interdependent analysis of leverage, …...will be tested using two-stage least squares. five...

179

Hardjopranoto—Interdependent Analysis of Leverage, Dividend, and Managerial Ownership Policies

Gadjah Mada International Journal of BusinessMay-August 2006, Vol. 8, No. 2, pp. 179–199

INTERDEPENDENT ANALYSIS OFLEVERAGE, DIVIDEND, AND MANAGERIAL

OWNERSHIP POLICIESAgencies Perspectives

Wibisono Hardjopranoto*

This paper attempts to investigate interdependent mechanismamong leverage, dividend, and managerial ownership policies. Thispaper considers firm size and economic conditions to control theireffect on the relationship among the three policies. The interrelation-ship between leverage, dividend, and managerial ownership policieswill be tested using two-stage least squares. Five exogenous vari-ables are employed in simultaneous equation: current assets andstructure of assets as leverage determinants, book to market andreturn on investment as dividend determinants, and relative return torisk as managerial ownership determinant. The research employsyear 1994-2004 data, with 1717 firm years. The research findingscan be summarised as follows. First, there is a negative relationshipbetween managerial ownership and leverage policies as suggestedby agency theory. Second, there is a relationship between managerialownership and dividend policies, but the relationship betweenleverage and dividend is insignificant. Third, the relationship be-tween leverage and dividend is insensitive to economic conditionand firm size. Fourth, all exogenous variables have significant effecton endogenous variables, except relative return. Fifth, the effects ofexogenous variables are not sensitive to control variables. Sixth, wefind that managers show self-interest behaviours by reducing mana-gerial ownership when the economic condition worsens.

* The author would like to thank Professor Eduardus Tandelilin for his invaluable comments,insights, and financial data sharing.

Keywords: dividend; interdependent mechanism, leverage, managerial ownership

180

Gadjah Mada International Journal of Business, May-August 2006, Vol. 8, No. 2

Background

Jensen and Meckling— (1976)define a company as a locus of contractor a number of contracts between anagent and a principal caused by theseparation of ownership and control.Contract system tends to create moralhazard. The moral hazard occurs whenthe agent is able to undertake actionsthat cannot be controlled by the princi-pal. An increase in leverage or debt isone of the alternatives to shift themonitoring costs from the principal tobondholders. The use of debt alsopushes managers to be more disciplineso as to shun bankruptcy (Harris andRaviv 1990). Debt also functions as abonding means to management(Megginson 1997: 335).

This study discusses the interde-pendence of dividend, leverage, andmanagerial ownership policies.Agency cost of equity is difficult toobserve directly; hence, variables ofdividend and leverage policies thatconsider the separation between mana-gerial interests and shareholders’ in-terests are used. The relationship be-tween leverage and dividend is math-ematically examined by DeAngelo andMasulis (1980), who find that lever-age and dividend are relevant if thereare taxes and not in equilibrium condi-tion. Koch and Shenoy (1999) provethat there is interdependence betweenleverage policy and dividend policythat simultaneously and significantlyinfluence future cash flows. Hartono(2000) finds that dividend policy is amechanism to reduce debt policy.

The interdependence amongst thethree policies (debt, dividend, and in-sider ownership) is researched byJensen et al. (1992), finding that lever-age policy and dividend policy do notinfluence insider ownership, whilstinsider ownership influences financ-ing and dividend decisions. Crutchleyand Hansen (1989) harnesses agencytheory and variables that influence le-verage and dividend using three-stagereduced-form regression model. Theyfind that managers utilize three poli-cies (ownership, leverage, and divi-dend) to decrease the agency cost.This finding supports the agency theorythat agency cost of equity and agencycost of debt can be managed and con-trolled optimally using the interdepen-dence of leverage, dividend, and in-sider ownership policies. They alsopoint out that managerial ownership isinfluenced by relative return.

Myers (1977) reveals that compa-nies with high assets in place are in-clined to use higher leverage as theassets in place are basically a sunkcost. Several studies show that assetsin place, proxied by fixed assets, havea positive impact on leverage (Ferriand Jones 1979; Marsh 1982; Longand Malitz 1985; Friend and Lang1988; and Jensen et al. 1992). Collat-eral hypothesis argues that fixed assetscan function as a bailout, showing thata company has enough resources tofulfil its liabilities. Through currentassets, collateral hypothesis is sup-ported by perquisites hypothesis. Per-quisites hypothesis proposed in this

181


research argues that current assets canbe utilized by management to exertperquisite actions. Shareholders incur-ring loss due to these perquisite ac-tions will exert bonding mechanismthrough leverage policy.

Free cash flow hypothesis has arelation to investment opportunities;the higher the investment opportuni-ties, the higher is the internal cashflows to be used to finance or to realizethe investment opportunities (Myersand Majluf 1984; Jensen 1986). Re-maining profits used to pay dividendsare smaller such that growing compa-nies will pay low dividends (Jensen etal. 1992). This argument is contradic-tory with signalling hypothesis thatargues that declining dividend reflectsmanagement’s pessimism towards thecompany’s prospect such that it is anegative signal. An increase in divi-dend is a positive signal since it showsmanagerial confidence on thecompany’s prospect.

Difference in company size caninfluence the relationship amongst thethree policies and variables. Some re-search shows that company size canfunction as a control variable to clarifythe relationship amongst variablestested (Murphy 1985; Smith and Watts1992). Difference in economic condi-tion can also influence the managerialbehavior in making decisions. The re-lationship amongst financial decisionvariables and their determinants canalso change when the economic condi-tion changes substantially. This studycontrols the company size effect and

the change in economic condition bycutting off year 1997 as the beginningperiod of financial crisis in Indonesia.

Predicated upon the discussionabove, the research questions can beformulated as follows:

1. Is there a relationship amongst le-verage, dividend, and managerialownership policies?

2. Do current assets and fixed assetspositively influence leverage?

3. Do investment opportunities andprofitability influence dividendpolicy?

4. Does relative return influence mana-gerial ownership?

General description on the agencyproblem in determining leverage, divi-dend, and managerial ownership poli-cies is still an interesting issue to re-search. Some studies have yet to showresults fully supporting the predictionof agency theory. The issue of agencyconflict is valuable and worth to ana-lyst as La Porta et al. (2000a) arguethat ownership structure has suitablecontrol over agency conflict. Owner-ship structure has a significant impacton reducing agency conflict especiallyin developing countries (e.g., Indone-sia). Indonesian capital market is re-nown as a market with concentratedownership structure. Baker andWurgler (2004) test the effectivenessof dividend to control agency conflict.They find that dividend has a signifi-cant impact on reducing agency con-flict between agent and principal. In-donesia experiences low dividend pay-ment for listed firms (Mahadwartha

182


2004). The phenomenon suggests thatthe agency conflict is lower in coun-tries with high dividend payment, asBaker and Wurgler (2004) examineusing United States data.

This research’s finding is expectedto contribute to empirical and policyaspects. Empirical contribution of thisstudy is expected to support the agencytheory, especially in explaining lever-age, dividend, and managerial owner-ship policies to reduce agency con-flict. The finding is also expected toprovide additional explanations on theagency theory since it involves twopanel data in the periods of normaleconomy and of economic crisis. Sen-sitivity testing model on the change ineconomic condition will elaborate onwhether there is a difference in mana-gerial behavior reflected by leverage,dividend, and managerial ownershippolicies in the different economic con-ditions.

Meanwhile, the policy contribu-tion of this study is intertwined with itsprovision of literature and practicebackground and reference to decisionmakers, both companies and inves-tors. It is expected that managers andinvestors understand the explanationof the agency relationship. Sharehold-ers policy with respect to stock invest-ment in firms should concern theagency conflict between managers andshareholders. Shareholders will jus-tify that financial policies such as divi-dend and leverage are better controlmechanisms for the agency conflict.

Managers offering ownership compen-sation scheme will consider theirwealth more correlated with firmwealth and accordingly will be con-cerned with the firm value. Their inter-ests should align with shareholders’interests so as to reduce the agencyconflict.

Literatures Review andHypotheses Development

The interdependence amongst le-verage, dividend, and managerial own-ership policies in agency theory per-spective is included in behavioral re-search since it explains the effect ofmanagerial behaviors and sharehold-ers’ behaviors on corporate financialdecisions. The research is directed to-wards positivist agency theory as itexamines the agency relationship be-tween management or agent and share-holders or principal (Eisenhardt 1989).

Interdependence of LeveragePolicy and ManagerialOwnership

Managerial ownership policy isincluded into remuneration policy pur-porting to reduce the agency conflictbetween management and sharehold-ers. Murphy (1985); Brickley et al.(1988); Jensen and Murphy (1990);and Smith and Watts (1992) explainhow fixed compensation package (sal-ary) and contingent package (bonusesand options) can be utilized as a meansto align the managerial interests and

183


the shareholders’ interests.1 Megginson(1997: 335) uses the term “bondingmechanism” that tries to align the in-terests of shareholders and those ofmanagement through programs thattie the personal wealth into the wealthof the company.

Management has lower risk aver-sion level than shareholders have sincemanagement is faced with high risk oflosing jobs if the company finds diffi-culties in its operations. Meanwhile,shareholders can diversify their port-folios to reduce risk (Megginson 1997:19). If the company risk increases, forinstance the risk of financial distress,management may lose their jobs whilstshareholders who are well diversifiedwill be faced with lower risk of finan-cial distress.

This phenomenon leads to man-agement more focusing on total risk ofthe company whilst shareholders aremore concerned with market risk orbeta since the firm specific risk hasdisappeared due to diversification. Thisrisk phenomenon then possibly resultsin ambiguous research findings on re-muneration (managerial ownership) inrelation to leverage. Leland and Pyle(1977) and Kim and Sorenson (1986)find that the relationship between le-verage and managerial ownership ispositive. On the other hand, the re-search of Friend and Lang (1988), andJensen et al. (1992) find a negativerelationship between leverage andmanagerial ownership.

Leland and Pyle (1977) use the“signalling” argument that the lever-age of company that has managerialownership is a signal that the leverageis truly utilized for investment andincreasing value of the company. Kimand Sorensen (1986) use demand-sup-ply hypothesis, arguing that a com-pany will increase leverage since in-siders (management) are able to con-trol the company more effectively bymanagerial ownership program. Theprogram shifts the attention of man-agement to risk towards market riskalthough their main concern is still ontotal risk. It purports to align the per-spective on risk of management and ofshareholders. Hence, a company hav-ing managerial ownership will employhigher leverage such that the relation-ship between managerial ownershipand leverage is positive.

The studies of Leland and Pyle(1977), and Kim and Sorensen (1986)can be countered by the argument thatmanagement has possibly diversifiedtheir personal portfolios prior to de-ciding to buy the stocks of the com-pany where they work. This argumentconsiders that like investors and share-holders, management are also indi-viduals able to invest their wealth. Theargument is supported by Crutchleyand Hansen (1989) finding that man-agers will proceed to own the com-pany stocks if the stocks are approach-ing the efficient market line such thatthe diversification loss due to buying

1 Jensen and Murphy (1990) uses the term “incentives” for fixed package and “insurance” forcontingent compensation. Included in option-related package is managerial ownership program.

184


their own company’s stocks declines.If managers pay high attention to thediversification loss, it means that themanagers have diversified their per-sonal wealth.

Other empirical results contradict-ing with Leland and Pyle (1977) andKim and Sorensen (1986) are Friendand Lang (1988), Jensen et al. (1992),Mahadwartha and Hartono (2002),Ismiyanti and Hanafi (2004), andMahadwartha and Ismiyanti (2006)who find that leverage policy is nega-tively influenced by ownership struc-ture. Increasing managerial ownershipwill lead to a stronger tie betweenmanagement personal wealth and thecompany wealth; consequently, themanagement will strive for reducingthe risks of losing job and personalwealth by decreasing debt which alsomeans decreasing the company’s fi-nancial risk.

The prevalence of financial dis-tress arouses agency conflict throughasset substitution and underinvestment(Copeland and Weston 1992: 332),meaning that managerial ownership isrelated to bankruptcy risk triggered byleverage. If the leverage is low, thecompany will increasingly hinge onequity financing, either by right issueor by managerial ownership, such thatleverage negatively influences mana-gerial ownership.

H1: The relationship of managerial

ownership to leverage is negative.

Interdependence of DividendPolicy and ManagerialOwnership

Dividend payment will preventmanagement from undertaking perqui-site actions since the internal cash flowswill be utilized to deliver dividends tostockholders (bonding mechanism).Companies that have a control mecha-nism and broad ownership structureare usually big companies that tend topay dividends to alleviate the agencyconflict between management andstockholders. In contrast, small com-panies with centered ownership struc-ture on several individuals will pay outlow dividends as the potential agencyconflict is also low (Megginson 1997:375).2 Hence, size variable is of impor-tance in controlling the impact of divi-dend on managerial ownership, andvice versa.

In relation to H1, if the managerial

ownership increases, it will lower le-verage so that in order to finance in-vestments, management will be de-pendent on retained earnings whichalso means reducing dividend. Jensenet al. (1992); Rozeff (1982);Mahadwartha and Hartono (2002), andIsmiyanti and Hanafi (2004) find thatdividend policy is influenced by own-ership structure in a negative relation-ship.

Dividend payment will decreaseasymmetric information in a company.However, if the company has manage-

2 The explanation on company size as a control variable is discussed in the research methodsection.

185


rial ownership program, the asymmet-ric information will automatically de-cline. This condition renders dividendpayment irrelevant to be harnessed toreduce the asymmetric information(Megginson 1997: 373).3

H2: Relationship of managerial own-

ership to dividend is negative.

The existence of managerial own-ership leads signalling hypothesis tobe less adequate to elaborate on theinfluence of dividend policy on thecompany value. Agency theory will bemore robust to explain the phenom-enon of dividend policy as a means toreduce asymmetric information thatcauses agency conflict between man-agement and stockholders.

Interdependence betweenLeverage Policy and DividendPolicy

The explanation of agency theoryin the relationship between dividendand leverage, and the relationship be-tween leverage and dividend, is pro-vided through free cash flow hypoth-esis (contracting model of dividend)and balancing model of agency cost.Free cash flow hypothesis (contract-ing model of dividend) predicts thatdividend influences leverage in a posi-tive direction (Megginson 1997: 362).A company delivering dividends inbig amount will need additional fundsthrough leverage to finance its invest-ments such that the dividend policyinfluences the leverage policy in one

direction (Emery and Finnerty 1997:568). The company’s internal cash isutilized to pay dividends so that itneeds additional external funds throughdebt (free cash flow hypothesis).

The argument of Emery andFinnerty is substantiated by Miller andRock (1985) in their conceptual ar-ticle, revealing that high dividend pay-out is a signal of the company’s in-creasing profitability in the future.Management gives the positive signalthrough the dividend payment suchthat investors recognize the future in-vestment opportunities promising forthe company value. Furthermore, highdividend payout means that in order tomaintain its optimum capital struc-ture, the company will employ higherdebt to finance its investments (Emeryand Finnerty 1997: 568; Easterbrook1984).

Easterbrook (1984) also arguesthat stockholders will exert monitor-ing actions towards management; nev-ertheless, if the monitoring costs arehigh, they will harness the third party(debtholders or bondholders) for help-ing undertake the monitoring actions.Debtholders that have invested theirmoney in the company by themselveswill strive for undertaking the controlactivities on the use of their funds.Monitoring mechanism is usuallyimplemented through debt covenantmechanism.

Hartono (2000), Mahadwarthaand Hartono (2002), and Ismiyanti and

3 By itself contradicting with signalling hypothesis.

186


Hanafi (2004) find that dividend policyinfluences a company’s leverage policyin a positive relationship. However,Hartono (2000) also finds that the le-verage policy has insignificant impacton the dividend policy. The dividendpolicy, based on Hartono’s (2000) find-ing, is influenced by the company’saccounting decision. This empiricalevidence shows that the relationshipof dividend with leverage is positive.

H3: The relationship of dividend with

leverage is positive.

Balancing model of agency costexplains the influence of leverage ondividend policy. This model of agencytheory elaborates on the trade-off be-tween agency cost of equity and agencycost of debt (Megginson 1997: 338).Leverage policy influences dividendpolicy in a negative relationship. Acompany with high leverage (highagency cost of debt) will try to allevi-ate its agency cost of debt by decreas-ing the use of debt, thus financing itsinvestments by internal cash flows.Stockholders will give up the internalcash flows, previously paid out as divi-dends, to finance investments such thatthe agency cost of equity will increasecorresponding with the decrease in theagency cost of debt. The relationshipbetween leverage policy and dividendpolicy based on the balancing modelof agency cost is negative.

Balancing model of agency cost,supported by the finding of Jensen etal. (1992), Mahadwartha 2004, andIsmiyanti and Hanafi (2004), positsthat management will be faced with atrade-off between dividend payment

and fixed charge of leverage; accord-ingly, the higher the leverage, the lowerthe dividends will be.

Determinants of Leverage

Current Assets

Companies possessing high de-gree of current assets (CA) will usehigher leverage. It is argued that thecurrent assets are the companies’ bail-outs that the companies will be able tofulfil their leverage. This collateralhypothesis is linked to the possibilitythat management harnesses the cur-rent assets for self-interests, especiallycash and inventories. Agency theorypostulates that the use of leverage maydecline managerial perquisite capabil-ity such that the relationship of currentassets with leverage is positive.

Some research such as that of Ferriand Jones (1979), Marsh (1982), Longand Malitz (1985), Friend and Lang(1988), and Jensen et al. (1992) findthat current assets have a positive rela-tionship with leverage. The currentasset variable is also utilized to exam-ine the perquisite hypothesis that le-verage policy can be harnessed to pre-vent and suppress the managerial per-quisite behaviors. Leverage variableused has separated managerial owner-ship from nonmanagerial ownership(outside shareholders ownership).

H4: The relationship of current assets

with leverage is positive.

Asset Structure. Myers (1977) pointsout those assets in place are betterfinanced by leverage as the assets inplace have a characteristic of sunk

187


costs. The assets in place are reflectedby property, plant, and equipment (athistorical costs) or net fixed assets.The higher the proportion of assets inplace in a company’s total assets, thehigher will be the tendency of usingleverage. This argument is supportedby Skinner (1993). Barton et al. (1989)also add inventories into the assets inplace. Their empirical finding showsthat asset structure measured by con-sidering inventories has a weak posi-tive effect on leverage. The assets inplace also function as collateral suchthat the higher the value of collateral,the higher the company’s capability ofacquiring debt so as to increase lever-age.

H5: The relationship of asset structure

with leverage is positive.

Determinants of Dividend

Investment Opportunities

Future investments will highlyinfluence the value of company. Myers(1977) reveals that the company valueis derived from a combination of fu-ture assets and future investments. Highinvestment opportunity set (IOS) leadsto high growth of the company. Everycompany has different IOS, depend-ing on specific assets owned (Kester1986), the variation of cross-sectionalspecific investments such as humanresources and capital, and historicalinvestments that make barriers to theentrance of competitive forces (Smithand Watts 1992). The difference in

IOS leads to the difference in optimalityof leverage and dividend policies.

IOS can not be directly observed;instead, it is calculated using proxies.There are several proxies for IOS, cat-egorized into price-based, investment-based, and variance measures. Kallapurand Trombley (1999) find that theprice-based proxy for IOS is moredominant than the other proxies. Thisstudy uses book-to-market value ofequity (price-based category) as it isconsidered the best proxy for IOS.Similar to the research of Kallapur andTrombley (1999), in this study, lowlevel book-to-market value of equityindicates high growth of company.

Companies with high growth willbe forced to choose between payingdividends and incurring capital expen-ditures related to the investment op-portunities. Imperfect capital market4

creates a kind of competition betweendividend policy and investment financ-ing using internal cash flows. Agencytheory predicts that high-growth com-panies will pay out high dividendssince high growth firms especially indeveloping countries lack of investor’sprotection (La Porta et al. 2000b).

Then how about growth options?Management of a company with highgrowth options will have better infor-mation on the existence of growth op-tions; hence, shareholders need to makesuch “luring offers” to managementthat the management gives the infor-mation on growth options for the share-holders’ interests. Asymmetric infor-

4 Often called capital market friction.

188


mation creates conflict such that theinformation needs alignment. Bond-ing strategy in this case is executed bylinking growth option to the companyvalue and dividend payment.

H6: The relationship between book to

market value and dividend is posi-tive.

Profitability

Jensen (1986) argues that exces-sive free cash flows will lead to man-agement committing moral hazard. Afirm with high profitability is potentialto have high free cash flows if highinvestment opportunities do not exist.Therefore, the agency theory arguesthat outside shareholders push manag-ers to pay higher dividends in the pres-ence of higher firm profitability.

H7: The relationship between profit-

ability and dividend is positive.

Determinants of ManagerialOwnership

Management will buy more sharesof the company where they work in ifthe relative return of the ownership islow. Relative return is the loss if anindividual decides to add one share ormore to her portfolio (Crutcley andHansen 1989). As a share is morealigned to the capital market line, theshare is said to have a low relativereturn, therefore is more attractive tobe held. This research differs fromCrutcley and Hansen (1989) in whichthey use risk premium approach tocalculating diversification losses of ashare.

In relation to managerial owner-ship in the presence of higher relativereturn, a manager will have more in-centives to buy the shares of the com-pany where she works in, i.e., the prob-ability that the managerial ownershipis undertaken will be higher.

H8: The relationship between the rela-

tive return and managerial owner-ship is positive.

Research Methods

The sample are manufacturingcompanies listed on the Jakarta StockExchange (JSX) with financial dataand supporting data (returns, right is-sues, and ownership proportion) fullyavailable for the sample period of 1994-2004. The observation years total 1,717observation years. This research usespooling data method and excluded allcompanies that went public after 1996and delisted firms.

The model is constructed usingthree endogenous variables: leverage,dividend and managerial ownership.The endogenous variables are:

a. Leverage (LEV): This research usesthe ratio of the long-term liabilitiesto total liabilities as the proxy forleverage.

b. Dummy Dividend (DDIV): Thisresearch uses dummy variable;DDIV=1 for the observation yearthat pays out dividend and DDIV=0,vice versa, for the proxy for divi-dend. Data that tend to be binomialcause bias if absolute figures areused for dividend (Mahadwartha2002).

189


c. Managerial ownership (DMOWN):This research uses dummy variablefor managerial ownership due to thetendency of the data to be binomial.DMOWN= 1 for observation yearswhich have a proportion of manage-rial ownership and DMOWN= 0,vice versa.

Five exogenous variables in firm-specific characteristics are:

a. Current Assets (CA): Different fromthe research of Jensen et al. (1992)that uses fixed assets in explainingleverage, this research uses liquidassets variable. Besides testing col-lateral hypothesis, it is also used totest perquisite hypothesis by link-ing liquid assets to management per-quisite actions because liquid assetscontain cash and supplies that canbe manipulated by management inperquisite actions. CA variable useslog-linear measurement to decreasedispersion with other variables.

b. Asset Structure (AS): Asset struc-ture variable is calculated by divid-ing net fixed assets from total as-sets. This approach is different fromprevious studies. Barton et al. (1989)add inventories to the net fixed as-sets, whilst Skinner (1993) usesmarket value as the denominator.Different measurement consider-ation in this study is based on somearguments. First, net fixed assetsreflect assets that directly influencecompany operations. Second, inven-

tory items fluctuate in short run, soit can be used as the collateral ofdebt in the long term.

c. Investment Opportunity Set(IOSBM): IOS variable uses book-to-market equity (BE/MVE). Basedon the research of Kallapur andTrombley (1999), BE/MVE vari-able, as the most valid proxy, is usedas a growth proxy. Moreover, BVE/MVE is a growth proxy often usedby researchers in finance (Gaverdan Gaver 1993).5 Low book-to-market value of equity shows highgrowth of company.

d. Profitability uses return on invest-ment measured from operation profitdivided by total assets.

e. Relative Return (RR): uses averagereturn (52 weeks) divided by annualstandard deviation of return withweekly data

This research utilizes sensitivitytest using two conditions: firm sizeand financial crisis. Firm size is acontrol variable, and it uses dummyvariable. Size is measured by choosingthe first 50 percent with null value oftotal asset ascended (DSIZE= 0) andthe low 60 percent with 1 value(DSIZE= 1). It is done to avoid biaseffect on cross section in polling cross-sectional data (Murphy 1985).

E (Ri)

1

RR=

5 The researchers are, amongst the others, Chung and Charoenwong (1991); Collins and Kothari(1989); Lewellen et al. (1987).

190


The period of economic crisis inIndonesia uses dummy variable. Esti-mation period is divided into two sub-periods from 1998 to 2004 with DCRSas the dummy variable for the periodduring the crisis. Period before crisis1994-1997 is represented by the inter-cept. If DCRS is significant, so there isan intercept difference between andduring the crisis periods. DCRS isexpected to create more robust 3SLSestimation.

This research uses three-stage leastsquares (3SLS) as the technical analy-sis. Endogenous, exogenous, and con-trol variables for sensitivity test ofregression function are:

a. Endogenous variables: Leverage(LEV), Dummy Dividend (DDIV),and Dummy Managerial Ownership(DMOWN).

b. Exogenous variables: Log-linearLiquid asset (LnCA), Asset Struc-ture (AS), Investment Opportuni-ties (IOSBM), Return on Invest-ment (ROI) and Relative Return(RR).

c. Sensitivity test of control variable:Dummy Size (DSIZE) and DummyPeriod Crisis (DC).

Research models are as follows:

LEVit=

1 +

11DDIV

it +

12

DMOWNit +

13

LnCAit

+ 14

ASit +

1

DDIVit=

2 +

21LEV

it +

22

DMOWNit +

23

IOSBMit +

24ROI

it +

2

DMOWNit=

3 +

31LEV

it +

32

DDIVit +

33RR

it +

3

Research Findings

Table 1 shows the description ofresearch variables in the period beforecrisis 1994-1997 (DC = 0) and theperiod during crisis 1998–2004 (DC =1). Average leverage does not showsubstantial change, that is 27.88 per-cent in the period before crisis and25.87 percent in the period during cri-sis. Average dividend shows a decreasein the period during crisis. Dividenddecrease occurs since in the crisis pe-riod, most companies suffer from lossand hence do not have enough cash topay dividends. This result is supportedby the decrease in average return oninvestment (ROI measured from op-eration profit divided by total assets)in the period during crisis.

On the other hand, the averageliquid asset rates (LnCA) and assetstructure (SA) in both periods do notshow any substantial changes. Theaverage relative return (RR) showsquite big changes, from 27.86 percentin the period before crisis to 7.73 per-cent in the period during crisis.

Table 2 shows the interdependentanalysis of leverage, dividend, andmanagerial ownership policies usingthree-stage least squares (3SLS). Si-multaneous relationship betweenmanagerial ownership and leverage is

191


Table 1. Statistics Descriptive of Variables before and during Crisis Period

DC N Mean Std. Dev Std. ErrorMean

Leverage 0 632 0.2788 0.2533 0.0101(LEV) 1 1085 0.2587 0.3090 0.0088

Dummy Dividend 0 632 0.7832 0.4124 0.0164(DDIV) 1 1085 0.3278 0.6785 0.0145

Dummy Managerial Ownership 0 632 0.1851 0.3887 0.0155(DMOWN) 1 1085 0.1265 0.3546 0.0124

Log-linier Current Assets 0 632 25.7543 1.3521 0.0538(LnCA) 1 1085 26.2987 1.5089 0.0589

Asset Structure 0 632 0.3608 0.1880 0.0075(AS) 1 1085 0.3908 0.3349 0.0086

Investment Opportunity Set 0 632 1.6718 2.4132 0.0960(IOSBM) 1 1085 0.0773 0.9282 0.0368

Note: DC= dummy crisis; 1994-1997= before crisis; 1998-2004= during crisis

shown in Panel A and Panel C. Panel Ashows that managerial ownershipbrings negative and significant effecton leverage, and the effect is higherthan that of dividend on leverage. Thisresult is consistent with the first re-search hypothesis that there is a nega-tive relationship between managerialownership and leverage.

Simultaneous relationship be-tween managerial ownership and divi-dend is shown in Panel B and Panel C.Both panels show that the relationshipbetween the two variables is positiveand insignificant. This result fails tosupport the second hypothesis that thereis a negative relationship of manage-rial ownership with dividend.

Panel A and B show simultaneousrelationship between leverage policyand dividend policy. Both panels show

that the relationship between the twovariables is positive but insignificant.This result is in line with the thirdhypothesis prediction that there is apositive relationship between lever-age policy and dividend policy, but theinsignificance renders the result mean-ingless.

Panel A shows the effect of twoexogenous variables, which are thedeterminants of leverage: current as-sets and asset structure. Both variablesare evidenced to have positive andsignificant effect on leverage as pre-dicted in the fourth and fifth hypoth-eses.

Panel B shows the effect of twoexogenous variables, which are thedeterminants of dividend: book tomarket and return on investment thatare proxies for investment opportuni-

192


ties and profitability. Both variablesare proven to have positive and signifi-cant effect on dividend as predicted inthe sixth and seventh hypotheses.

Panel C depicts the effect of exog-enous variable, which is relative re-turn on managerial ownership. Therelative return variable has positiveeffect, albeit its insignificance, onmanagerial ownership. The positiveeffect of relative return is expected bythe eight hypotheses. Nevertheless, theinsignificant result does not cause theinterpretation of this variable to be lesscrucial.

In general, the result of interde-pendent analysis of leverage, dividend,and managerial ownership yields a rela-tively moderate result. The mecha-nism of the relationship amongst en-dogenous variables is still partial innature. The particular result requiresfurther investigation to find out thesensitivity of the simultaneous modeltowards the effect of firm size and ofeconomic condition whether in nor-mal or crisis state.

Table 3 examines the sensitivityof the simultaneous model towards theeffect of firm size and crisis period.

Table 2. Three-Stage Least Squares Analysis (Leverage, Dividend, andManagerial Ownership)

Variable Coefficient t-Stat Goodness of fit

Panel A: Endogenous variable (leverage)

Constant -0.8732 -2.4125 ** R2 = 0.1245DDIV 0.0276 1.1365 Adjusted R2= 0.1178DMOWN -0.0523 -7.3239 ***LnCA 0.0652 11.1871 ***AS 0.3012 2.4933 **

Panel B: Endogenous variable (dividend)

Constant 0.2264 12.5332 *** R2 = 0.1378LEV 0.0183 0.5239 Adjusted R2= 0.1322DMOWN 0.2332 1.8675IOSBM 0.0447 3.4386 ***ROI 2.2985 12.8150 ***

Panel C: Endogenous variable (managerial ownership)

Constant 0.2273 5.2852 *** R2= 0.0419LEV -1.3274 -6.9710 *** Adjusted R2= 0.0382DDIV 0.0398 0.2373

RR 0.0461 0.8456

Note: *= 0.10 significant level; **= 0,05 significant level; ***= 0,01 significant level; D = dummyvariable

193


Panel A, B and C show that substantialchanges prevail in the linkage amongstresearch variables. The results in Table3 relatively differ from those in Table2 discussed in the previous section.Based on the particular result, it can beseen that the simultaneous model isalso sensitive to the size of the firm andcrisis period.

The interdependent relationshipbetween managerial ownership anddividend, which is insignificant inTable 2, becomes significant and inline with prediction when firm sizeand crisis period are added. This ishighly caused by the significant de-cease in managerial ownership beforecrisis period and likewise dividend.

Table 3. Three-Stage Least Squares Analysis (Leverage, Dividend, andManagerial Ownership with Sensitivity Analysis to Firm Size andCrisis Period)

Variable Coefficient t-Stat Goodness of fit

Panel A: Endogenous variable (leverage)

Constant -0.1137 -0.0047 R2 = 0.1379DDIV 0.0150 1.1191 Adjusted R2 = 0.1303DMOWN -0.2854 -3.6301 ***LnCA 0.0295 2.5533 **AS 0.2451 8.2382 ***DSIZE 0.1640 2.7220 **DC -0.0499 -2.8720 **

Panel B: Endogenous variable (dividend)

Constant 0.6971 5.5504 *** R2 = 0.3240LEV 0.4926 0.5583 Adjusted R2 = 0.3145DMOWN 0.0337 2.4614 **IOSBM 0.0274 2.1856 **ROI 2.8862 7.5224 ***DSIZE -0.5620 -2.8867 **DC -0.4512 -9.6701 ***

Panel C: Endogenous variable (managerial ownership)

Constant 0.5711 6.8387 *** R2 = 0.0388LEV -0.2582 -4.0081 *** Adjusted R2 = 0.0327DDIV 0.1930 2.4432 **RR 0.0205 1.0519DSIZE -0.0529 -1.1576DC -0.0881 -6.8836 ***

Note: *= 0.10 significant level; **= 0,05 significant level; ***= 0,01 significant level; D = dummyvariable

194


On the other hand, firm size effectis evidenced to have a negative influ-ence on two equations: endogenousvariables (significant) and managerialownership (insignificant). Firm sizepositively and significantly influencesleverage. It indicates that big compa-nies are inclined to have higher debtcompared to small companies. Divi-dend payouts of big companies aresignificantly lower than that of smallcompanies. The level of managerialownership is practically higher for bigcompanies than for small companies,but this finding is not significant.

The effect of economic conditionchange is proven to significantly influ-ence endogenous variables: leverage,dividend, and managerial ownership.This result indicates that leverage levelused in the economic crisis period willbe significantly lower than that in thenormal period. This is mainly causedby the success of restructuring pro-gram done by government, bankingsector, and creditors in suppressingthe problematic companies’ debts. Be-sides, the decrease in debt level is alsotriggered by declining investment op-portunities in the crisis period. Thedividend paid out in the crisis period issignificantly lower than that paid outin the normal period. Likewise, mana-gerial ownership in the crisis period issignificantly lower than that in thenormal period.

Discussion and Implications

This research finding shows thatthere is a negative relationship be-

tween managerial ownership and le-verage. A company with managerialownership has a lower level of lever-age compared with that without mana-gerial ownership. The wealth of man-agers owning managerial ownership isstrongly related to the company wealth.Risk of losing personal wealth willincrease as the risk of company in-creases. One of the causes of risk isdebt. Less debt will decrease the com-pany risk, and so will the risk of man-agers’ personal wealth. Less leverageis expected to decrease the bankruptcyrisk and financial distress. Therefore,if the leverage is higher, managementchance to have managerial ownershipwill be lower.

The result partially fails to testthat there is a significant and negativerelationship between managerial own-ership and dividend. Managerial own-ership does not affect dividend policy,and vice versa. Hence, the result iscontradictory with Rozeff (1982);Jensen et al. (1992); and Mahadwarthaand Hartono (2002), finding that divi-dend policy is affected by thecompany’s ownership structure with anegative relationship. Higher manage-rial ownership will decrease leverageand it forces management to dependmore on retained earnings (retentionratio is higher) to finance the invest-ments that also tends to decrease divi-dend. The effect of dividend policy onmanagerial ownership is not signifi-cant either such that there is no simul-taneous relationship between thosepolicies. This result is on the contrarywith Mahadwartha (2002) who finds

195


that dividend affects managerial own-ership with a negative relationship.

The main rationale behind the factis probably managers’ inclination totake managerial ownership decisionby considering risk factor more thanprofit factor (return from stock owner-ship). Managerial ownership does nothave any substitution characteristic ondividend in decreasing the agency con-flict. The result shows that managerialownership plays a minor role in reduc-ing agency conflict. Firms with low orhigh managerial ownership have in-different impact on agency conflict.This result suggests that Indonesianfirms have more concentrated owner-ship compared to that in other devel-oping countries. Mahadwartha (2004)suggests that Indonesian firms are fam-ily-controlled firms. Hence, future re-search should be more focused on ul-timate shareholders to control agencyconflict in Indonesia. Unfortunately,ultimate shareholders data in Indone-sia are difficult to access fully, if notunavailable.

The relationship between lever-age and dividend with or without con-trol variable is proved to be positive,but not significant. There are two theo-retical explanations for this being onthe contrary with agency theory pre-diction. First, pecking order theoryassumes that dividend is sticky, so itdoes not have any effect on and is notaffected by leverage policy. Leveragepolicy tends to fluctuate, dependingon the need for funds of the companyto invest.

The second is clientele effect. In-vestors have different preferences ondividend. Some investors prefer thecompany to pay high dividends whilstsome prefer low dividend rate or eventzero dividend. Clientele effect explainsthat the company will decide its divi-dend policy based on the investors’preferences, so the leverage policy doesnot influence or is not influenced bydividend policy.

Liquid assets variable tests theperquisite hypothesis that leveragepolicy can be used to suppress perqui-site managerial actions using liquidassets because leverage variable is usedto separate the ownership by manage-ment from that by nonmanagement(outside shareholders). Liquid assetsaffect leverage with a positive andsignificant relationship. Higher liquidassets lead to higher leverage level.Based on the agency theory predic-tion, the liquid assets are utilized tosuppress managerial perquisite actionsrelated with the ability to fulfil itsobligations.

Collateral hypothesis is related tothe management possibility to harnessthe liquid assets for their own busi-nesses, especially cash and invento-ries. Asset structure is used as a guar-antee of the company’s obligations,thus higher asset structure leads tohigher company’s leverage level. Be-sides, agency theory suggests that theuse of leverage will decrease manage-ment perquisite ability and hence makethe relationship between liquid assetsand leverage level positive. The rela-

196


tionship between liquid assets and as-set structure with positive leverage isbased on the prediction of collateraland perquisite hypotheses.

The effect of book to market ondividend is positive and significant.Lower book to market means highergrowth and lower dividend payment.Based on the free cash flow hypoth-esis, higher growth will enable thecompany to pay lower dividends be-cause most of retained earnings areused for investments. Higher growthwill decrease perquisite ability to ex-ploit the company’s cash flows. Thisresult is supported by Jensen et al.(1992), also by Mahadwartha andHartono (2002) who conclude thatgrowing companies pay less dividends.

The relationship of return on in-vestment is significantly positive todividend. Higher company profit willalso give higher dividend becauseshareholders will require higher returnthrough dividend payment. Agencytheory explains that an increase in profitwill means higher cash flows, henceincreasing management ability to takeperquisite actions (Jensen 1986).Therefore, shareholders will apply abonding mechanism through the in-crease in dividend.

The relationship of relative returnto managerial ownership is positive,but not significant. This result shows

that management does not considerstock relative return to hold thecompany’s stocks. It is possibly be-cause stock return is not able to com-pensate the management for the risksuffered from owning the stocks. In-vestors are suggested to avoid compa-nies that have high leverage and do nothave managerial ownership. Compa-nies with high leverage tend not tohave managerial ownership, reflect-ing high managerial expectation onthe company risk. Investors havingalready invested in companies withhigh leverage can compensate for thehigh financial risk through the divi-dend payment.

Managers are suggested to avoidthe use of high leverage because it willincrease the financial risk. Managersthat have managerial ownership in acompany with high leverage tend tosuffer from higher risk than do inves-tors. It causes managers to have doublerisk. First, high leverage brings im-plications on higher company risk, andthen decreasing the value of the com-pany. It also means that the managers’personal wealth as the owners willdecline. Second, a company with highleverage is faced with higher bank-ruptcy risk. It leads to higher manage-ment possibility to lose their jobs. Thefailure of managers to keep and de-velop the company will plunge theirreputation in the labor market.

197


References

Ang, J., J. Chua, and J. McConnell. 1982. The administrative costs of corporate bank-ruptcy: A note. The Journal of Finance: 219-226.

Barton, S. L., N. C. Hill, and S. Sundaram. 1989. An empirical test of stakeholder theoryprediction of capital structure. Financial Management (Spring): 36-44.

Baker, M., and J. Wurgler. 2004. Appearing and disappearing dividends: The link tocatering incentives. Journal of Financial Economics 73: 271-288.

Brickley, J. A., R. C. Lease, and C. W. Smith. 1988. Ownership structure and voting onantitakeover amendments. Journal of Financial Economics 20: 267-291.

Chung K., and C. Charoenwong. 1991. Investment options, assets in place and the risk ofstocks. Financial Management 20: 21-33.

Collins, D. W., and S. P. Kothari. 1989. An analysis of intertemporal and cross sectionaldeterminants of earnings response coeficients. Journal of Accounting and Economics11: 143-181.

Copeland, T. E., and J. F. Weston. 1992. Financial Theory and Corporate Policy (3rd ed.).Addison-Wesley Publishing Company.

Crutchley, C. E., and R. S. Hansen. 1989. A test of the agency theory of managerialownership, corporate leverage, and corporate dividends. Financial Management: 36-46.

DeAngelo, H., and R. W. Masulis. 1980. Leverage and dividen irrelevancy undercorporate and personal taxation. The Journal of Finance: 453-464.

Easterbrook, F. 1984. Two agency-cost explanations of dividend. American EconomicReview: 650-659.

Eisenhardt, K. M. 1989. Agency theory: An assesment and review. Academy of Manage-ment Review: 57-74.

Emery, D.R., and J. D. Finnerty. 1997. Corporate Financial Management (Internationaled.). New Jersey: Prentice Hall Inc.

Ferri, M., and W. Jones. 1979. Determinants of financial structure: A new methodologicalapproach. The Journal of Finance 34: 631-644.

Friend, I., and L. H. P. Lang. 1988. An empirical test of the impact of managerial self-interest on corporate capital structure. The Journal of Finance 43: 271-282.

Gaver, J. J., and K. M. Gaver. 1993. Additional evidence on the association between theinvestment opportunity set and corporate financing, dividend, and compensationpolicies. Journal of Accounting and Economics: 125-160.

Harris, M., and A. Raviv. 1990. Capital structure and the information role of debt. TheJournal of Finance (June): 321-349.

Hartono, J. 2000. An agency-cost explanation for dividend payments. Working Paper.Yogyakarta, Indonesia: Universitas Gadjah Mada.

198


Ismiyanti, F., and M. M. Hanafi. 2004, Struktur kepemilikan, risiko, dan kebijakankeuangan: Analisis persamaan simultan. Jurnal Ekonomi dan Bisnis Indonesia.Yogyakarta, Indonesia: Faculty of Economics- Universitas Gadjah Mada.

Jensen M. C., and K. J. Murphy. 1990. Performance pay and top-management incentives.Journal of Political Economy 98 (2): 225-264.

Jensen, G. R., D. P. Solberg, and T. S. Zorn. 1992. Simultaneous determination of insiderownership, debt and dividend policies. Journal of Financial and Quantitative Analy-sis: 247-263.

Jensen, M. C. 1986. Agency cost and free cash flow, corporate finance, and takeovers.American Economics Review: 323-329.

Jensen, M. C., and W. H. Meckling. 1976. Theory of the firm: Managerial behavior agencycosts and capital structure. Journal of Financial Economics: 305-360.

Kallapur, S., and M. A. Trombley. 1999. The association between investment opportunityset proxies and realized growth. Journal of Business Finance and Accounting 26: 505-519.

Kim, S.W., and E.H. Sorensen (1986), “Evidence on The Impact of The Agency cost ofdebt on Corporate Debt Policy,” Journal of Financial and Quantitative Analysis 21,131-144.

Koch, P. D., and C. Shenoy. 1999. The information content of dividend and capitalstructure policies. Financial Management: 16-35.

Leland, H. E., and D. H. Pyle. 1977. Information asymmetries, financial structure, andfinancial intermediation. The Journal of Finance 32: 371-387.

Lewellen, W., C. Loderer, and K. Martin. 1987. Executive compensation contracts andexecutive incentive problems: An empirical analysis. Journal of Accounting andEconomics 9: 287-310.

Long, M.S., and I. B. Malitz. 1985. Investment patterns and financial leverage. In NationalBureau of Economic Research: Corporate Capital Structures in the United Statesedited by B.M. Friedman (ed): 325-348.

Mahadwartha, P. A., and H. M. Jogiyanto. 2002. Uji teori keagenan dalam hubunganinterdependensi antara kebijakan hutang dengan kebijakan dividen. SimposiumNasional Akuntansi (5-6 September). Universitas Diponegoro, Semarang.

Mahadwartha, P. A. 2002. The association of managerial ownership with dividend policyand leverage policy: A logit model. Proceeding Simposium Surviving Strategies toCope with The Future (September 13th). Atmajaya Yogyakarta University.

Mahadwartha, P. A. 2004. Pengawasan dan pengikatan berbasis kepemilikan institusionalinternal. Desertation (unpublished). Universitas Gadjah Mada.

Mahadwartha, P. A., and F. Ismiyanti. 2006. Debt policy, free cash flow hypothesis, andbalancing of agency theory through ownership: Evidence from Indonesia. Journal ofCorporate Ownership and Control 5 (1).

Marsh, P. 1982. The choice between debt and equity: An empirical study. The Journal ofFinance 37: 121-144.

199


Megginson, W. L. 1997. Capital Structure Theory, Corporate Finance Theory. ReadingMassachusetts: Addison-Wesley.

Miller, M., and K. Rock. 1985. Dividend policy under asymmetric information. TheJournal of Finance: 1031-1052.

Murphy, K. J. 1985. Corporate performance and managerial remuniration: An EmpiricalAnalysis,” Journal of Accounting and Economics 7, 11-42.

Myers, S. C. 1977. The determinant of corporate borrowing. Journal of FinancialEconomics 5: 147-176.

Myers, S. C., and N. S. Majluf. 1984. Corporate financing and investment decisions whenfirms have information investors do not have. Journal of Financial Economics 13;187-221.

Porta R. L., F. Lopez-De-Silanes, A. Shleifer, and R. W. Vishny. 2000a. Agency problemsand dividend policies around the world. The Journal of Finance 55 (1): 1-33.

Porta R. L., F. Lopez-De-Silanes, A. Shleifer, and R. W. Vishny. 2000b. Investorprotection and corporate governance. Journal of Financial Economics 58: 3-27

Ross, S., R. W. Westerfield and J. Jaffe. 1999. Corporate Finance (5th ed.) Irwin: McGraw-Hill.

Rozeff, M. S. 1982. Growth, beta and agency costs as determinants of dividend payoutratios. Journal of Financial Research: 249-259.

Skinner, D. J. 1993. The investment opportunity set and accounting procedure choice.Journal of Accounting and Economics 16: 407-445.

Smith Jr. C. W., and R. L. Watts. 1992. The Investment opportunity set and corporatefinancing, dividend, and compensation policies. Journal of Financial Economics 13:187-221.

interdependent analysis of leverage, …...will be tested using two-stage least squares. five...

Documents