000
10,000
-2.1 -1.6 -1.1 -0.5 0.0 0.6 1.1 1.6
LA5Other emerging markets
U.S. short-term interest rates(Frequency)
Structural Drivers of Capital Inflows
25
Country-specific structural factors strongly associated with higher average levels capital flows…
Sources: IMF staff calculations; Caceres et al (2017; forthcoming A). Notes: green denotes a positive coefficient estimate; red denotes a negative coefficient estimate. *, **, *** denote statistical significance at the 10, 5, and 1 percent confidence level.
Latin America Other Emerging Markets All Emerging Markets
Country-specific structural factors
Government effectiveness *** *** ***
Regulatory quality *** *** ***
Control of corruption *** *** ***
Rule of law *** *** ***
Law and order *** *** ***
Voice an accountability *** *** ***
Political stability *** *** ***
Political risk *** *** ***
Institutionalized democracy * *** ***
Corporate tax rate *** *** ***
Credit rating *** *** ***
III. Sensitivity of Capital Flows:The Role of the ‘Investor Base’
Response of Capital Flows to Shocks
27
High vulnerability of capital inflows to external shocksImpulse Response Functions of Gross Inflows:
Unrestricted Sample(Percent of trend GDP)
Source: IMF staff calculations; Caceres et al (2017; forthcoming B).Note: Bars represent the response of gross inflows four quarters after a positive one unit shock to each of the four exogenous global variables. A unit shock for the VIX and global commodity prices correspond to a unit change in the logarithm of these two variables; a unit shock for G7 growth and the “U.S. monetary shock” corresponds to a 1 percentage point change in these variables. Solid bars with black borders denote that the response is statistically significant at the 5 percent confidence level. G7 = Canada, France, Germany, Italy, Japan, United Kingdom, United States; VIX = Chicago Board Options Exchange Volatility Index.
-6
-3
0
3
6
9
12
Commodityprices
VIX G7 growth U.S. monetaryshock
Role of the Investor Base
28
Higher participation of non-resident investors increases the sensitivity of capital flows to external shocks…
Impulse Response Functions of Gross Inflows: Non-Residents’ Participation in Domestic Markets
(Percent of trend GDP)
Source: IMF staff calculations.Note: Bars represent the response of gross inflows four quarters after a positive one unit shock to each of the four exogenous global variables. A unit shock for the VIX and global commodity prices correspond to a unit change in the logarithm of these two variables; a unit shock for G7 growth and the “U.S. monetary shock” corresponds to a 1 percentage point change in these variables. Solid bars with black borders denote that the response is statistically significant at the 5 percent confidence level. G7 = Canada, France, Germany, Italy, Japan, United Kingdom, United States; VIX = Chicago Board Options Exchange Volatility Index.
-6
-3
0
3
6
9
12
15
18
Commodityprices
VIX G7 growth U.S. monetaryshock
High Low
Role of the Investor Base
29
Deeper and more liquid domestic markets lower the sensitivity of capital flows to external shocks…
Impulse Response Functions of Gross Inflows: Domestic Market Capitalization
(Percent of trend GDP)
Source: IMF staff calculations; Caceres et al (2017; forthcoming B).Note: Bars represent the response of gross inflows four quarters after a positive one unit shock to each of the four exogenous global variables. A unit shock for the VIX and global commodity prices correspond to a unit change in the logarithm of these two variables; a unit shock for G7 growth and the “U.S. monetary shock” corresponds to a 1 percentage point change in these variables. Solid bars with black borders denote that the response is statistically significant at the 5 percent confidence level. G7 = Canada, France, Germany, Italy, Japan, United Kingdom, United States; VIX = Chicago Board Options Exchange Volatility Index.
-6
-3
0
3
6
9
12
15
18
Commodityprices
VIX G7 growth U.S. monetaryshock
High Low
Role of the Investor Base
30
A large stable domestic investor base can also help reduce the sensitivity of capital flows to external shocks…
Impulse Response Functions of Gross Inflows: Domestic Pension Funds Presence
(Percent of trend GDP)
Source: IMF staff calculations; Caceres et al (2017; forthcoming B).Note: Bars represent the response of gross inflows four quarters after a positive one unit shock to each of the four exogenous global variables. A unit shock for the VIX and global commodity prices correspond to a unit change in the logarithm of these two variables; a unit shock for G7 growth and the “U.S. monetary shock” corresponds to a 1 percentage point change in these variables. Solid bars with black borders denote that the response is statistically significant at the 5 percent confidence level. G7 = Canada, France, Germany, Italy, Japan, United Kingdom, United States; VIX = Chicago Board Options Exchange Volatility Index.
-6
-3
0
3
6
9
12
15
Commodityprices
VIX G7 growth U.S. monetaryshock
High Low
Role of Policy Choices
31
Exchange rate flexibility can also contribute to lower the sensitivity of capital flows to external shocks
Impulse Response Functions of Gross Inflows: Exchange Rate Flexibility
(Percent of trend GDP)
Source: IMF staff calculations; Caceres et al (2017; forthcoming B).Note: Bars represent the response of gross inflows four quarters after a positive one unit shock to each of the four exogenous global variables. A unit shock for the VIX and global commodity prices correspond to a unit change in the logarithm of these two variables; a unit shock for G7 growth and the “U.S. monetary shock” corresponds to a 1 percentage point change in these variables. Solid bars with black borders denote that the response is statistically significant at the 5 percent confidence level. G7 = Canada, France, Germany, Italy, Japan, United Kingdom, United States; VIX = Chicago Board Options Exchange Volatility Index.
-8
-4
0
4
8
12
16
Commodityprices
VIX G7 growth U.S. monetaryshock
Fixed Flexible
IV. Conclusions
Conclusions:
33
• Capital flows in emerging markets exhibit a largely common cyclical behavior, but a relatively differentiated structural component
• Predominant role of global factors - particularly commodity prices - in driving the cyclical behavior of capital flows in emerging markets
• Country-specific structural factors, such as good governance and strong institutional and regulatory frameworks, play a key role in attracting inflows over longer time horizons
Conclusions:
34
• Countries with high ‘foreign participation’ tend to exhibit higher sensitivity of capital flows to external shocks…
• …whereas countries with deeper and more liquid financial markets, with a large stable domestic investor base, appear to be less sensitive to shocks
• Exchange rate flexibility is another line of defense to lower the sensitivity of capital flows to external shocks
Thank You!
Auxiliary Slides:
Gross Inflows to Latin America
37
Similar behavior of capital inflows across LA5 countries
-5
-4
-3
-2
-1
0
1
2
3
-2
0
2
4
6
8
10
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
Median Mean Principal component (right scale)
LA5: Gross Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.Note: LA5 = Brazil, Chile, Colombia, Mexico, Peru.
Subcomponents of Capital Flows
38
Sub-components of capital flows are positively correlated with one another and with total flows…
LA7: Gross Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.
-2
0
2
4
6
8
10
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
Total Flows FDIPortfol io Other investment
Subcomponents of Capital Flows
39
…and FDI tends to be larger in magnitude, whereas portfolio (and other) flows tend to be more volatile
Other Emerging Markets: Gross Inflows(Percent of trend GDP; median)
Sources: IMF, Balance of Payments Statistics Yearbook database; Haver Analytics; and IMF staff calculations.
-2
0
2
4
6
8
10
12
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
Total Flows FDIPortfol io Other investment
Cyclical Behavior of Capital Flows
40
Broadly similar behavior observed across regions
Gross Inflows (Percent of trend GDP; median)
Haver Analytics; IMF, Balance of Payments Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.
Haver Analytics; IMF, Balance of Payments Statistics Yearbook database; IMF, World Economic Outlook database; and IMF staff calculations.
0
2
4
6
8
10
12
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
LA7 Other emerging markets
0
4
8
12
16
20
-4
-2
0
2
4
6
8
10
2000:Q1 04:Q1 08:Q1 12:Q1 16:Q1
LA7 Emerging Asia Emerging Europe (rhs)