does working capital management affect profitability of belgian firms

8/12/2019 Does working capital management affect profitability of belgian firms

1/16

joimial of usinessFiiunife if .Aicmmting 30(.^) & 4), April/May 20 03 , 030 6-6 86X

Does Working Capi ta l ManagementAffect Profitability of BelgianFirms?

M A R C D E L O O F *

I. INTRODUCTIONMost firms bave a large amount of cash invested in workingcapital, as well as substantial am ounts of short-term payables as asource of financing. For instance, according to the National Bankof Belgium, in 1997 accounts receivable and inventories wererespectively 17% and 10% of total assets of all Belgian non-financial firms. Accounts payable were 13%oftotalassets of thesefirms. It can be expected tbat the way in which working capital ismanaged will have a significant impact on the profitability offitms. Accordingly, for many firms working capital management( \ V ( ; M ) is a very important component of their financialmanagement.Firms may have an optimal level of working capital thatmaximizes their value. On the one hand, large inventory and agenerous tiade credit policy may lead to higher sales. Largerinventory reduces the risk of a stock-out. Trade credit maystimulate sales because it allows customers to assess productquality before paying (Long, Malitz and Ravid, 1993; andDeloof and Jegers, 1996). Because suppliers may have


2/16

574 DELOOFsignificant cost advantages over financial institutions inprov iding credit to their custom ers, it can also be an inexpensivesource of credit for customers {Petersen and Rajan, 1997). Theflip side of granting trade credit and keeping inventories is thatmoney is locked up in working capital.Another component of working capital is accounts payable.Delaying payments to suppliers allows a firm to assess thequality of the products bought, and can be an inexpensive andflexible source of fmancing for the firm. On the other hand, latepayment of invoices can be very costly if the Rrm is offered adiscount for early payment. In a 1996 survey of trade creditpolicies in Europe, Svensson (1997) found that 75% of Belgianfirms offered a'discou nt for pro m pt pay ment {within 10 days),and the average discount offered was 3% . T he averageontr tu l credit period was 41 days, but the average ctu payment period was6 days: 49% of all trad e credit was paid toolate. In a 2001 survey by the Institute for Credit Management ofthe Vlerick Leuven Ghent School for Management of tradecredit policies of Belgian fhms, it was found that the medianBelgian firm offers a 2/10n30 discount for quick payment.

Compared with payment periods in the USA, paymentperiods of Belgian firms appear quite long. In Belgium, as inmany other continental European countries, capital markets areunderdeveloped in the sense that information and agencyproblems are particularly pro no un ced . La Poita et al. {1997 and1998), who find that countries with poorer investor protectionhave smaller and narrower external capital markets, show thatBelgium has weak legal protections of corporate shareholdersand creditors, making bank financing and trade credit moreattractive. Fisnian and Love {2001) argue that trade creditorsmitigate weak creditor protection and imperfect informationbetter than formal lenders, and find that firms in countries withless developed financial markets use informal credit pr(n ided bytheir suppliers to finance growth.A popular measure of WCM is the cash conversion cycle, i.e.


3/16

WORKING CAPITAL MANAGEMENT AND PROFITABILITY 5 7 5also decrease with the cash conversion cycle, ifth e costs of higherinvestment in working capital rise faster than the benefits ofholding more inventory and/or granting more trade credit tocnstomers. Shin and Soenen (1998) investigate the relationbetween a measure of the cash conversion cycle and corporateprofitability. For a large sample of listed American firms for the1975-1994 period, they find a .strong negative relation. Thisresult indicates that managers can create value for theirshareholders by reducing the cash conversion cycle to areasonable minimum.

In this paper, I investigate the relation between WCM andcorporate profitability for a sample of 1,009 large Belgian non-financial firms for the 1992-1996 period. Number of daysaccounts leceivable, inventories and accounts payable are usedas measures of trade credit and inventory policies. The cashconversion cycle is used as a comprehensive measure of WCM.The results suggest that managers can increase corporateprofitability by reducing the number of days accounts receivableand inventories. Less profitable firms wait longer to pay theirbills.The paper proceeds as follows. In the next section the sampleand the variables used in the empirical analysis are presented.The results of the empirical analysis are presented in Section 3.Section 4 discusses tbe causality in the relation between WCMand corporate profitability. Section 5 concludes.

2. SAM PLE AN D VARIABLESThe sample is based on a database provided by the NationalBank of Belgium that consists of financial statements of the2,000 most important Belgian firms. The sample wasconstructed as follows. I started with the 1,637 firms forwhich a financial statement was available for each year of the1991-1996 period. Because of the specific nature of their


4/16

576 DELOOFreceivable, number of days inventories, tiumber of daysaccounts payable, net operating income and gross operatingincome were left out. Thus a balanced panel set of5 045firm-year observations was obtained, with observations of 1,009 firmsover the 1992-1996 period.Profitability is measured by gross operating income, which isdefined as sales minus cash costs of goods sold, and is divided bytotal assets minus financial as.sets.^ For a number of firms in thesample, financial assets, which are mainly shares in other firms,are a significant part of total assets. That is also the reason whyreturn on assets is not considered as a measure of profitability:when a firm has mainly financial assets on its balance sheet, itsoperating activities will contribute little to the overall return onassets. Profitability measures based on stock market valuationare not considered because only a limited number of Belgianfirms is listed on a stock exchange.

Number of days accounts receivable is calculated as [accountsreceivable x 365]/sales. Number of clays inventories is[inventories x 365]/cost of sales. Number of days accountspayable is [accounts payable x 365]/purchases. A more detaileddescription of the definitions of these variables, which involve anumber of Belgian financial statement items, can be found inTheunisse and Jegers (1994).The cash conversion cycle is used as a comprehensivemeasure ofWCM.The cash conversion cycle is simply [numberof days accounts receivable -I- number of days inventorynumber of days accounts payable]. Gentry, Vaidyanathan andLee (1990) developed aweighte cash conversion cycle, whichscales the liming by the amount of funds in each step of thecycle. However, this measure cannot be used because not allinformation necessary for calculation is available. *In addition, size (the natural logarithm of sales), sales growth([this year's sales - previous year's salesj/pievious years sales), thefinancial debt ratio (financial debt/total assets), and the ratio of fixedfinancial assets to total assets are included as control variables in the


5/16

WORKINC, CAIMTAL MANAGEMENT AND PROFn ABILI IV 577

in

Q 1rtC

CZ

CQCTJO

CM T :^dd o c or TfX q,ri d o o d

ooon (N OCTa-^ in cc in _;

CMmCMmo mt =N

O O)CM *_qdd

bc.=

u -Jl3u3 c =^* : CLl0 c Q rt

f Z ? .3 o^ iJ :^^ P 3= .S

rt^ u . _.

p ^ C ^~

O- C r^IJ rt=bX -^ y

rtw IJ' c'ra~1ra

a- trs da

y

--^ Z^ 5

Stol

iala

1 ^

re0D

5rt^1 i^ ji

5sy - i>^^^re

o,.

cia

crtC

j jre

i3

-g-

b lretion

X-p-i1

.= .2 IE


6/16

578 DELOOFI also consider variability of net operating income (divided bytotal assets minus financial assets) as a control variable.Variability is the standard deviation of net operating incomeof each firm over the 1991-1996 period.Table 1 presents descriptive statistics. Gross operating incomeis on average 12.2 of [total assets - financial assets], whilethe median is 10.6 . T he average cash conversion cycle is 44.5days (median is 59.6 days). Firms receive payment on sales afteran average of 54.6 days (the median is 51.4 days). It takes onaverage 46.6 days to sell inventory (median is 33.8 days) and firms

wait on average 56.8 days to pay their purchases (median is 52days).Mean sales growth is only2.H , while median sales growth even less at 1 69 On average about a quarter of all assets arefinanced with financial debt (median is 22.4 ). It is alsonoteworthy that while for the median firm in the sample,fixed financial assets are only 2.4 of total assets, the meanfixed financial assets ratio is much higher at 12 . For a n um be rof firms, a large proportion of total assets are fixed financialassets.

3. EMPIRICAL ANALYSIS i) orrelation AnalysisTable 2 presents Pearson correlation coefficients for all variablesconsidered. There is a negative relation between grossoperating income on the one band and tbe measures of WCM(number of days accounts receivable, inventories and accountspayable and casb conversion cycle) on the other hand. This isconsistent with the view tbat the time lag between tbeexpenditure for tbe purchases of raw materials and thecollection of sales of finished goods can be too long, and thatdecreasing this time lag increases profitability.


7/16

WORKING CAPITM, MANAGEMENT AND PROEITABILITY 579

S 7CT)

H i:0uc0(u

Ert

s

CO

o

qq q I I I I ID XO X...

I

O 5 CX 5~l 5^1r^ ^ j ^ ^

I I

I-- ;c -r -r o1-- X O |-~- O

10 cr. oo ^ c^j CN^^ ^_. _^ ^_,

I I

I I I I

O^ Tf V^ 00X C^ Tf CM t; X qd d d d

tC eo Dq lo CNdi d d

o oI

55h r

ra>

ac. 3.i5

5.CLi

30"p'-'

^ V ^ ^ OJ '^ - (f l _S

'wi ti - CJ ^ qj_ - ^ o5 ? : H Ec ^ ^ .2 be

s/ ^^ d S =

- - o . . S

l j s s

J uE 5

" ? J= ' "

.E ..2 ; i Ii. TJ


8/16

580 DELOOFpolicy, and not vice verm. An alternative explanation for anegative relation between the tiumber of days accounts payableand profitability could be tbat firms wait too ong to pay tbeiraccounts payab e. Speeding up paytiieiits to suppliets migbtincrease profitability because Belgian firms often receive asubstantial discount for prompt payment. However, in Belgianfuiancial statements, discounts received for prompt paytnentsbould be booked as financial iticome, atid .sbould not aflectoperating income.

ii) R egression AnalysisNext, I use regression analysis to investigate the impact ofW Mon corporate profitability. Ih e de tertn inants of corpo rateprofitability are estimated with a fixed effects model. Fixedeffects estimation assumes firm specific intercepts, whichcapture the effects of those variables that are particular to eachfirm and that are constant over time.^ A disadvantage of fixedeffects estimation is that it eliminates anything that is time-invariant from the model. Variability of income, which ismeasured by the standard deviation of net operating incomeover the 1991-1996 period, can therefore not be included in afixed effects model. I also estimate plain OLS-models, which notonly include all variables of the fixed effects tnodel, but alsovariabihty of income, 4 year dummies and 37 industrydummies, which are based on 2-digit NACE-code. In allregressions, standard errors are calculated using White'scorrection for heteroscedasticity.The determinants of gross operating income are investigatedfor all 5 045 firm-year observations. The results can be foundin Table ?>. Regression (1) is estimated with fixed effectsand includes tiumber of days accounts receivable as a measureof accounts receivable policy. The coefficient of the accountsreceivable variable is negative and highly significant, and


9/16

W (iRKIN (; CAPITAL MANAGEMENT AND PROFITABtLITY 5 8 1operating income increases with firm size {measured by thenatural logarithm of sales), sales growth and fixed financialassets, and decreases with financial debt. In tegtession (2), asignificant negative relation is found betweeti gross operatingincome and number of days inventories (p-value = 0.015).Regression ('i) shows a very significant negative relationbetween gross operating income and number of days accountspayable. This negative relation confirms the negativecorrelation between operating income and number of daysaccoutits payable in Table 2.

The cash cotiveision cycle is included in regression (4). Ihecoefficient ofthe cash conversion cycle variable is negative, bu t itis not significantly cfifferent from zero{/? value= 0.668). That isnot a stu prise: gross operating income declines with the tiutiiberof days accounts receivable and inventories, but also with thenutnber of days accounts payable, which is subtracted tocalctdate the cash conversion cycle.rh e tesults of regressions (1) to (4) suggest that m anagers can

increase corporate profitability by reducing the number of daysaccounts receivable and inventories. An explanation for thenegative relation between accounts payable and gross operatingincome is that less profitable firms wait longer to pay their bills.In regressions (5) to (8), the determinants of gross operatingincome are estimated using plain OLS instead of fixed effectsestimation, and include variability of income, 4 year dummiesand ;i7 itidttstry dtttnmies as independent variables. OLSestimation does not take into account firm specific differencesin profitability. The results are generally consistent with theresults of regressions (1) to (4). Gross operating incomedecreases with niunber of days accounts receivable, inventoriesand accounts payable. One difference between fixed efFectsestimation and OLS estimation is that in regression {8} grossoperating income decreases with the cash conversion cycle:the coefficient is highly significant (p-value = 0.000).In regression (4), it was not significant. It is interesting to


10/16

582

3

.2

ura

ICT)

^ CDO

o2

^

DELOOE

00d

2 0c-j 00 0d d1 ^

mooo

oo mo -^o -^oqq qq qq q

:oooqq

O C-J OO O m oO c: O O e , Od d

d d d d

f O r~d d d

5 . O^7 qd o d

^^ o m r o '' omo --O mo too ic _ c Ci q qdd d d d d d d

O r- TfO Oo o o

X o m X rrO Oq q q od d d d

m or o X Oin o ,

X Oin o

mo tDm o t-o-fo o mo ^ odd dd dd dd

ox^ od o d

IaJ>r

ns

3


11/16

WORKING CAPITAL MANAGEMENT AND PROFITABILITY 583o

o dc id

oo d

CMC-J

d

a;^ ^ ÔO O

d o d

OO O

^ 3i Ji S '5-C XJ -^ ^ ^^ ? n ^^ i ii lo '^ IS : , C I ' -S 23 c I.J i^ . 3 be

j re >^ 3 3 S

- Z = '^m ô -

S I 3. 2IIIII

I > X i?= 1 i E o y

S . = - 2 5 s

^ +

ra0u

o 1

1

i _0

z CV 5 1c0

cC2

' î2u'ac.2Vea


12/16

584 DELOOF3 . DOE S P R O F I l AB IL IT Y AF F E C T W C M, OR I ICK VERSA}

It cannot be rtiled ont that the negative relation between WCMand profitability is to some extent a consequence ofprofitability affecting WCM, and notvice versa. Indeed, the mostplausible explanation for the negative relation between accountspayableand profitabilityisthatlessprofitable firm s wait longertopaytheirbills. negative relation between inventoryand profitability canbe caused by declining sales, leading to lower profits and moreinventory.

An alternative explanation for the negative relation betweenaccountsreceivable and profitability could be that customerswant more time to assess the quality of products they buy fromfirms with declining profitability. However, finance basedmodels explaining trade credit (e.g. Schwartz, 1974) arguethat firms able to obtain funds at a low cost will offer tradecredit to firms facing higher financing costs. Emery (1984) seestrade credit as a more profitable short term investment thanmarketable securities. These models imply that higher profitsshould lead to more accounts receivable, because firms withhigher profits have more cash to lend to customers. This isconfirmed by Deloof and jegers (1996), who find that Belgianfirms with a shortage oi cash reduce investment in accountsreceivable.

Figure 1 shows the median nu m be r of days accoimtsreceivable, accounts payable and inventories, all partitionedby gross operating income deciles. Consistent with thehypothesis that less profitable firms wait longer to pay theirbills, the number of days accounts payable is much higher forthe lowest income deciles than for the other income deciles.The median number of days accounts payable is 62.5 daysfor the first income decile (gross operating income = 0.009)and 65.8 days for the second income decile (gross operatingincome 0.039). For the other income deciles, the mediannumber of days accounts payable ranges from 47.1 days


13/16

WORKINC; CAPITAL MANAGEMENT .ANO PROFITABILITY 585Figure 1

Median No. of Days Accounts Receivable, Accotints Payable andInventories, Partitioned by Gro.ss Operating Income Deciles (1,009Belgian Non-financial Firms, 1992-1996:5 045 Firm-yearObservations)

70656055504540353025

A

-

s.-A..-.. ^ ^ f

1 2 3 4 5 6 7 8 9 10Gross Operating Income - Deciles

Accounts Payable - - Inventories Accounts receivable

4. CONCLUSIONMost firms have a large amount of cash invested in w{)rkingcapital . I t can therefore be expected that the way in whichworking capital is managed will have a significant impact on theprofitabili ty of firms. Shin and Soenen (1998) find a strongnegative relat ion between the cash conversion cycle andcorporate profitahili ty for a large sample of l isted Americanfirms for the 1975-1994 period. In this paper, I find asignif icant negative relat ion between gross operat ing incomeand the iuimber of days accounts receivable, inventories andaccounts payable of Belgian firms.

These results suggest that managers can create value for their


14/16

586 DELOOF

NOTES1 The NACE industrial classification serves the same purpose as the well

known SIC classification, and has been established for industry taxonomywithin the European Union. For details, see Eurostat (1985).2 The 1991 data were needed to calculate some oi the variables.3 All regressions in the paper were also estimated with net operating income(sales minus cashanf ihf regressions with the cash conversion cycle.5 An important feature of the fixed effects model is that it concentrates ondiflerences 'within' firms. Ilie fixed effects model is estimated by (1)computing the means lor each variable h /i lii, (2) subtracting the firmmeans from each variable and {?>]running a regression on tht- transformeddata. Fixed effects estimation explains why the variables differ from theirmeans, hut not why the fnm means diffier from each other.6 For a number of fn nis in the sample, a large proportion of total assets arcfixed financial assets, which might influence the results. I re-estimated all

panel regressions for the firms for which the average fixed financial assetsratio over the 1992-1996 period is below ihe median value (0.034). All OLSregressions were re-estimated for the observations for which the fixedfuiancial assets ratio is below the median value (0.024). The results of allthese regressions (not reported) are very similar to the ones reported in thepaper.

REFERENCESDeloof, M. and M. Jeger (199fi). 'Irade Oedit. Product Quality, and

Intragroup Irade: Some European Evidence . Financial Mana^i-menl,Vol.25, No. 3, pp. 945-(i8.Emery, G.W. (1984), 'A Pure Financial Explanation for Trade Credit ,fournalof Fhianuialand Qiiantilative Analysis Vol. 9, No. :i pp. 2 7 1 - 8 5 .Eurostat 1985), NACE: Gen eral Industrial C lassificatum of Economic AcliviliesivithintlwEuropean Com munities Brussels, ECSC-EEE-EAEC).Fisman, k. and I. Love (2001). 'Trade Credit. Financial intermediaryDevelopment and Industry Clrowtb', Unpublished Manuscript ((ColumbiaUniversity).


15/16

WORKING CAPITAL MANAGEMENT AND PROFITABILITY 5 8 7

Lon g, M.S., LB . Malitz and S.A. Ravid (1993), T r a d e C red it, Q ualityG uara nte es , and Produ ct M arketabil i ty , Fmancial Management, Vol. 22,N o. 4 , pp. 117-27.PiMersen, M.A. and R.G. Rajan (199 7), T ra de C redit: The ori es an d Evid ence ,Rn ien of Fuiantial Studies, Vol. 10, No. 3, p p . 6 ( i l - 9 1 .Schwartz, R.A. (1974), An Kcononiic Model (jf Ttade Cred'il', Joiinuil ofFitiaiina and Q jiaiitildlive Aria/ysi.s, Vol. 9, No. 4, pp . 64. i-57.Shin, H.H. atid L. Soenen (199H), Efficiency of Working Capital andC or po ra te Profitability , Finanrinl Praclire and Education, Vol. 8, No. 2,p p . 3 7 - 4 5 .Svens.son, K. (1997) , T ra de CJedits in Euro pe lo d a y : Credi t C ulture s ,Paym ent Moral ity and Legal Systems , Unp ublishe d Man uscript (Lun dLlniversity). | heun isse H. and M. Je ge rs (1994), Elementen van Boekhouden en Analyse vanJaanrkeningen (Brussels , VU BPress) .


16/16

does working capital management affect profitability of belgian firms

Documents