The effect of the firm's monopoly power on the earnings response coefficient.

Lee, Kyung Joo ; Jin, Jongdae ; Huh, Sung K. 等

ABSTRACT

The purpose of this study is to provide further evidence on the determinants of the earnings response coefficient (ERC). More specifically, we have attempted to examine the 'monopoly power' of a firm as an additional factor affecting the ERC.

Using a firm valuation model that explicitly incorporates the degree of monopoly power in its product markets (Thomadakis, 1976; Subrahmanyan and Thomadakis, 1980), we demonstrate that the ERC is positively related to the firm's monopoly power.

This theoretical prediction is empirically examined using a sample of 144 Korean firms listed in the Korean Stock Exchange during the period extending from 1986 to 1992. The sample firm's monopoly power is measured by whether or not the firm is designated as a market-dominant enterprise by the Korean Fair Trade Commission according to the Monopoly Regulation and Fair Trade Act. Such designation implies that the firm has a monopoly power.

The empirical results are generally consistent with the theoretical prediction. Specifically, the ERC is higher for the designated firms than for the non-designated firms. This result is robust across different methods.

INTRODUCTION

The determinant of cross-sectional and/or inter-temporal variations of the earnings response coefficient (hereafter, ERC in short) has been investigated in quite a few previous studies (e.g., Kormendi and Lipe, 1987; Collins and Kothari, 1989; Easton and Zmijewski, 1989; Dhaliwal, Lee and Fargher, 1991; Dhaliwal and Reynolds, 1994; Ahmed, 1994; Kallapur, 1994; Choi and Jeter, 1992; Biddle and Seow, 1991; Teets, 1992; Collins and Salatka, 1993; Bandyopadbyay, 1994). The determinants of the ERC identified in previous studies are characteristics of the firm's earnings generating process, systematic risk of common stock, firm size, the default risk, growth opportunity, cost structure, dividend payout ratio, audit opinion, industry, and interest rates. However, the effect of a firm's monopoly power on the ERC has not been extensively investigated, so far. Thus, the purpose of this study is to examine the effect of a firm's monopoly power on the ERC using Korean capital market data.

Using a firm valuation model that explicitly incorporates the degree of monopoly power in its product markets (Thomadakis, 1976; Subrahmanyan and Thomadakis, 1980), we demonstrate that the ERC is positively related to the firm's monopoly power. This theoretical prediction is empirically tested by comparing ERC's between the firms designated as market-dominant enterprises by the Monopoly Regulation and Fair Trade Act and the other firms. To the extent that designation as a market-dominant enterprise is an appropriate proxy for the degree of monopoly power, we expect the ERC's of the designated firms to be higher than those of the non-designated firms.

The remainder of this paper is organized as follows. In the next section, we derive the theoretical relationship between the firm's monopoly power and the ERC within the framework of a firm valuation model developed by Thomadakis (1976), Subrahmanyan and Thomadakis (1980), and Ahmed (1994). Section three contains our research hypothesis and research methodology. Section four describes sample selection procedure and descriptive statistics for the variables used. The empirical results are presented in section five. A summary of the results and some suggestions for future research appear in final section.

MONOPOLY POWER AND EARNINGS RESPONSE COEFFICIENT

A firm valuation model based on cash flow has been used in many previous ERC literature such as Kormendi and Lipe (1987), Collins and Kothari (1989), Dhaliwal, Lee and Fargher (1991), and Dhaliwal and Reynolds (1994). On the other hand, Thomadakis (1976) and Subrahmanyan and Thomadakis (1980) developed a model that incorporates the degree of monopoly power in the valuation of a firm. By combining these two valuation approaches, we develop a valuation model that describes a functional relationship between the monopoly power and the ERC.

To simplify the analysis, we make the following assumptions:

Assumption 1: The demand function faced by the firm is

[p.sub.t] = [a.sub.t] [q.sup.-n.sub.t]

Where p = the price of a unit of product in period t;

q = the quantity of output chosen by a firm in period t;

a = a random variable representing the uncertainty concerning the demand function. Its mean, variance and covariance are constant through time;

n = the measure of a firm's monopoly power, 0 [less than or equal to] n [less than or equal to] 1.

From Assumption 1, the marginal revenue (MR) is (1-n) [aq.sup.-n] and the average revenue (AR) is [aq.sup.-n]. Therefore, the firm is in a perfect competition when n=0 and hence MR=AR= p, while the firm is in a monopolistic position when n>0. Also, since n is the inverse of the elasticity of demand, the demand becomes more inelastic as n increases. Therefore, n is reliable measure of a firm's monopoly power. This is also evident from the fact that Lerner's index, a popular measure of the monopoly power, is usually defined as 1/(1-n).

Assumption 2: The cost function for the firm is

[TC.sub.t] = [c.sub.t] [q.sub.t]

Where TC = the total cost;

c = the operating cost per unit of output in period t. c is invariant to the level of output, and its mean, variance and covariance are also constant through time.

Assumption 3: As a discount rate for the capitalization of a firm's future cash flow, a single period CAPM is applicable to each period. Also, it is assumed that the market parameters in the CAPM (risk-free rate, market returns, and systematic risk) are exogeneous and constant through time. Thus, the risk-adjusted expected return for a firm in period t (Kt) is: [K.sub.t] = [R.sub.f] + [beta] [ E ([R.sub.m])-[R.sub.f]]

Where [beta] = the systematic risk of the firm;

[R.sub.f] = the risk-free interest rate;

[R.sub.m] = the rate of return on market portfolio

Using the above assumptions, a firm valuation model is developed in a two period world (Thomadakis, 1976). The firm's problem is to choose its output level that maximizes the present value of its future cash flows. In a two period world, the sequence of events is as follows: At the beginning of the first period (t=0), the firm chooses the optimal level of output (Q) for the first period based on expectation about future cash flows (i.e., prices and costs). At the end of the first period (t=1), prices and operating costs for the first period are realized. The firm revises its expectations about period 2 cash flows and chooses the optimal quantity for period 2 based on the revised expectations. The firm is liquidated at the end of the second period. Under this setting, the firm's value at time 0 ([V.sub.0]) can be described as follow.

[V.sub.0] = [k.sub.1] [Q.sub.1] + n[E.sub.0] ([P.sub.1][Q.sub.1] - [c.sub.1][Q.sub.1]) / 1 + [K.sub.1] + n[E.sub.0] ([p.sub.2][Q.sub.2] - [c.sub.2][Q.sub.2]) / (1 + [K.sub.1]) (1 + [K.sub.2])........ (1)

Where [p.sub.t] = the price of a unit of product in period t;

[Q.sub.t] = the quantity of output chosen by a firm in period t;

[K.sub.t] = the risk-adjusted expected return for a firm in period t;

[k.sub.1] = the actual risk-adjusted return for a firm in period 1;

n = the measure of a firm's monopoly power, 0 [less than or equal to] n [less than or equal to] 1.

Abnormal returns or excess returns for the first period (AR1) are computed by the difference between realized returns (R1) and expected returns (ER1) as follows:

[AR.sub.1] = [R.sub.1] - [ER.sub.1] = [V.sub.1] - [V.sub.0] + [D.sub.1] / [V.sub.0] - [E.sub.0] ([V.sub.1]) - [V.sub.0] / [V.sub.0]

Where [D.sub.1] = the dividend paid to stockholders after deducting investments for the second period from the realized cash flows in period one.

Two additional assumptions regarding the firm's earnings generating process are made to develop a model for abnormal returns. First, cash flows to the firm and accounting earnings ([X.sub.t]) are identical (i.e., [X.sub.t] = [p.sub.t] [q.sub.t] [-c.sub.t] [q.sub.t]). Second, the firm's earnings have time-series characteristics described by the following model:

[E.sub.1] ([X.sub.2]) - [E.sub.0] ([X.sub.2]) = [lambda] [[X.sub.1] - [E.sub.0] ([X.sub.1])]

Where [lambda] = the extent to which the current period's earnings shock affects the revisions in expectations of future earnings, usually referred to as persistent coefficient. The sign and value of will depend on the time-series properties of the firm's earnings.

It can be shown that [lambda] is a function of time-series model parameters even when earnings generating process is specified by a general ARIMA(pdq) model (Collins and Kothari, 1989).

Then Abnormal returns or excess returns for the first period ([AR.sub.1]) can be described as follow:

[AR.sub.1] = [1 + n[lambda] / 1 + [R.sub.f] + [beta][E([R.sub.m]) - [R.sub.f]]] [X.sub.1] - [E.sub.0] ([X.sub.1]) / [V.sub.0] ........ (2)

It is obvious from equation (2) that the impact of [beta], [lambda], E(Rm)-[R.sub.f], and n on the ERC (the bracket term) are, ceteris paribus:

[partial derivative]ERC/[partial derivative][beta] < 0, [partial derivative]ERC/[partial derivative][lambda] > 0, [partial derivative]ERC/[partial derivative][E ([R.sub.m]) - [R.sub.f]] < 0, [partial derivative]ERC/partial derivative]n > 0

The first three results reveal that, if other factors be constant, the ERC is negatively related to both the systematic risk of the firm ([beta]) and the market risk premium (E(Rm)-[R.sub.f]), but positively related to the persistence coefficient ([lambda]). Previous studies such as Kormendi and Lipe (1987), Easton and Zmizewski (1989), and Collins and Kothari (1989) provide empirical evidence consistent with these predictions. The fourth comparative static result indicates that the ERC is a positive function of the firm's monopoly power (n) in its product markets.

HYPOTHESIS AND RESEARCH DESIGN

Hypothesis

The research question addressed in this study is whether there is an association between firm's monopoly power and the ERC. The analytical results in the preceding section suggest, among other things, that the a firm's monopoly power is positively related to the ERC.

As a surrogate for the firm's monopoly power, the firm's designation as a market-dominant enterprise by the Monopoly Regulation and Fair Trade Act is used. According to the Monopoly Regulation and Fair Trade Act, the Korea Fair Trade Commission designates and notifies market-dominant enterprises at the beginning of each year. A firm with its annual domestic sales exceeding 100 billion won is designated as a marker-dominant enterprise if its market share is over 50% or 75% (combined with up to 3 other designated firms) in a same or similar industry. If a firm is designated as such, the firm (hereafter, designated firm) has a higher degree of monopoly power relative to other firms that are not designated (hereafter, non-designated firms).

A testable hypothesis for the positive relationship between the ERC and the firm's monopoly power derived herefrom would be, Hypothesis: Earnings response coefficients of designated firms are higher than those of non-designated firms.

Measurement of Variables

Under an assumption that earnings are described by the random walk with a drift model. Expected earnings, E(X), can be written as follows:

[E.sub.t-1] ([X.sub.t]) = [X.sub.t-1] + [[delta].sub.t]

Where [X.sub.t] = the earnings at time t;

[delta] = a drift term obtained by averaging earnings changes for the 5 previous years.

Unexpected earnings (UE), excess of actual earnings over expected earnings, can be described as follow:

[UE.sub.it] = [X.sub.it] - ([X.sub.it-1] + [[delta].sub.it]) / [P.sub.it-1]

Where [P.sub.it-1] = the market value of the equity of firm i at the beginning of period t (stock price times number of shares outstanding).

Expected earnings as well as stock price are often used as a deflator. Stock price is chosen because it was shown to be a theoretically superior deflator (Christie (1987)) and has been used in a number of previous studies (e.g., Easton and Zmizewski (1989), Collins and Kothari (1989)). To avoid the problem of extreme values, observations with [absolute value of (UE)]>200% are truncated to 200%.

The systematic risk (BETA) of firm i in time t, is obtained by estimating the following market model:

[R.sub.itj] = [[alpha].sub.it] + [[beta].sub.it] [R.sub.mtj] + [[epsilon].sub.itj] ............ (3)

Where [R.sub.itj] = the rate of return on firm i during month j in year t;

[R.sub.mtj] = the rate of return on market portfolio during month j in year t;

[[alpha].sub.it], [[beta].sub.it] = the intercept and the slope coefficient, respectively, from the market model.

The above model is estimated using four years (48 months) of monthly return data up to 3 months after the beginning of a fiscal year. If less than 24 monthly returns were available, the firm-month observation is excluded from the analysis.

The estimated parameters, it and it, from the market model (3) are used to calculate monthly abnormal returns (AR) as follows:

[AR.sub.itj] = [R.sub.itj] - [[alpha].sub.it] + [[beta].sub.it] [R.sub.mtj])

Where i = the firm index;

t = the year index;

j = the month index.

The monthly abnormal returns are then cumulated over twelve months up to the three months after the end of the fiscal year to get cumulative abnormal returns (CAR):

[CAR.sub.it] = [15.summation over (j=4)] [AR.sub.itj]

Where [CAR.sub.it] = the cumulative abnormal returns for firm i in year t;

[AR.sub.itj] = the abnormal returns of firm i for the jth month of year t.

To test the hypothesis that ERC's of designated firms be higher than those of non-designated firms, we estimated the following regression model:

[CAR.sub.it] = a + b[UE.sub.it] + [phi][D.sub.it][UE.sub.it] + [e.sub.it] ........... (4)

Where [UE.sub.it] = the unexpected earnings for firm i in year t,

[D.sub.it] = the dummy variable which takes a value of one if firm i is designated as a market-dominant enterprise ('designated firm') in year t, and zero if otherwise.

Test for any significant difference in ERC's between the designated firms and the non-designated firms is equivalent to testing the significance of the estimated coefficient in the regression model (4). Thus, our hypothesis can be formally stated as:

Ho: [phi] = 0, Ha: [phi] > 0

SAMPLE SELECTION AND DESCRIPTIVE STATISTICS

The sample firms examined in this study are Korean firms listed on the Korean Stock Exchange as of December 31, 1992. To be included in the sample, the firm must satisfy the following criteria: (1) Sufficient accounting data including net income and equity are available over the study period (1981-1992); (2) Monthly security returns data are available from January 1981 to December 1992; (3) Firms in banking and finance industry are excluded. Criteria (1) and (2) are imposed to ensure the data availability of accounting earnings and returns data enough to carry out empirical analyses. The firms in banking and finance industry are excluded because they tend to have different characteristics from the other firms. The above selection criteria yielded a sample of 144 firms. The number of firms designated as market-dominant enterprises by the KFTC varies over time. For example, 181 firms were designated in 1981 while 209 firms were designated in 1992.

The breakdown of sample firms by industry is shown in Table 1. The sample consists of 14 industries and there are some clustering in particular industries. For example, designated firms in foods & beverage do have 29% market share, while those in textile industries have 14.5% market share. On the other hand, the medical products industry appears to be very competitive in the sense that only 1 out of the total 17 sample firms is designated. In general, designated sample firms consist of large firms with a relatively long history and hence there may be a potential problem of survivorship bias.

Table 2 provides descriptive statistics for selected variables of the sample firms. Also reported are Wilcoxon rank test statistics for the differences in these variables between designated firms and non-designated firms. Selected variables include unexpected earnings (UE), cumulative abnormal returns (CAR), systematic risk (BETA), growth as measured by the ratio of market value to book value of equity (GROWTH), leverage as measured by the ration of total liabilities to total asset (LEVG), Tobin's Q ratio (QRATIO), return on asset (ROA), return on equity (ROE) and firm size as measured by the market value of equity (SIZE).

As expected, the average firm size of the designated firms is much greater than that of the non-designated firms: i.e., 2,366 billion Won for the designated firms ($1.57 billion at the exchange rate of 1500 Won per dollar as of August, 2004), while 528 billion Won ($0.35 billion) for the non-designated firms. This difference is statistically significant at less than 0.01 confidence level. There is no significant difference in UE and CAR between the two groups. However, mean (median) systematic risk (BETA) of the designated firms is 0.816 (0.801), which is much smaller than that of the non-designated firms. These differences are consistent with the theoretical prediction that monopoly power is negatively correlated with firm's systematic risk (Subrahmanyan and Thomadakis, 1980).

The median ROA and ROE are statistically significantly greater for the non-designated firms, which appears to be contrary to our expectation from a monopoly gain perspective. On the other hand, QRATIO of the designated firms are significantly greater than that of the non-designated firms. Higher QRATIO for the designated firms implies that the designation as a market dominant enterprise is recognized as having monopoly power in that product markets.

EMPIRICAL RESULTS

Table 3 presents the results from estimations of equation (4). We estimate equation (4) for the designated firms and the non-designated firms, as well as total sample. The results are reported for two types of samples, one for the total sample (Sample A) and the other for the reduced-sample (Sample B) excluding those firms that changed their designation status. Overall results are consistent with the theoretical prediction.

Panel A of Table 3 shows the results for the total sample (Sample A). The ERC for the designated firms is 1.159, while that of non-designated firms is 0.408. The regression coefficient ([phi]) of [D.sub.it] [UE.sub.it] in equation (4) are positive as predicted and statistically different from zero at the significance level of 0.05, supporting the Hypothesis.

Sample A may have some estimation bias because the number of the 'designated firms' is not symmetrical with that of the 'non-designated firms' each year. Thus, we delete those firms that changed their designation status during the test period and hold only those firms that consistently keep designated or non-designated status over the whole test period. The results are shown in Panel B, which are quite consistent with the results in Panel A.

Overall, these results lend empirical support to our maintained hypothesis that the ERC is a positive function of a firm's monopoly power measured by the designation as a market dominant enterprise.

In general, the above results support our hypothesis. However, the empirical estimation procedure might include the following potential problems. First, the firm size of the two sample groups is significantly different from each other, which may contaminate the estimation results. Secondly, different industry distributions of the two groups may also contaminate the estimation results. To resolve these potential problems, matched-paired sample based on firm size and industry is used. Industry is classified based on the classification by the Korean Listed Companies Association, while firm size is measured as the market value of equity. Through this procedure, 48 matched paired sample firms (total 96 firms) are selected.

Table 4 provides the estimation results of equation (4) for the matched paired sample. For the sample A including those firms who changed their designation status, the regression coefficient ([phi]) is statistically significantly positive at the significant level 0.05 as predicted. The results for sample B excluding those firms who changed their designation status are similar to those for sample A (panel B). Overall, the results for the matched paired sample also support the hypothesis that the ERC is a positive function of a firm's monopoly power.

The variables, RISK and GROWTH, have been shown to affect ERC's in the previous literature (e.g., Easton and Zmijewski, 1989 and Collins and Kothari, 1989). Thus, our findings in the previous section may be due to systematic differences between these two groups in the variables that affect the ERC's. In an attempt to investigate this possibility, we estimated the following regression model:

[CAR.sub.it] = [b.sub.0] + [[b.sub.1] + [b.sub.2][RISK.sub.it] + [b.sub.3] [GROW.sub.it] + [phi][D.sub.it]] x [UE.sub.it] + [e.sub.it] ........(5)

where [RISK.sub.it] = 1 if the systematic risk of common stock (BETA) for firm i in year t is above the sample median, and 0 if otherwise,

[GROW.sub.it] = 1 if growth rate (GROWTH) for firm i in year t is above the sample median, and 0 if otherwise.

In equation (5), the coefficient b1 of UE is predicted to be positive as a measure of usefulness of accounting earnings information. The b2 and b3 are predicted to be negative and positive, respectively.

Table 5 provides the empirical results for both total sample and matched paired sample. Each sample includes two different groups: one group includes those firms that changed their designation status while the other group does not include those firms that changed their designation status. Overall, the coefficients on RISK and GROW have their predicted signs. Furthermore, the coefficient b2 of RISK is statistically significant at the significance level of 0.05.

As expected, the estimate of the coefficient [phi] on [D.sub.it] x [UE.sub.it] are positive, which is similar to earlier results. The coefficients are statistically significant at the 0.10 significance level for sample A of total sample and sample B of matched-paired sample.

CONCLUSIONS

The purpose of this paper is to provide further evidence on the factors that affect the coefficient relating unexpected earnings and abnormal stock returns, viz., the ERC. In particular, this study examines whether a firm's monopoly power has a systematic impact on the ERC. From analytical results, we derive a theoretical prediction that the ERC is a positive function of the firm's monopoly power in its product markets.

Using a sample of 144 Korean firms listed in the Korean Stock Exchange during the period from 1986 to 1992, we empirically test this theoretical prediction. A firm's monopoly power is measured by whether or not the firm is designated as a market-dominant enterprise by the Monopoly Regulation and Fair Trade Act.

The empirical results are generally consistent with the theoretical prediction. Specifically, the ERC is higher for the designated firms than for the non-designated firms. This result is robust across different methods and samples. The results from this study may provide additional insights into the effect of the monopoly power on the ERC and the economic effect of the monopoly regulation policy in Korea.

One related issues left for future research is a time-series approach that examines the direction of changes in ERC's associated with shifts in the firm's monopoly power would provide meaningful results. For example, we can compare the ERC's over time using a sample of firms that are newly designated as a market-dominant enterprise or de-listed from the designation.

REFERENCES

Ahmed, A. (1994). Accounting earnings and Future Economic Rents. Journal of Accounting and Economics, 377-400.

Ball, R. and P. Brown (Autumn 1968). An Empirical Evaluation of Accounting Income Numbers. Journal of Accounting Research, 159-178.

Bandyopadhyay, S. (October 1994). Market Reaction to Earnings Announcements of Successful Efforts and Full Cost Firms in the Oil and Gas Industry. Accounting Review, 657-674.

Biddle, G. and G. Seow (Spring 1991). The Estimation and Determinants of Association Between Returns and Earnings: Evidence from Cross-Industry Comparisons. Journal of Accounting, Auditing & Finance, 183-232.

Choi, S. and D. Jeter (June/Sept. 1992). The Effects of Qualified Audit Opinions on Earnings Response Coefficients," Journal of Accounting & Economics, 229-247.

Christie, A. (December 1987). On Cross-sectional Analysis in Accounting Research. Journal of Accounting and Economics, 231-258.

Collins, D. and S. Kothari (July 1989). An Analysis of Intertemporal and Cross-sectional Determinants of Earnings Response Coefficients. Journal of Accounting and Economics, 143-181.

Collins, D. and W. Salatka (Fall 1993). Noisy Accounting Earnings Signals and Earnings response Coefficients: The Case of Foreign Currency Accounting. Contemporary Accounting Research, 119-160.

Dhaliwal, D. and S. Reynolds (April 1994). The Effect of the Default Risk of Debt on the Earnings Response Coefficient. Accounting Review, 412-419.

Dhaliwal, D., K. Lee and N. Fargher (Fall 1991). The Association Between Unexpected Earnings and Abnormal Security Returns in the Presence of Financial Leverage. Contemporary Accounting Research, 20-41.

Easton, P. and M. Zmijewski (July 1989). Cross-sectional Variation in the Stock Market Response to Accounting Earnings Announcements. Journal of Accounting and Economics, 117-141.

Kallapur, S. (1994). Dividend Payout Ratios as Determinants of Earnings Response Coefficients. Journal of Accounting and Economics, 359-375.

Kormendi, R. and R. Lipe (July 1987). Earnings Innovations, Earnings Persistence, and Stock Returns. Journal of Business, 323-345.

Subrahmanyan, M. and S. Thomadakis May 1980). Systematic Risk and the Theory of the Firm. Quarterly Journal of Economics, 437-451.

Thomadakis, S. (1976). A Model of Market Power, Valuation and the Firm's Returns. Bell Journal of Economics, 150-162.

Teets, W. (Autumn 1992). The Association Between Stock Market Response to earnings Announcements and Regulation of Electric Utilities. Journal of Accounting Research, 274-285.

Kyung Joo Lee, Cheju National University

Jongdae Jin, University of Maryland-Eastern Shore

Sung K. Huh, California State University-San Bernardino Table 1: Industry Classification of Sample Firms (1) Industry Designated Firms N % Foods and Beverage 18 29.0 Textiles 9 14.5 Pulp, Paper and Paper Products 1 1.6 Chemicals & Chemical Products 6 9.7 Medical Products 1 1.6 Rubber Plastic Products 4 6.5 Non-metalic Mineral Products 5 8.1 Basic Metals 2 3.2 Fabricated Metal Products 1 1.6 Machinery and Equipment 5 8.1 Radio, TV and Communication Equipment 3 4.8 Electrical Machinery and Apparatus 3 4.8 Motor Vehicles and Trailers 3 4.8 Medical, Precision and Optical Instruments, Watches 1 11.6 Total 62 100.0 Industry Non-designated Total Firms Firms N % N % Foods and Beverage 6 7.3 24 16.7 Textiles 11 13.4 20 13.8 Pulp, Paper and Paper Products 5 6.1 6 4.2 Chemicals & Chemical Products 11 13.4 17 11.8 Medical Products 16 19.5 17 11.8 Rubber Plastic Products 1 1.2 5 3.5 Non-metalic Mineral Products 9 11.0 14 9.7 Basic Metals 9 11.0 11 7.6 Fabricated Metal Products 4 4.9 5 3.5 Machinery and Equipment 1 1.2 6 4.2 Radio, TV and Communication Equipment 7 8.5 10 6.9 Electrical Machinery and Apparatus 0 0.0 3 2.1 Motor Vehicles and Trailers 2 2.5 5 3.5 Medical, Precision and Optical Instruments, Watches 0 0.0 1 0.7 Total 82 100.0 144 100.0 Table 2: Descriptive Statistics of Selected Variables Non-designated Variable Designated Firms Firms Mean S.D. Median Mean UE (2) -0.001 0.131 0.001 -0.002 CAR (3) 0.061 0.362 0.058 0.057 BETA (4) 0.816 0.239 0.801 0.936 GROWTH (5) 2.381 3.601 1.241 1.925 LEVG (6) 0.706 0.120 0.723 0.646 QRATIO (7) 1.356 0.945 1.051 1.268 ROA (8) 0.023 0.022 0.017 0.030 ROE (9) 0.076 0.056 0.068 0.083 SIZE (10) 236.59 558.53 75.24 52.78 Variable Non-designated Firms Wilcoxon Significance S.D. Median Z-value (11) (p-value) UE (2) 0.196 0.002 0.542 0.5871 CAR (3) 0.424 0.045 0.535 0.5922 BETA (4) 0.282 0.935 -7.483 0.0001 GROWTH (5) 3.051 1.001 5.036 0.0001 LEVG (6) 0.135 0.660 7.759 0.0001 QRATIO (7) 0.884 1.000 4.557 0.0001 ROA (8) 0.029 0.024 -4.334 0.0001 ROE (9) 0.186 0.073 -1.072 0.2833 SIZE (10) 84.90 26.59 12.519 0.0001 (1) Total 1,064 observations were used for 144 sample firms during 7 years (1986-1992) (2) Cumulative abnormal returns are cumulated over 12 months form April to March of the year t + 1 (3) Unexpected earnings as measured by subtracting expected earnings described by the random walk with drift model from actual earnings, and then deflated by total market value of equity at the beginning of the fiscal period. (4) Systematic risk of common stock, estimated from market model. (5) Growth as measured by the ratio of market value to book value of equity. (6) Leverage as measured by the ratio of total liabilities to total assets. (7) Tobin Q ratio = (Total Liabilities + Market value of equity) / Total assets (8) Return on total assets = Net Income / Total Assets (9) Return on equity = Net Income / Equity (10) Firm size is measured by the market value of equity (10 billion won). (11) Wilcoxon signed ranks tests statistics Table 3: Effect of Monopoly Power on the ERC [CAR.sub.it] = a + [bUE.sub.it] + [phi][D.sub.it] [UE.sub.it] + [e.sub.it] Panel A: Sample A (including firms that changed their designation status) Independent Expected Sign Designated Firms Variables Intercept ? -0.177 ** (5.340) UE + 1.159 ** (3.708) D * UE + R2 (%) 5.35 Panel B: Sample B (excluding firms that changed their designation status) Intercept ? -0.206 ** (5.599) UE + 1.406 ** (2.473) D * UE + R2 (%) 3.06 Panel A: Sample A (including firms that changed their designation status) Independent Non-designated Total Sample Firms Variables Firms Intercept -0.142 ** (5.109) -0.156 ** (7.303) UE 0.408 * (2.397) 0.418 ** (2.510) D * UE 0.744 * (2.023) R2 (%) 1.5 2.91 Panel B: Sample B (excluding firms that changed their designation status) Intercept -0.124 ** (4.144) -0.155 ** (7.303) UE 0.437 ** (2.400) 0.461 ** (2.510) D * UE 0.764 + (1.273) R2 (%) 1.82 2.14 1) Dit is a dummy variable which takes a value of one if firm i for the year t belongs to designated firms, and zero if firm i belongs to non-designated firms 2) t-values are in parenthesis. + : Significant at a = 0.10; * : Significant at a = 0.05; ** : Significant at a = 0.01. Table 4: Effect of Monopoly Power on the ERC: Matched Paired Sample based on Firm Size and Industry [CAR.sub.it] = a + [bUE.sub.it] + [phi][D.sub.it] [UE.sub.it] + [e.sub.it] Panel A: Sample A (including firms that changed their designation status) Independent Expected Designated Firms Variables Sign Intercept ? -0.191 ** (5.063) UE + 1.182 ** (3.551) D * UE + R2 (%) 6.25 Panel B: Sample B (excluding firms that changed their designation status) Intercept ? -0.188 ** (4.406) UE + 1.565 ** (2.593) D * UE + R2 (%) 4.78 Panel A: Sample A (including firms that changed their designation status) Independent Non-designated Total Sample Firms Variables Firms Intercept -0.226 ** (5.339) -0.208 ** (7.362) UE 0.279 (0.886) 0.261 (0.879) D * UE 0.947 * (2.073) R2 (%) 0.41 3.20 Panel B: Sample B (excluding firms that changed their designation status) Intercept -0.144 ** (3.109) -0.165 ** (5.256) UE 0.099 (0.289) 0.122 (0.375) D * UE 1.362 ** (1.942) R2 (%) 0.06 2.08 1) Dit is a dummy variable which takes a value of one if firm i for the year t belongs to designated firms, and zero if firm i belongs to non-designated firms. 2) t-values are in parenthesis. + : Significant at a = 0.10; * : Significant at a = 0.05; ** : Significant at a = 0.01. Table 5: Effect of Monopoly Power on the ERC: After controlling for Systematic Risk and Growth [CAR.sub.it] = [b.sub.0] + [[b.sub.1] + [b.sub.2] [RISK.sub.it] + [b.sub.3] [GROW.sub.it] = [phi][D.sub.it] * [UE.sub.it] + [e.sub.it] Total Sample Independent Variables Expected Sign Sample A Sample B Intercept ? -0.160 (7.469) ** -0.158 (6.796) ** UE + 0.723 (3.044) ** 0.866 (3.439) ** RISK * UE - -0.699 (2.107) * -0.853 (2.340) ** GROW * UE + 0.236 (0.617) 0.055 (0.138) D * UE + 0.508 (1.308) + 0.589 (0.977) R2 (%) 3.6 3.29 Matched Paired Sample Independent Variables Expected Sign Sample A Sample B Intercept ? -0.211 (7.514) ** -0.164 (5.217) ** UE + 0.834 (2.111) ** 0.581 (1.339) + RISK * UE - -1.265 (2.453) ** -0.989 (1.667) * GROW * UE + 2.468 (2.261) * 0.320 (0.231) D * UE + 0.372 (0.747) 1.105 (1.538) + R2 (%) 5.75 3.12 1) RISKit = 1 if the systematic risk of common stock for firm i in year t is above sample median, and 0 otherwise GROWit= 1 if growth (ratio of market value to book value of equity) for firm i in year t is above sample median, and 0 otherwise. 2) t-value is in parenthesis + Significant at a = 0.10; * Significant at a = 0.05; ** Significant at a = 0.0; all two tailed tests