Infation,operating cycle,and cash holdings

Yncho W ng,Yu Ji,Xu Chen,Chunlei Song

aSchool of Accountancy,The Central University of Finance and Econom ics,China

bGuanghua School of M anagement,Beijing University,China

Infation,operating cycle,and cash holdings

Yanchao W anga,*,Yu Jib,Xu Chena,Chunlei Songa

aSchool of Accountancy,The Central University of Finance and Econom ics,China

bGuanghua School of M anagement,Beijing University,China

A R T IC L E I N F O

Article history:

Received 24 M ay 2012

Accep ted 19 Ju ly 2013

Availab le online 2 January 2014

JEL classif cation:

G 31

G 20

E32

Inf ation

A co rporate cash-ho ld ing strategy is a trade-o f betw een the costs and benef ts of holding cash.A t themacro level,f rms are inclined to adjust and optim ize their cash-ho lding strategies in response to changes in purchasing power due to in f ation.A t them icro level,the operating cycle,which indicates the speed and turnover o f corporate cash f ow,also in fuences thecorporate cash-ho lding strategy.Firm s f exibly adjust their cash-ho lding strategies in response to changes in the internal and external environm en t,w hich is referred to as the cash ad justment strategy.W e exam ine these predicted relationships using a samp leo f listed f rm s in China’sstock marketover the1998–2009 period.Consistentw ith our predictions,the em pirical resu lts indicatea signif cantnegative association between cash holdingsand the CPI,but the relationship is reversed when the CPI reaches a certain level.There is also a U-shaped relationship betw een operating cycle and cash ho ld ings,and this relationship is sim ilarly inf uenced by changes in the in fation level.In exam ining themacroeconom ic environment and m icro level f rm-specif c characteristics simultaneously,our fndingssupp lement the literatureon f rms’cash-ho lding strategiesand p rovide theoretical and p ractical im p lications.

?2013 Production and hosting by Elsevier B.V.on behalf o f China Journalo f A ccounting Research.Founded by Sun Yat-sen University and City University o f Hong Kong.

1.Introduction

A f rm’s cash-ho ld ing strategy determ ines its fate and futu re.In an incom p letem arket,f rm s tend to ho ld and accum ulate cash for fu ture trad ing and to m itigate risk(K eynes,1936).However,ho ld ing too m uch cash can increase opportunity costs,while holding too little can induce shortage costs that disadvantage invest-m ent.The op tim um level of co rpo rate cash ho ld ings is a trade-of between the costs and benef ts o f holding cash(Opler et al.,1999;Harford,1999).Increases in infationary pressure undoubted ly in f uence the amount and cost o fmonetary supp ly,thus af ecting the trade-o f between the costs and benef ts of holding cash.In addition,a f rm’soperating cycle,which is directly related to the time and speed atwhich it gains cash,also in fuences itscash-holding level.By considering both theeconom ic cycleat themacro leveland the f rm-specif c operating cycle at them icrolevel,this study extends our know ledge of the factors af ecting corporate cash holdings under in fation.

Studieson corporate cash holdingsdiscuss theefectso f corporate f nancial conditions,governancesystems, capitalstructures,external fnancing costs,and so forth on the levelo f cash ho ldings.In the fnancialcondition literatu re,fo r exam p le,the co rpo rate cash-holding level is found to have a signif cant relationship w ith f rm s’fnancial features(X in and Xu,2006b);a positive relationship w ith leverage,in form ation asymm etry,corporate age and size,investm en t opportunities and cash fow changes(Fau lkender,2002);and a negative relationship w ith corporatesize and credit-ratings(Ozkan and Ozkanm,2004;Wang,2009).Batesetal.,2008 fnd that riskier cash fow,decreasing inventory holdingsand receivables,and increasing R&D lead to increasing levels of corporate cash ho ldings in the US.In the corporate governance systems and capital structures literature, Kusnadi(2003)fnds that the relationship between cash ho ldingsand board size is positive,whereas the relationship between cash ho ldings and non-management block-ho lder ownership is negative.Zhang and Liu (2005)fnd a negative linear relationship between corporate cash ho ldingsand shareholder p rotection,dependent on ownership structure.Sim ilarly,Xin and Xu(2006)f nd that Chinese listed f rmsw ith better corporate governancemechanisms havemore reasonable levelso f cash holdingsw ith less cash redundancy or shortages. Chen and Chuang(2009)fnd that CEO ownership,risk investm en t,and the num ber of independent directors also have a strong efect on cash policy.Faleye(2004)exam ines the relationship between takeover defenses and cash holdings from a control-rightsmarketperspectiveand f nds that themore cash f rm shold,themorebidders they face in a takeover bid.Studies on the external fnancing environment also report important fndings.For example,f rmsw ith greater grow th opportunitiesand those that f nd itmore risky andmoredif cult to enter the capitalmarket holdmore cash(Op ler etal.,1999).A f rm’s fnancial restraints policy playsa decisive role in its cash ho ldings(Wang,2009).The monopoly power o f banks(Pinkow itz and W illiamson,2001)and past fnancing dif culties(Faulkender,2002)also have important efects on cash ho ldings.Some studies fnd that diversif ed enterprises hold less cash compared w ith those in the same industry(Subramaniam and Tang, 2010).Firms are likely to hold m ore cash to help smooth R&D expenditure but face fnancing frictions (Brown et al.,2010).

In contrast,recen t research on corporate cash ho ld ings focuses on them acrolevelo rm ed ium level,including political,econom ic,legislative,and industry-environm en t facto rs.In a cross-coun try com parison, Pinkow itz et al.(2006)f nds that f rms in countriesw ith low levels o f investor protection and more po litical risk tend to hold more cash.Firms facing fnancing constraints increase their cash ho ldings during econom ic recession periods and raise their liquidity levels during credit crunch periods(Custodio et al.,2005).The increase in tax costs due to dif erences in tax ratesbetween countriesalso contributes to increases in the proportion o f cash ho ldings ofmulti-national corporations(Titman et al.,2004).Some studies exam ine the specif cmacroenvironment in China.In China’sweak institutional environment,f rms that face high f nancing constraints can gain higher yieldsw ith high-level cash-holding po licies(Zhou and Xie,2007).In China’s specif c institutional and governance environm ent,the judgmentsmade by Chinese enterprises and their agents regard ing the costs and benef ts of cash ho ld ingsm ay be dif erent o r distorted,which m ay afect their cashhold ing decisions(Gu and Sun,2009).Cash holdings vary w ith the tightening o fm onetary po licies.During periods ofm onetary con tractions,f rm s that grow rapid ly w ill increase their cash ho ldings to reduce external fnancing constraints and thusm eet the future dem and for investment(Zhu and Lu,2009).Viewed from an industry competition andmarket characteristics perspective,a f rm’s cash-holding level is the equilibrium outcomeo f the corporate fnancing,investment,andmarketenvironm ent.Enterp rises tend to holdmore cash asa precautionary measure to maintain their market share and reduce the risk of being exploited(Haushalter et al.,2007).Competition intensity and life-cycle stage also have signif cant ef ects on cash-holding levels (Zhang and Zhang,2009).H igher levelso f cash holdingsenable enterprises to benef t from market share segmentation and competition in the future,especially in the face of fnancing constraints and ferce competition (Fresard,2010).

In conclusion,although num erous stud ies focus on them icrodeterm inants of cash holdings,few exam ine the operating cycle,and studiesat themacrolevel andm edium levels could be further developed or extended. A lthough Zhu and Lu(2009)exam ine the relationship between monetary po licy tightening and cash-holding levels,few studiesexam ine themutual inf uence ofmacro level and m icrolevel factorson cash ho ldings.Investigating the efectso f inf ation levelson f rms’cash-ho lding po licieswould p rovide direct and useful evidence because it is usually inf ation that induces governments to tighten theirmonetary po licies.

From the perspective of the costsand benef tsof cash ho ldings,in fation af ects the capitalsupp ly cost and value.Increasing inf ation also inducesgovernmentm acroeconom ic contro l,which directly afects the capital supp ly costand value.A f rm’soperating cycle indicates thespeed and scale of its cash fow acquisition,which then inf uences the costsand benef tsof internal cash ho ld ings.A business is inseparab le from itsexternalenvironm ent.In this study,we no t on ly consider the inf uence o f f rm s’operating characteristics on the costs and benef ts of hold ing cash,bu t also exam ine changes in the external environm en t and how f rm sm ake ad justm ents to the tradeof.Hence,this study p rovidesa comprehensive exam ination o f how corporate enterp rises determ ine their cash-holding policies in response to changes in themacroeconom ic environment.The current m acroeconom ic inf ation p rovides usw ith an opportunity to extend the study.

Fig.1 illustrates the large fuctuations in the consumer price index(CPI)between 1998 and 2009.China’s CPI indicates that inf ationary p ressure increased consistently from 2003 to the beginning of 2008.The huge efecto f the2009 globaleconom ic crisison Chinese realeconom ics,togetherw ith the reversion and rebound in thedomestic econom ic cycle,pu lled theCPIdown for the f rst time.This fuctuation prom pted f rm s to ad just or op tim ize their p roduction and operation plans,eitheractively or passively,and then change their cash-ho lding behavio r(Rao and Jiang,2013).Changes in m acroeconom ic policy,especially m onetary policy,due to f uctuations in the CPI also inf uence f rm s’external fnancing costs and capabilities,which in tu rn inf uence their cash-ho lding levels.We exam ine the relationship between operating cycle and cash holdingsby high lighting the interaction between the operating cycle and the“generating blood”function,which becomesmore comp lex when the CPIor inf ation is f uctuating.Therefore,it is important to understand themutual inf uence o f the operating cycle and CPIon corporate cash-ho lding levels.

This study investigates the efect of inf ation levels and operating cycles on corporate cash holdings from both themacro-and them icroperspective in China’s current situation.The empirical results conf rm that corporate cash-holding levels decline as the level of infation increases.Once in fation reachesa certain level,corporate cash-holding levels start to rise;hence,there isa U-shaped relationship between a f rm’soperating cycle and its cash holdings.W e further exam ine the interactive efect of infation and operating cycle on the level of cash.

Them ain contribu tionso f our study are as follow s.W e reveal the efectso f inf ation and operating cycle on corpo rate cash hold ings from bo th them acroperspective of the f nancial environm en t and them icroperspective of f rm characteristics,thusenriching the literature on cash ho ldingsathome and abroad.W ealso present a prelim inary analysis o f the factors that in fuence the relationship between infation,operating cycle,and cash-holding levels to enhance our understanding of corporate cash-ho lding behavior and to provide listed f rmsw ith empirical evidence on how best to adjust their cash holdingsaccording to p rice changes and theiroperating cycle.G iven the cu rrent background of accelerated in fation and increasing CPI in China,this study is particularly timely.It also reveals the efect o fmacrocontrol po liciesonm icrolevel corporate cash holdings and p rovidesa reference for enterprises to optim ize their allocation of resourcesbased on changes in themacrofnancial environment and their own fnancial circumstances.

Figure 1.N ational CPI from 1998 to 2009.

The remainder o f thisarticle isstructured as fo llows.Section 2 p resents the theoreticalanalysisand assumptions.Section 3 describes the research design.Section 4 reports the results and Section 5 p rovides the conclusions.

2.Theoretical analysis and hypothesis development

2.1.Efect of inf ation on corporate cash holdings

In recent years,China’s CPIand in fation have been consistently increasing.Periodso f inf ation inf uence f rms’operating performance,management situation,and need for and supply o f cash;thus,they need to actively or passively change their levels of cash holdings.The level o f corporate cash holdings is a balance between the costs and benef ts(Op ler et al.,1999;Peng and Zhou,2006).As cash is a non-p rof tab le asset, especially during periods of inf ation,cash holdings can cause a loss o f pu rchasing power which increases the cost of ho ld ing cash and,as prices and in terest rates rise,capital costs also rise.As a resu lt,cu rrency fnancing demands higher interest rates to compensate for the loss o f purchasing power,investors require an additional investment rate of return,which determ ines the expected investment return rate.W hen evaluating an investment project,the expected investment return rate isusually used as the discount rate to ad just the amounto f future cash fow to thepresent value.This results in a positivenet present valueo fenterp rise investment p rojects,as the value o f investment opportunities and the potential future investment returns of corporate cash ho ldings are reduced.As a result,the enterprise w ill reduce their cash holdings for investment (Ferreira,2003).In addition,when f rmshold a cash equivalent that can be converted into cash ata lower cost, it is econom ical to raise funds through selling these assets(Shleifer and Vishny,1984).Firms that hold many realizab le assets tend to hold less cash.They w ill transfer assets that can be held at a lower cost in to cash to raise m oney,particularly during periods of infation when p rices are higher and the fair value o f assets is generally above their book value.Therefore,rather than ho lding cash,f rms tend to increase their liquid assets, for instance by stocking up on inventory,to obtain p rice gains,and avoid their monetary assets shrinking. For these reasons,we believe that during periodso f infation,when the cost of corporate cash holdings is high and income is reduced,enterp risesw ill have less demand for cash and may actively reduce their levels of cash holdings.

Inf ation may also afect an enterp rise’s cash supp ly.In timeso f inf ation,f rms need moremoney to purchase the same amounto f raw materialsand other goods,thus taking upmore o f theirworking capital,while they also generate lessmoney through their operating cycles.As p rices continue to rise,f rmsm ay anticipate fu rther inf ation and thusw ill pu rchase raw m aterials in advance to avoid increases in cost o r to serve asexcess reserves.O thersm ay invest in gold o r real estate to avoid the loss o f pu rchasing power and generate excess earnings,which m ay result in f rm s ho ld ing too m uch work ing cap ital and reducing their cash ho ld ings.In the capitalmarket,rising prices can lead to interest rate rises and increasing uncertainty over investment income,so that the residentsor institutionsare less likely to invest in stocksand bonds,thus increasing stock prices and making itmore dif cult to raisemoney in capitalmarkets(Friedman,1977).

However,when inf ation reaches a certain level,the continued deterioration in macroeconom ic regulation will cause the government to exert tightermacroeconom ic control,which in turn changes the external fnancing environm ent.As the level of inf ation continues to rise,the governmentmust adopt tightermonetary policies for macroeconom ic regulation,such as raising the deposit reserve rate,improving the level o f interest rates,and contro lling the scaleo f credit in commercialbanks.Banksw illbecomemore cautiousabout lending and loan cond itionsw ill be stricter,thus creating strong external f nancing constraints thatm ake itm o re diff cu lt fo r f rm s to ob tain loans(Stiglitz and W eiss,1981).External fnancial constraints in the fnancem arket cause the value ofm oney to becom em ore apparent due to the sharp drop in the overall supp ly o f funds and extremely scarce cash resources.Under such conditions,corporate cash holdings inevitably change and f rmstend to ho ld m o re cash.The prim ary benef t o f holding m o re cash is to p rotect against the increasing risk o f bankruptcy.

To sum up,due to the current econom ic situation in China,ongoing inf ationary p ressures have devalued the currency assetsheld by f rms.Under such conditions,f rms signif cantly increase their inventories to m inim ize risk and fulf ll long-term contracts.Consequently,they experience reduced liquidity for resource allocation,production,salesdistribution and operating capacity,togetherw ith a deteriorating businessenvironment and a shortage of cash.M eanwhile,the emergence o f inf ation also causes a tightening o fmacrocontro l policies,especiallym onetary policies,which also af ects the cost and capacity of external f nancing and in turn in fuences the level of corporate cash ho ldings.Based on the above theoretical and practicalanalysis,we propose H ypo thesis 1.

H 1.As in f ation rises,corporate cash holdings decline.However,when inf ation reaches a certain level, corporate cash holdings increase as the level o f in fation increases.

2.2.The ef ect of the operating cycle on cash holdings

The length of a f rm’soperating cycle isafected by industry factors,the f rm’sbusinessmodel,and itsmanagementef ciency.Generally,corporate operating assetsand working capitalneed to be kep t in balance if the f rm is to continue as a going concern.In other words,from the demand perspective,if the p roducts have a shorter operating cycle,capital takes relatively less time to circu late and the f rm m ust continuously invest funds to com p lete the rap id cycle o f pu rchase,production,and sales.Such f rm s have to hold m ore cash to cover ongoing transactions,which is consistentw ith the transactionsm otive fo r ho ld ing cash.From the supp ly perspective,a shorter operating cyclemeans that thep rocessof obtaining inventory,selling and regaining cash is shorter.Internal capital accumulation is then m ore ef cient,the amount of current capital taken up by inventory and accounts receivable declines,and the f rm’s own cash supply increases,w ith a subsequent increase in internal cash ho ldings.

Conversely,in f rms w ith longer operating cycles,it takes longer to purchase raw materials and produce p roducts.From the demand perspective,the period o f continuous investment in production and operation is extended accordingly,which reducesaverage cash holdings.From thesupply perspective,a longer operating cycle meansa slower turnover of cash fow,inventory sales,and recovery o f accounts receivable.M ore cash is frozen in current assets for a longer tim e,which decreases the f rm’s own cash supp ly and reduces its cash holdings.

As the operating cycle increases,a f rm’sown“blood-m aking”capacity becom esweaker,cap ital tu rnover is likely to create prob lem s,the cap ital chain cannotm eet the f rm’s norm al production needs,and operational demandsand good investment opportunitiesmay be lost.Opler et al.(1999)analyze the ef ects of cash fow risk and fnancing capacity on cash ho ldingsand fnd that f rm ssu fering from higher risk and w ith lessaccess to the capitalmarket tend to holdmore cash.If f rms cannot fu lly sp read the risk o f breaks in their cash f ow, theymust ho ld more cash tomeet future transactionsand specu lativedemand.Thus,once the operating cycle reachesa certain point,f rms raise their cash-ho lding levels to prevent risk and to meet uncertain p roduction and managem ent demandsand investmentneeds,which is consistentw ith thep recautionarymotive theory o f cash holdings.

Based on theaboveanalysis,weargue that there isa U-shaped relationship between theoperating cycle and cash hold ings.W ithin a certain range,extending the operating cyclem eans that there isa continuous supp ly o f m oney pouring through the longer cash cycle,and f rm s do no t need to hold large am oun ts of cash for transactions.A t the sam e tim e,extending the operating cycle reduces the ef ciency o fwork ing cap ital,m o rem oney is taken up by inventory and accounts receivab leand the lower speed o f inventory realization leads to a decline in cash holdings.However,at a certain level,w ith the product turnover cycle increasing,especially when the collection rate of accounts receivable falls,f rmsexperience a longer operating cycle and w ith the consequent risk need to hold an increasing amount o f cash to guard against future risk and uncertainty.Therefore,we state the above relationship as fo llows:

H 2.As the operating cycle increases,cash ho ldings decline.However,when the operating cycle reaches a certain level,cash holdings increase as the operating cycle increases.

3.Data and empiricalmethodology

Our sample consistsof allA-share f rms listed on the Shanghaiand Shenzhen stock exchanges from 1998 to 2009.W e select our sample f rms as follows.First,we exclude f nancial f rms.Second,we remove f rmsw ith total assets and operating cycles below zero.Third,we remove f rms with m issing related data.Fourth,we exclude the top and bottom 1%(greater than 99%quantile)of the variablesasextreme values.Asourmodel is based on the dif erencesbetween consecutive years,all of the data are from f rm s that have been listed for more than two years and the overall data range is from 1997 to 2009.Our f nal samp le comprises9165 f rmyear observations.The data on regional CPI are on ly availab le from 1999 to 2009,so we use the national rather than the regional CPI for 1998.A ll of the data were retrieved from the W IND,CSM AR,and CCER databases.

A ccord ing to A lm eida et al.(2004),the actual situation in China and the above theo reticalanalysis,we estimate the fo llow ing equation to test our hypotheses:

Table 1 provides the variab le def nitions.Internal cash f ow(CF)refects corporations’self-blood-making function,and internal cash fow changes also afect cash holdings in the f rm.The size and grow th o f investment opportunities can also change cash holdings.Tobinq isused to control for corporate grow th opportunities.Lnasset is used to control for f rm size,as levels of corporate expansion and fnancing constraintsmay difer among diferent-sized f rms.Financial conditionsmay also inf uence corporate cash-holding behavior, so we use△NWC and△SDebt(working capital and current liabilities)from the previous to the current year to m easu re the efect o f changes in fnancial cond itions on co rpo rate cash ho ldings(Op ler et al.,1999).Firm s m ay ho ld cash to p rotect against future risk,so we choose beta to m easu re Risk(beta is the regression coeff cient o f the daily stock return and m arket retu rn for every f rm in a f scal year).Therefo re,we include the fo llow ing fve contro lvariables:cash fow(CF),size(Lnasset),grow th(Tobinq),beta risk(Risk),and f nancial status(△NWC and△SDebt).Fo llow ing p reviousstudies,weuse fxed year efects to reduce theef ecto fmacroenvironment changes to a certain extent.Price f uctuation is a macro level variab le,so when the model includes CPI,we on ly control for f rm fxed ef ects,but when themodel excludes CPI,f rm and year f xed efectsare controlled for at the same time.W e decompose the operating cycle into the inventory turnover period(Inventi,t)and the accounts receivab lepayback period(Recvi,t)to analyze the efectso f the inventory turnover period and the accounts receivable payback period on the level of cash holdings.

Table 1 Variable defnitions.

4.Results

4.1.Descriptive statistics

Table2 p resents thedescriptivestatistics for themain variables.On average,the change in the cash holdings to total assets ratio for our sample f rms is 1.1%,the standard deviation is8.6%,them inimum is-63.7%and themaximum is69.6%,indicating that thereare largediferencesbetween the cash-ho lding changeso f diferent f rms.The average operating cycle in our sample is 282 days,w ith a standard deviation of 396 days and a range from 14 to 277 days,indicating that the operating cycle variesw idely between f rms.Theaverage invento ry tu rnover period in our sam p le is 203 days,and the average accounts payab le turnover period is 75 days.Compared w ith receivables,inventory accounts for a greater amount o fworking capitaland slower turnover, and the inventory turnover period constitutes the greatest p roportion of the operating cycle.

Table 2 Summary Statistics.

Table 3 CPIand△Cash.

Table 4 Operating cycle and△Cash.

Figu re 2.CPI and△Cash.

Figure 3.Operating cycle and△Cash.

In Tables 3 and 4,we divide our sample into 10 groups according to changes in the CPI and operating cycle,which are def ned from 0(low)to 9(high).Table 3 and Fig.2 reveal a strong negative relationship between the cash rate of public f rms and the CPI.W hen inf ation falls,the cash-ho lding ratio increases, and when inf ation rises,the cash-hold ing ratio decreases.Specif cally,as the in fation level d rops toward its lowest point,the change in cash ho ld ings is greater.From Tab le 4 and Fig.3,we can also see a close relationship between the cash-ho ld ing rate and the operating cycle.W hen the operating cycle is longer,the cashholding rate is lower,indicating a negative relationship,butwhen the operating cycle is very long,the ratio o f cash holdings begins to increase.

4.2.M ain results

The OLS regression resu lts are shown in co lumns 4–6 of Table 5.To control the problem o f biased estimates due to potentially important om itted variables,we construct non-balanced panel data to enhance the reliability.The panel regression results are shown in colum ns 1–3 of Table 5.The OLS resu lts show that if we only take CPI and CPI2 in to accoun t and do not consider the in fuence o f Cycle,the regression coef cien t of CPI is-0.1227,which is signif cant at the 0.001 level.The regression coef cient of CPI2 is 0.0006,which is signif cant at the 0.001 level.These results ind icate a U-shaped rather than a sim p le linear relationship betweenin fation and changes in the cash holdingso f listed f rms.The panel resultsare sim ilar to theOLS results.Ifwe only take CPI and CPI2 into account and do not consider the in fuence of Cycle,the regression coef cient o f CPI is signif cantly negative and the regression coef cientof CPI2 is signif cantly positive.Even when we take Cycle and Cycle2 into consideration,the regression coef cients for CPI and CPI2 remain at a sim ilar level. This indicates that as the inf ation level increases,f rms reduce their levels of cash ho ldings for the purposes o f trading and speculative dynam ic opportunities.However,when themacroenvironment in fation deteriorates to a certain extent,the government’smacrocontro l lim its the size of funds in the capitalm arkets,which creates f nancial constraints.Theextremescarcity of cashmeans itbecomesan objecto f pursuit in themarket, and listed f rm s increase their levels of cash ho ld ings.This fnding suppo rts H ypo thesis 1.

Table 5 Study of the U-shaped relationship between changes in the in fation rate and changes in cash holdings.

The OLS regression resu lts in colum n 5 show thatw ithout considering CPI and CPI2,the regression coeff cient of Cycle is-0.0125,which is signif cant at the 1%level and the regression coef cien t o f Cycle2 is signif cantly positive at the 10%level.Compared w ith the OLS regression,the regression resu lts for the panel data in co lumn 2 aremuch stronger.The regression coef cients also remain steady when we take CPI and CPI2 into consideration.This indicates that the efect o f operating cycle on cash ho ldings isnot a simple negative linear relationship.The cash ho ldingsof listed f rms decrease as theoperating cycle increases,but once it exceedsa criticalvalue,listed f rms increase their cash holdingsasa p recautionarymeasure,in response to the gradually increased operational risk.Them athematicalm eaning of the regression resu lts is that in a coordinate system w ith Cycle along the x axisand△Cash along the y axis,Cycle isa concave function.The regression based on thepaneldata calculates the range of the in fection point distribution around 1.6164,w ithmost o f the data distributed on the left of the axis of symm etry,and few data points on the righ t(see Fig.4).These results validate H ypo thesis 2.

Figure 4.U-shaped curve o f the relationship between operating cycle and cash ho ldings.

4.3.Further analysis

4.3.1.Inventory turnover period,accounts receivable turnover period and cash holdings

Table 6 further tests the ef ect of listed f rms’cash ho ldings on the inventory turnover period and the accounts receivab le turnover period.From the regression results in colum ns 1 and 4,we can see that Invent has a negative coef cient,signif cant below the 1%level.Invent2 is not signif cant,indicating that the cash holdings of listed f rmsgradually reduce as the inventory turnover period increases and that there is a linear relationship between the inven tory turnover period and cash holdings.The regression results in co lum ns2 and 5 indicate that Recv has a larger coef cient than Invent and is signif cantly negative at the 1%level,and Recv2 hasa positive coef cient and is signif can t below the 10%level.The resu lts rem ain robust regard lesso fwhether we use the CPI or itssquared term.As the accounts receivab le turnover period increases,listed f rmsgradually reduce their cash ho ldings,but start to increase them once the accounts receivable turnover period increases abovea critical threshold.From the combined regression resu lts in Tables5 and 6,we can in fer that due to the relationship between the levelof cash ho ldingsand theaccounts receivab le turnover period,f rm sw ill increase their cash holdingswhen the operating cycle increasesbeyond a certain level,perhaps because they aremore sensitive to the extension o f the receivables payback period.Inventory back logs and slow-m oving and defective risks can be dispersed by sales channels,p rice discounts,and re-working.However,receivables’bad debt losseshave an immediate efect on the normal turnover of funds.W hen the accounts receivab le turnover period reaches the critical level,acco rding to the p recau tionary m otive,listed f rm s w ill pay m o re attention to in ternal capital accum ulation and increase their cash ho ld ings to pro tect them selves against fu ture uncertainties and the risk of loss.

4.3.2.Investigation of the interaction ef ect between infation and the operating cycle

Inf ation not on ly has a direct efect on f rms’cash ho ldings by afecting the demand and supp ly for cash, but also an indirect af ect via the operating cycle.Changes in the level of inf ation,inventory turnover,and accounts receivab le turnover have diferent efectson cash holdings.The level o f inf ation has diferent ef ects on cash ho ldings as the operating cycle(inventory turnover,accounts receivab le recovery period)extendsor shortens.Hence,we conduct further analysis to exam ine the interactive efect of CPI and Cycle on cash holdings.

Taking the level of in fation as the standard,we divide thewhole sample into three groups according to CPI,w ith group 0 represen ting low CPIand group 2 representing high CPI.W e useM odel(1)to test the ef ect of operating cycle on the level of cash ho ldingsunder diferent levelso f in fation and the resultsare p resented in Table 7.

Table 6 Infation,inventory tu rnover period,receivab le accounts turnover period,and cash ho lding changes(panel results).

For group 0,when in fation is at a low level,the coef cient of Cycle isnot signif cant under a linear specif cation.However,the coef cient of Cycle is signif cantly positive and the coef ciento f Cycle2 is signif cantly negative in the quadratic specif cation.For group 1,when in fation is at a medium level,the coef cient o f Cycle is not signif cant.W hen inf ation is high(G roup 2),the coef cient of Cycle is signif cantly negative and the coef cient o f Cycle2 is signif cantly positive.

The regression resu lts indicate thatwhen the inf ation level is low,there isa signif cant positive association between the operating cycle and cash ho ld ings up to a certain point,afterwhich the association turnsnegative. W hen the in fation level ishigh,there is a signif can t negative association between the operating cycle and cash holdings up to a certain point,afterwhich the association tu rns positive.This d if erencem ay occu r because in periodsof low inf ation,enterp rises faceamore relaxed fnancing environment and fewer fnancial constraints, and thusare able to obtain low-costexternal funds.M eanwhile,f rmsw ith longer operating cycles realize that the value of their funds is reduced;hence,they increase their cash holdingsasa p recaution.As inf ation rises, the excess liquidity in the country usually causesmoney supp lies to tighten.The governmentm ay exercise m acrocontrol by raising the deposit-reserve ratio to reduce themoney supp ly.The size of the banks’credit w ill shrink,so the banksbecom e particu larly cautious in granting loans and setmore demanding conditions, which causes strong external fnancing constraints.A t the same time,for enterp rises w ith long operating cycles,the usage ef ciency of working capital is relatively low.An insuf cient supply of operating cash fowleads to greater operational risks,m aking it harder for them to obtain bank loans and thus increasing their external f nancing constraints.In addition,increases in raw material prices due to inf ation demand more working capital,which makes internal funds generally tight and causesm ore serious delinquency,which extends theoperating cycle.In otherwords,theef ecto f inf ation on businessenterp rises leads to further deterioration in their ability to generate operating cash and external fnancing capacity.It becomes increasingly dif cult for them to maintain a certain level o f cash holdings.Therefore,when levels o f in fation are high, the cash ho ld ings o f f rm s w ith longer operating cycles w ill be signif cantly reduced.

4.4.Robustness tests

We rep lace operating cyclewith the cash turnover period to evaluate f rms’operating capacity.Verlyn and Laughlin(1980)take the cash cycle asameasure ofworking capitalmanagement because it can refect capital operating ef ciency more com prehensively than the turnover rates of current assets,such as the inventory turnover and accounts receivab le turnover rates.W ang et al.(2007)holds that the cash turnover period associatesworking capitalmanagementw ith supp ly chainmanagem ent and avoids the contradiction derived from exam ining the elements in isolation w ithout considering the relationship between them(for examp le,the easing of credit po licy shortens the inventory turnover period and increases the accounts receivab le,leading to a decrease in the accounts receivable turnover rate).In addition,diferent from the operating cycle,the cash turnover period takes into account both current liabilities in working capital and the time factor in cash in fows and outf ows,so that we can identify the ef ects o f the fow o f inventory,accounts receivable and accounts payab le on working capital as a whole.Therefore,we conduct this robustness test by substituting operating cycle with cash turnover period and achieve the sam e fndings(see Fig.5).

Figure 5.Cash turnover period=inventory turnover period+accounts receivable payback period-accounts payab le deferred period.

5.Concluding remarks

A t them acro level,con tinuously rising p rices and in fation,which are the focus o f statem acrocon trol policies,af ect f rm s’p roduction decision m ak ing and cash ho ld ings.U sing the consum er price index,this study verifes and exp lains the fnding that co rpo rate cash ho ld ings decrease as in fation increases from the perspective o f the loss in purchasing power o fm onetary assets.However,when inf ation reachesa certain level,f rms increase their cash holdings to guard against bank rup tcy.A t them icrolevel,we f nd that corporate cash ho ldings decreasew ith longer operating cycles,butwhen inf ation reachesa certain level,f rms increase their cash holdings to m itigate risk.Hence,there isa U-shaped relationship between operating cycle and cash holdings. This study also investigates the interactive inf uence o f infation and operating cycle on the amount o f cash holding,thusexpanding our know ledge o f this area.

Acknow ledgements

This study is suppo rted by grants from the pro ject No.71102123,N ational Natural Science Foundation o f China,and also from the“Project 211”Fund o f the Cen tral University o f Finance and Econom ics,China.

Almeida,H.,Cam pello,M.,Weisbach,M.S.,2004.The cash fow sensitivity of cash.Journal of Finance 59,1777–1804.

Bates,T.W.,Kahle,K.M.,Stulz,R.M.,2008.W hy do U.S.f rm s hold so m uch mo re cash than they used to?W orking Paper.

Brown,R.,Petersen,J.,B ruce,C.,2010.Cash ho ldings and R&D sm oothing.Journal of Corporate Finance 17,694–709.

Chen,Y.R.,Chuang,W.T.,2009.A lignment or entrenchment?Corporate governance and cash holdings in grow ing f rm s.Journal of Business Research 62,1200–1206.

Custodio,C.,Ferreira,M.A.,Raposo,C.,2005.Cash holdings and business conditions.ISCTE Business School,Working Paper.

Faleye,O.,2004.Cash and corporate con tro l.Jou rnal of Finance 59,2041–2060.

Faulkender,M.W.,2002.Cash holdings among smallbusiness.Washington University,St.Louis,SSRN,Working Paper.

Ferreira,V.,2003.Why do f rm s hold Cash?Evidence from EM U countries.ISCTE School of Business,SSRN,Working Paper.

Fresard,L.,2010.Financial strength and productmarket behavior:the real efects of corporate cash holdings.Journal o f Finance 65, 1097–1122.

Friedman,M.,1977.Nobel lecture:infation and unem ployment.Journalo f Econom ic Dynamics and Control 12,231–254.

Gu,N.K.,Sun,J.J.,2009.Financing constraints,cash f ow risk and cash ho ldings for precautionary motive.Business Economics and M anagem ent 4,73–81.

Harfo rd,J.,1999.Corporate cash reserves and acquisitions.Journal of Finance 54,1969–1998.

Haushalter,D.,K lasa,S.,M axwell,W.,2007.The infuence of productmarket dynamicson a f rm’s cash holdingsand hedging behavior. Journal of Financial Econom ics 84,797–825.

Keynes,J.M.,1936.The General Theory of Em p loyment,Interest and M oney.M cM illan,London.

Kusnadi,Y.,2003.Corporate cash hold ings and corporate governancem echanism s.H ong Kong Un iversity of Science and Technology, SSRN,W orking Paper.

Op ler,T.,Pinkow itz,L.,Stulz,R.,W illiamson,R.,1999.The determinants and im plications of corporate cash holdings.Journal of Financial Econom ics52,3–46.

Ozkan,A.,Ozkanm,N.,2004.Corporate cash ho ldings:an em pirical investigation of UK com pan ies.Jou rnalof Bank ing and Finance28, 2103–2134.

Peng,T.Y.,Zhou,W.,2006.M otivation for research on China listed com panies’high cash holdings–p rincipal-agent theory or balance theory?Accounting Research 5,42–49.

Pinkow itz,L.,Stulz,R.,W illiam son,R.,2006.D oes the contribution of corporate cash holdings and dividends to f rm value depend on governance?Journal of Finance 61(6),2725–2751.

Pinkow itz,L.,W illiam son,R.,2001.Bank power and cash holdings,evidence from Japan.Review of F inancial Stud ies 14,1059–1082.

Rao,P.G.,Jiang,G.H.,2013.The impact ofm onetary policy on the relationship between bank loans and business credits.Econom ic Research 1,68–82.

Shleifer,A.,Vishny,R.W.,1984.Liquidation values and debtmonitoring.Review of Economic Studies 51,393–414.

Stiglitz,J.E.,W eiss,A.,1981.Credit ration ing in m arkets w ith im perfect information.Am erican Econom ic Review 71,393–410.

Subram aniam,V.,Tang,T.T.,2010.Firm structure and corporate cash holdings.Jou rnal o f Co rporate Finance 17,759–773.

Titman,S.K.C.,W ei,J.,Xie,F.X.,2004.Capital investmentsand stock returns.Journal of Finance and Quantitative Analysis39,677–700.

Verlyn,R.D.,Laugh lin,E.J.,1980.A cash conversion cycle app roach to liqu idity analysis.Financial M anagemen t 9(1),32–38.

W ang,Y.C.,2009.Financial constrain ts,cash holdings and over-investm ent.Journal of Financial Research 7,121–133.

Wang,Z.Q.,Pang,Y.M.,Sun,J.Q.,2007.A review and p rospect of theworking capitalmanagement.Accounting Research 2,85–92.

Xin,Y.,Xu,L.P.,2006a.Corporate governance and excess cash ho ldings.M anagementW orld 5,136–141.

Xin,Y.,Xu,L.P.,2006b.The determ inants of cash hold ings in ch ina’s listed companies:fnancial characteristics,ownership structureand governance environm ent.China A ccounting Review 2,307–320.

Zhang,R.J.,Liu,C.J.,2005.Ownership structure,shareholder protection and listed com panies’cash holdings.Finance and Trade Econom ics2,3–9.

Zhang,J.G.,Zhang,J.R.,2009.Em pirical research on cash holdings in China listed companies.Collected Essays on Finance and Econom ics 2,74–81.

Zhou,W.,Xie,S.L.,2007.China’s listed com pany high cash holdings.World Economics 3,67–74.

Zhu,J.G.,Lu,Z.F.,2009.M onetary policies,enterprise’grow th,and the change in the levelof cash-holding.M anagementW orld 3,152–158.

*Corresponding author.

E-mail address:aeroant@126.com(Y.W ang).

http://dx.doi.org/10.1016/j.cjar.2013.07.001

1755-3091/?2013 Production and hosting by Elsevier B.V.on behalf of China Journalof Accounting Research.Founded by Sun Yat-sen U niversity and City University of H ong K ong.

Operating cycle

Cash holdings