DOES MARKET MICROSTRUCTURE MATTER? FOREIGN AND DOMESTIC INSTITUTIONAL OWNERSHIP TO AGENCY COST

This research provides measure of absolute and relative equity agency costs for corporations under different ownership and management structures. Miller (1977) argues that divergence of opinion among investors causes the price difference of the price of a security. The dispute mechanism causes the forming price to be further of closer to its intrinsic value. Greater the divergence of opmion, causes greater the gap between the price and its' intrinsic value. This study tests a new condition that reflects the existence of agency conflict, which is the conditions of stock price premium and stock price discount and related to agency cost control mechanism through foreign and domestic institutional ownership. The two conditions then called as price spread. This study tests four interrelated hypotheses in conditions of stock price premium and stock price discount that related to agency cost, foreign institutional ownership and domestic institutional ownership. Analysis method employs complete structural equation model (SEM), and multigroup S E M with constrained and unconstrained parameters. The direction of the study results consistent with result prediction. Nevertheless, there is one insignificant relationship, which is domestic institutional ownership towards agency cost. This indicates that the relationship hold but remains statistically unproven.


Research Background
The social and private costs of an agent's actions due to incomplete alignment of the agent and owner's interests brought to attention by the seminal contributions of Jensen and Meckling (1976) on agency costs. Agency theory has also brought the roles of managerial decision rights and various external and internal monitoring and bonding mechanisms to the forefront of theoretical discussions and empirical research. Great strides made in demonstrating empirically the role of agency costs in financial decisions such as in explaining the choices of capital structure, maturity structure, dividend policy and executive compensation. However, the actual measurement of the principal variable of interest, agency costs, in both absolute and relative terms, has lagged behind.
j > t r , ^ ^ To measure absolute agency costs, a zero agency cost base case must be observed to serve as the reference point of comparison for all other cases of ownership and management structures. In the original Jensen and Meckling agency theory, the zero agency cost base case is, by definition, the firm owned solely by a single owner manager. When management owns less than 100 percent of the firm's equity, shareholders incur agency costs resulting fi-om management's shirking and perquisite consumption. Because of limitations imposed by personal wealth constraints, exchange regulations on the minimum numbers of shareholders, and other considerations, no publicly traded firm entirely owned by management. Thus, Jensen and Meckling's zero agency cost base case unproven among the usual sample of publicly traded firms for which information is readily available. The absence of information about sole owner manager firms explain why agency costs often inferred but not directly measured in the empirical finance literature.
At one extreme of ownership and management, structures are firms whose managers own 100 percent of the firm. These firms, by their definition, have no agency costs. At the other extreme are firms, by their definition, having no agency cost. At the other extreme are firms whose managers are paid employees with no equity in the firm. In the between are firms where the managers own some, but not all, of their firm's equity.
Agency costs emerge when the interests of the firm's managers less aligned with those of the firm's owners, and take the form of preference for on the job perquisites, shirking and making self interested and entrenched decisions that reduce shareholder wealth. The magnitude of these costs is limited by how well the owners and delegated third parties, such as banks, monitor the actions of the outside managers. Miller (1977) proposes a theory explaining the creation of price between selling investor and buying investor. The theory in which Miller propose loosens the assumption of homogenous expectation in balance model. Miller (1977) argues that divergence of opinion among investors causes the price difference of the price of a security. The dispute mechanism causes the forming price to be further of closer to its intrinsic value. Greater the divergence of opinion, causes greater the gap between the price and its' intrinsic value.
This study utilizes Miller's theory, which states disputes between buying and selling investor caused by divergence of opinion, which in this study focused to the divergence of opinion the magnitude of agency cost. Divergence of opinion on the magnitude of agency cost in a period will certainly find a dominant party. This is shown by the formation of a closing price, which is agreed by buying investor and selling investor.
How does the trade activity reflect a company's agency cost? Stockholders as owners of a company have strong interest with the price of the stocks they own. Beside stockholders, other parties are also interested in the stock price are potential stockholders or potential investors. The selling and buying process of stocks becomes the process of accomplishment of the agreement point, which produces closing price. How is the process of reaching closing price? Sellers (old stockholders) will sell with an as high ask or offer price as possible and the buyer (potential stockholder or potential investor) will try to buy with as low bid price as possible. In the ending session of ask-bid, closing price produced from the price bargaining process between the seller and the buyer.
The next question is how do sellers set an offer price. How do buyers set a bid price? This study proposes an idea that the setting of offer price and bid price reflects conflict of interest between the parties in the company (management, stockholders and creditors). Conflict of interest between the parties in a company called agency conflict. Agency conflict, according to Jensen and Meckling (1976), reflected in agency cost that the company bears and influence the wealth of all stakeholders. If the agency conflict is low, then closing price achieved in the transaction process will be closer to offer price. In the contrary, if the agency cost is high, then closing price achieved in the transaction process will be closer to ask price.
Studies about agency theory ignore the existence of agency conflict that reflected in the process of achieving closing price through negotiations between offer price and bid price. The studies more focused on the agency conflict control mechanism. This study tests a new condition, which reflects the existence of agency costs, which are stock price premium condition and stock price discount condition and related to agency conflict control mechanism through foreign institutional ownership and domestic institutional ownership. Both conditions then called price spread. Stock price premium is a condition that shows that closing price of company's closing price tends to be closer to offer price. Meanwhile, stock price discount is a condition that shows that closing price of company's stocks tends to be closer to bid price. Stock price premium and stock price discount show expectations of stockholders for ownership structure and agency cost which effects company's performance. This study proposes that price stock price premium condition and stock price discount are important issue in identifying the agency cost the company has to bear. Both conditions also reflect the level of agency conflict in the company. This study argues that in stock price premium condition, agency conflict is low and agency conflict is high in stock price discount condition.
Closing price of the company's stock that is close to offer price shows that old stockholders (sellers) can obtain a price close to their offer price. Potential stockholders (buyers) are willing to buy with a price close to offer price is possibly because the stock is considered profitable in the future. Old stockholder tend to hold their company's value on to the offer price. If old stockholders are convinced that the value of the company can be increased then they will retain the stock price in high offer price. This causes closing price that agreed between the parties to be close to offer price. In this case, potential stockholders are also convinced that the value of the company can be increased in the future. This study assumes that an expectation towards high company value caused by low agency cost. If old stockholders and potential stockholders percept that agency conflict is low, they value the company higher than the value of other similar companies. Closing price condition that is close to offer price in this study called as stock price premium condition.
Stock price discount condition is a condition in the contrary of stock price premium condition. Closing price that is close to bid price shows that old stockholders (sellers) forced to sell their stocks with lower price then their offer. This is possibly because the offer price not responded by the market and old stockholder is in the position to sell their stocks immediately. Potential stockholders (buyers), in this condition, obtain the stocks with a price close to their bid price. In this condition, old stockholders realize that agency cost of the company is high, thus estimated that the future movement of stock price is unprofitable. Old stockholders forced to give a discount to potential stockholders. This condition causes the tendency of company's closing price agreed by both parties to be close to bid price. Potential stockholders (buyers) are convinced that high agency cost causes the company's value to be low but the value of the company can still be increased through the mechanism of ownership structure and financial policies of other companies.
This study conducts observations of information of daily closing price (agreed) between sellers and buyers and the difference between offer price and closing price, and between bid price and closing price. Therefore, we can say that this study employs the perspective of market microstmcture to explain agency cost. Studies in the subject of microstructure give a deep understanding in examining the behavior and operation of the capital market based on intraday movement. This study employs a microstructure approach that combined with corporate finance research model. Amihud and Mendelson (1986) states that bid ask spread measurement can be used to determine price of an asset (stocks, bonds, and others). Their studies in the microstructure are also useful to determine the value of an asset in corporate finance. This study relates the findings of bid-ask spread as an indicator in determining stock price with agency cost and introduce price spread condition, which consists of stock price premium and stock price discount.
So far, studies on the agency theory do not test the existence of bargaining between old stockholders (sellers) with potential stockholders (buyers) in achieving closing price. Different closing prices reflect different agency conflicts among companies. The effect of different agency conflicts among companies will cause a number of companies to be in stock price premium condition and others in stock price discount condition. The condition of stock price premium and stock price discount difference causes the effect of foreign institutional ownership and domestic institutional ownership as agency conflict control mechanisms towards agency cost to be different. Identification m Fitri Ismiyanti of stock price premium and stock price discount conditions in this study is expected to give better explanation the different effects various agency cost reduction mechanism. This study focuses on foreign institutional ownership and domestic institutional ownership as agency conflict reduction mechanism.
The existence of different results fi"om past studies about the relationship between foreign institutional ownership and domestic institutional ownership to agency cost urge researcher to test their relationship between the three constructs. This study also introduces price spread condition, which is expected to explain the difference among the effects of agency conflict control mechanism through foreign institutional ownership and domestic institutional ownership better.

Literature Review and Hypothesis Development 2.1. Agency Cost
The core of agency theory is the existence of conflict of interest between agents and principal. The agency cost, which occurs because of this conflict of interest, reduces the value of the company. Equity agency cost includes monitoring cost, bonding cost and residual loss (Jensen and Meckling, 1976). Monitoring cost include principals' expense, in the effort to control agent's behavior through budget tightening, compensation policy, and operational regulations. Bonding costs are the agent's expense to guarantee that agents will not conduct certain actions that will inflict financial loss towards principals or to guarantee that principals will give compensations if agents conduct certain actions.
Residual loss includes the monetary value of principals' wealth reduction because of different interests between agents and principals, which stimulate agents, conduct selfish actions and inflict financial loss to principals. The action of this agent can be in the form of inefficient actions such as investing in unprofitable investments or make wasteful expenses. Moreover, there is also debt agency cost that includes paying too much dividend, monitoring cost and bonding cost. Reduction of agency cost can be achieved through a number of mechanisms such as through manager stock ownership, combining financing sources from debts and equities, and dividend payout (Crutchley and Hansen, 1989).
This study will relate agency cost and price spread condition between stock price premium and stock price discount experienced by the company. Agency conflict experienced by the company will be reflected in the spread of stock price premium ad stock price discount of the company. Agency conflict reflected in the stock price premium and stock price discount condition are an agency conflict that is called perceived conflict. Therefore, this study employs perceived agency conflict (stock price premium and stock price discount) to explain actual agency conflicts. Wurgler (2004a, 2004b) employ the term stock price premium to explain reasons of companies that pays dividend and companies that do not pay dividend. This study adopts the term stock price premium and stock price discount but to test the influence of ownership structure and agency cost towards company performance. Stock price premium and stock price discount will be called price spread condition.

Price Spread: Stock Price Premium and Stock Price Discount
Price spread condition of stock price premium and price spread conditions of stock price discount are implications from company's agency conflict. Closing price that is close to offer price and bid price shows that buyers and sellers do take into account agency cost in daily transactions. The level of agency conflict will cause difference between closing price and the offer and bid price. The stock price premium condition shows that the closing price of a company is close to offer price. The stock price premium condition reflects the low level of agency conflict. On the other hand, the stock price discount condition shows that the closing price of a company is close to bid price. The stock price discount condition reflects the low level of agency conflict.

Foreign and Domestic Institutional Ownership
Ownership structure becomes important in agency theory because most agency conflict arguments are caused by ownership and control separation. Agency conflict does not occur in companies with 100% management ownership (Jensen and Meckling, 1976). The condition where new owners buy company's stocks causes discrepancy of interest between the parties in the company. Pure conflict occur between principals and agents as discussed in positivist agency theory and conflicts between stockholders, management, employees and other parties are within principal-agent research (Eisenhardt, 1989).
Institutional ownership can be used to reduce agency conflict (Shleifer and Vishny, 1986;Jarrel and Poulsen, 1987;Brickley et al., 1988;Graves and Waddock, 1990;Han et al, 1999;and Varma, 2001). The studies argue that institutions that invest in the company will monitor a company better. Institutions have professionals in the field of investing which understands the appraisal mechanisms of companies and conduct monitoring towards managers.
Institutional ownership selling will drive down the price of the stocks therefore institutional owners avoid selling their stocks and them conduct monitoring towards the company instead. Institutional ownership expects by condicting effective monitoring the value of the company would increase. The most efficient method employed by institutional owners is by informal discussions with managers.
Hypothesis development based on the argument that this study is developed in the conceptual element. Hypothesis development consists of a number of components, which are foreign institutional ownership, domestic institutional ownership, agency cost and company performance in stock price premium and stock price discount condition.
Agency theory argues that institutional ownership will decrease agency conflict because the institution will help monitor the company so management will not conduct actions that will inflict financial losses towards stockholders (Crutchley et al., 1999;Chen and Steiner, 1999). This is valid in the condition where institutional owner partially monitor the management. However, in the condition where institutional owner is the majority owner, then monitoring would be focused only for the interest of owning institution and ignores public stockholders interest. Foreign institutional ownership can be utilized as control method to decrease agency cost. The higher the foreign institutional ownership, the lower the agency cost and the lower the foreign institutional ownership, the higher the agency cost.
Hi: Foreign institutional ownership has negative influence towards agency cost.
This study assumes that agency conflict in stock price premium condition will be lower compared to stock price discount condition. This assumption cause's agency conflict reduction mechanism through foreign institutional ownership in stock price premium condition will have influence less negative compare to stock price discount condition. Companies with low agency conflict will closely observe the cost to control agency conflict. So they tend to decrease conflict reduction mechanism through ownership structure to drive cost down.
H2: Foreign institutional ownership will affect agency cost negatively; lower when firm is in stock price premium than stock price discount condition. ^ Domestic institutional ownership also acts as a monitoring party, similar to foreign institutional ownership. Core and Larcker (2002) found a negative relationship between stock performance and domestic institutional ownership. Companies with high institutional ownership (more than 5%) indicate its ability to monitor the management. Great institutional ownership causes the utilization of company's assets to be more efficient. Therefore, the proportion of institutional ownership acts as a method to prohibit management from inefficient. Ismiyanti and Hanafi (2004) found that the average institutional ownership in 1997-2001period reaches 66% of total stocks outstanding. This result shows that public (individual) investors, management, directors and institutional ownership hold 34% of stocks. This is different in the United States, institutional ownership reaches 52.36%) from total stocks outstanding in 1999 (Chen and Steiner, 1999). Domestic institutional ownership can be utilized as confrol method to decrease agency cost.

H3: Domestic institutional ownership has negative influence toward agency cost
This study assumes that agency conflict in stock price premium condition will be lower compared to stock price discount condition. This assumption cause's agency conflict reduction mechanism through foreign institutional ownership in stock price premium condition will have influence less negative compared to stock price discount condition. Companies with low agency conflict will closely observe the cost to control agency conflict. So they tend to decrease conflict reduction mechanism through ownership structure to drive cost down.

H4: Domestic institutional ownership will affect agency cost negatively;
lower when firm is in stock price premium than stock price discount condition The relationship between foreign institutional ownership, domestic institutional ownership and agency cost through stock price premium and stock price discount condition, describes in research framework as below:

Data and Sample
Samples employed in this study are non-fmancial companies listed in Jakarta Stock Exchange from 1995 to 2004. Financial data obtained from annual financial report, which consists of balance sheet, income (profit and loss) statement, cash flow report and financial report notes. The data source for the study is Jakarta Stock Exchange Library, Indonesian Capital Market Directory (ICMD) and Indonesian Securities Market Database (ISMD) published by Faculty of Economics Gadjah Mada University.

Operational Definition Variables
Agency cost proxy employed in this study refer to asset utilization and operational cost (Ang et al., 2000); and free cash flow (Hackel et al., 1996) but with adaptation based on certain benchmark value. Agency cost calculation method ideally uses the difference of residual loss from 100% company ownership minus residual loss from non 100% ownership. Sample collection constraint towards 100% company ownership causes the utilization of a benchmark value. The asset utilization measure agency cost based on asset turnover. Asset turnover is a ratio between total sales and total asset. Selling and General Administrative (SGA) is included in operational expense proxy.
Operational expense measures the agency expenses based on SGA which a ratio between operational expense and total sales. Free cash flow used in this study employs free cash flow counting method developed by Hackel et al. (1996) which modifies the traditional free cash flow calculation. This method I Fitri Ismiyanti was selected because it is the most appropriate method for the condition of cash flow statement in Indonesia to avoid sample decrease. FCF = TFCF + DOCO + DCEX TFCF (OCR -OCX)) -CFX where: TFCF = traditional free cash flow OCR = operating cash inflow OCO = operating cash outflow CEX = capital expenditure DOCO = (OCOgrowth -salesgrowth) * 2.0 (* OCO) where: DOCO = discretionary operating cash outlay; Hackel et al. (1996) assumes 20% of OCO is discretionary of OCO and sales growth. Foreign institutional ownership is the sum and percentage of stocks owned by foreign institution. Domestic institutional ownership is percentage ownership by a legal entity registered as non-public stockholder.

Method of Analysis
This research employs Structural Equation Modeling (SEM) in hypotheses testing because SEM has the ability to combine measurement model and structural model. This research applied two stage approaches for multigroup structural equation modeling (MSEM). MSEM do not require nested model to estimate different hypotheses groups in path-analytic model coefficient or model fit coefficient. A series of statistical goodness-of-fit indicators were employed to test a complex model for every group.
Moderating variable test is SEM was conducted in two structural models, which are constrained parameters model and unconstrained parameters model. In models with constrained parameters, regression estimate weight controlled for both sample groups thus having similar estimated relationship. Moderating variable is significant if models with unconstrained parameters are better than models with constrained parameters.

Result and Discussion
Result of full structural equation model in this research will be used to analyze research hypotheses that do not contain stock price premium and stock price discount moderating variables, which are Hi and H3. Hypotheses which used stock price premium and stock price discount moderating variables (H2 and H4), are tested by employing multigroup structural equation model by using constrained parameters and unconstrained parameters models.  Table 2 shows test result by multigroup structural equation model with constrained parameters. Regression coefficient value of ownership structure influence (foreign institutional ownership and domestic institutional ownership) towards agency costs is not different when compared between stock price premium sample and stock price discount sample. The numbers of data used for the study are 1559 samples comprising of 713 samples with stock price premium and 846 samples in stock price discount.  Table 3 shows test result by multigroup structural equation model with unconstrained parameters. Regression coefficient value of ownership structure influence (Foreign institutional ownership and domestic institutional ownership) towards agency cost is not different when compared between stock price premium sample and stock price discount sample. The numbers of data used for the study are 1559 samples comprising of 713 samples with stock price premium and 846 samples in stock price discount.
Goodness of fit of model with unconstrained parameters (GFI= 0.976) is found to be better than model goodness of fit of model with constrained parameters (GFI= 0.942). In addition, the difference at chi square value is 56.585 with 4 degree of freedom show a significant result (p < 0.10). Therefore, base model and alternative model based on the difference of stock price premium and stock price discount are significantly different. This indicates that different price spread condition significantly influential as moderating variables. Variable moderation of price spread condition mainly seen on the difference between foreign institutional ownership, domestic institutional ownership and agency cost on stock price premium and stock price discount. Comparison between base model and alternative model shown in Table 4.  Table 4 shows comparison of test result between base model (constrained model) and alternative model (unconstrained model). The values being compared are goodness of fit value, chi square value and degree of freedom of both test models to determine whether stock price premium and stock price discount is significantly moderate relationships in this model. Table 4 shows the result of test comparison using constrained parameters and unconstrained parameters. The table shows increase of goodness of fit values, from base model to alternative model. Goodness of fit value analyzed is chi square value, which changed 56.585 points, and degree of freedom, which changed 4 points. Based on goodness of fit of base model and alternative model, it can be concluded that relationship between variable of agency cost and performance, moderated by stock price premium and stock price discount.   Table 6 shows comparison resuh prediction with model test resuh using full structural equation model. The direction of the study results found to be consistent with result prediction, which is negative. Nevertheless, there is one insignificant relationship, which is domestic institutional ownership towards agency cost. This indicates that the relationship is practically proofened but remain statistically unproven.  Table 6 is a summary table between result predictions with research results, which utilizes multigroup structural equation model in stock price premium, and stock price discount as moderating variable. The result of the study shows that coefficient value of stock price discount should be lower than coefficient value of stock price premium. The influence of agency cost towards performance found to be different. Coefficient of stock price premium is lower than stock price discount.  Table 7 show a summary of hypotheses test result of the study. Next studies expected to reexamine agency conflict magnitude proxy through stock price premium and stock price discount. This is to support the findings of this study and that the proxy can be used as measurement method of agency conflict in Indonesia. Further studies can also develop other mechanisms in agency conflict control such as debt policy and dividend policy. firm is in stock price premium than stock price discount condition. H4: Domestic institutional ownership will affect agency cost negatively; lower 4i when firm is in stock price premium than stock price discount condition.

Result and Discussion
Explanation: •*: Empirical result consistent to theoretical prediction and significant •: Empirical result consistent to theoretical prediction and not significant This will enrich findings that support measurements of stock price premium and stock price discount conflict. For investors, this study suggests investors to select companies, which have low agency conflicts. This will then influence the company's stock prices. Companies will then be encouraged to reduce conflict and increase their financial performance. If this situation can be, achieved Indonesia will be a profitable investment area.