Market Efficiency and Signaling: An Event Study Analysis for Athens Stock Exchange

Aristeidis G. Samitas Department of Business Administration, Business School, University of the Aegean, 6 Christou Lada str., Office 18, Athens 10561, Greece tel: +3 210-3689247, fax:+3 210-3689046, e-mail: asamitas@econ.uoa.gr

Dimitris F. Kenourgios Department of Accounting and Finance, Athens University of Economics and Business 5 Stadiou Str., Office 115, Athens 10562, Greece tel.: +3 210-7275012, fax: +3 210-7275010, e-mail: dkenourg@econ.uoa.gr

Proceedings of the 1st Applied Financial Economics (AFE) International Conference on Advances in Applied Financial Economics, pp. 163-175, Samos island, Greece, May 28-30, 2004

Abstract

This paper applies event study methodology to provide evidence on the semi-strong Efficient Market Hypothesis (EMH) in the Greek capital market (Athens Stock Exchange-ASE). It measures the effects of the announcements for issuing new stocks by the listed companies for the period 1998-2003. The empirical results indicate that there are findings with significant implications for the rejection of the semi-strong form efficiency. This paper also provides evidence, according to signaling theory, which suggests that the announcement of issuing new stocks gives support (total or partial) to specific hypotheses of capital structure changing. The existence of non-zero abnormal returns around the announcement periods leads to rejection of the semistrong EMH in the ASE.

JEL Classification: G14, C12 Key Words: Athens Stock Exchange, event study methodology, market efficiency, signaling.

1. Introduction

During the last decade, the Athens Stock Exchange has appeared enormous growth rates of its basic figures and contributed decisively to the economic development in Greece. Before early 1990s, local firms, which were mostly familyowned, traditionally turned to the banking system for financing, despite the presence of an organized equity market for over 120 years. After the early 1990s, this has significantly changed with firms increasingly turning to the equity market for capital. The number of listed companies grew from 145 in 1990 to 215 in 1995 and 342 in 2003. Market capitalization grew from 11,8 billion euro at the end of 1995 to 84,5 billion euro in 2003. Moreover, the number of share issues (common and preferred) grew from 298 in 1995 to 382 in 2003. Table 1 reports the development of basic figures of the ASE during the period 1990-2003.

Table 1 Basic Figures of the ASE Year Market Capitalization (on 31/12 each year, in Euro million) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 12,023 10,450 8,283 11,624 12,453 13,440 19,720 31,806 688,882 206,601 117,956 96,938 65,750 84,547 229 253 259 232 280 298 313 302 313 336 380 383 377 382 145 159 164 150 196 215 235 237 258 294 342 349 348 355 Number of share issues (common and preferred) Number of listed companies

Source: ASE, Division of Dissemination of Information, Marketing Department

Given the growth experienced by ASE during the last decade, the purpose of this paper is to test the semi-strong market efficiency of the ASE, implementing an event study analysis. Unlike earlier studies on ASE, we investigate the EMH for the ASE through a major corporate event (new stocks issuance with bonus, rights and

both) over a more recent period (1998-2003), which captures the effects of the entrance of Greece to the Economic and Monetary Union and the extremely volatile market conditions in domestic and international level. The second section of this paper presents the market efficiency theory, while the third section analyses the main methodological issues. The fourth section describes the data and presents the empirical results and finally, the paper is completed with the inscription of the basic conclusions as they came forth from our research.

2. Market Efficiency in Capital Markets

In an efficient market, the competing market participants reflect information rationally and instantaneously on prices, making past relevant information useless in predicting future prices. An efficient market should react only to new information but since this is unpredictable by definition, price changes or returns in an efficient market, cannot be predicted. Fama (1970) distinguished three types of market efficiency. A market is said to be weak form efficient if past prices are useless in predicting future prices. A market is semi-strong efficient if all publicly available information like inflation, interest rates and earnings have no predictive power. Finally, a market is strong form efficient if all information is reflected on prices, including the inside information. Under the Efficient Market Hypothesis we have (Fama, 1976),

Pt-Pt*/It-1=ut

(1)

Where It-1 is the information set available at time t-1, P* is the expected price which is based on the information test It-1, so P* is uncorrelated with ut and additionally the forecast error Pt-Pt* is uncorrelated with variables in the information set It-1 so that,

E[(Pt-Pt*)/It-1]=0

or E(ut)=0

(2)

Under the assumptions of a zero constant equilibrium return and risk neutrality. Tests for market efficiency usually examine weather the forecast error is uncorrelated with variables in the information set It-1. According to the Efficient Market Hypothesis the best predictor of tomorrows price, is todays price. Thus in a efficient market, the series of stock returns are uncorrelated with the variables in the information set It-1. According to the Efficient Market Hypothesis (EMH) a market is said to be semi-strong efficient if all publicly available information like inflation, interest rates and earnings announcements have no predictive power. The main purpose of this study is to investigate whether it is possible to predict stock price changes in ASE that could be due to the announcement of issuing new stocks. ASE sometimes works with speculative characteristics as to the movement of stock prices and with erratic and sometimes unjustifiable price swings. The latter give an indication that market prices may not, at all times, rationally reflect all information existing in the market and that possibly other factors may affect security prices.

There have been several studies so far, testing the price behaviour in the ASE and performing tests of the efficient market hypothesis (e.g. Niarchos 1972; Panas 1990; Alexakis and Petrakis, 1991; Niarchos and Alexakis 1998).

3. Methodological Issues: Event Study Analysis

Event study methodology may be interpreted as analyzing the markets reaction to events or as an empirical investigation of the relationship between security prices and economic informational events such as mergers, acquisitions, dividends announcements, issuing new stocks1 etc. The most commonly used generating process is the Single Index Market Model (SIMM), which presents a linear relationship between the returns from a given security and some market portfolio. In this paper we used ASE General Market Index at time t as the portfolio tested. According to the above the market model is,

Ri,t = a + biRp,t + ei,t

(3)

Where, Ri,t = the firms stock returns Rp,t = the return of the ASE General Market Index bi ei,t
= securities = the

specific intercept and slope coefficients

unpredictable component of the return

In our case the events tested are firms capital stock increase by issuing new stocks either with bonus, or with rights, or possibly by using both ways (complex). For each event the market model was estimated on daily data for an estimation period of 80 to 140 trading days, according to the firm, and an event period of 20 to 60 days.

The estimation period is used to forecast the expected returns for the event period. This will show us how the returns would have reacted if the events havent occurred. The estimation period is an arbitrary period of time chosen to include no security specific market events. Event period is also arbitrary selected to capture ex ante or ex post effect of the event on security price. The abnormal return is estimated by,

u it = R it i i Rmt

(4)

where, t is the event time at T+1,., T+m and i=1,.,N Averaging these prediction errors (abnormal returns) across all securities yields to the average abnormal return for day t.

ARt = N 1 u it ,
i =1

t = T+1,..,T+m

(5)

These series of returns is used to investigate whether the event had any significant effects on returns. It also provides evidence for semi-strong market efficiency, since persistent non-zero abnormal returns around an event are consistent with the hypothesis that security prices adjust quickly to reflect new information. Finally, the cumulative abnormal returns (CARs) are estimated as the sum of the average abnormal returns for the period T+m1 to T+m2, around the event and shows if the events tested had positive or negative effect on the firm:

CARm1 ,m2 =

j = m1

AR

m2

T+ j

, 1m1m2m

(6)

In order to compute the Average Cumulative Abnormal Returns for a specific event period we test the null hypothesis, CAR = 0, of no reaction to the event. The statistical significance of the CARs found were tested with t statistic,

t = CAR/ T x S (AR)

(7)

where, S (AR) is the sample standard deviation of the abnormal returns during the estimation period and T = t2 t1 + 1 (t1 is the first day of the period for which the CAR is calculated and t2 is the last). As far as our events in this study are concerned, raising capital stock by issuing stocks with bonus creates enthusiasm to the investors. These results are known as Positive Impact Hypothesis, first presented and supported by Jensen and Meckling (1976), Ross (1977), Leland and Pyle (1977) and Heinkel (1982). They argue that the unexpected rise of the firms financial leverage signals positive managerial expectations for the forthcoming profits, positive re-estimation of the stock value and final increase of the stock price. On the contrary, issuing stocks with rights leads to the fall of security returns and negative reaction of the investors. Such a reaction is known as Negative Impact Hypothesis (Myers and Majluf, 1984; Miller and Rock, 1985).

4. Data and Empirical Results

The collection of data for this empirical work was based on 67 firms listed in ASE (from all sectors) for the period 1998-2003. Each enterprise was tested for one or

two events. A total number of 103 events were investigated. The sample was identified through a research of their annual balance sheets, the monthly and quarterly ASE bulletins and publications and the firms announcements on economic press. The daily price index of primary stock market in Greece was also used in this study. The index simply represents prices. The data was converted to natural logs and covered the period from January 1998 to December 2003. Data was drawn from the Athens Stock Exchange Financial Statistics. To avoid the problem of infrequent trading in ASE, the final data of our sample consists of all the dates of successful issuing stocks (by bonus, rights and complex) with at least 80 trading days during the respective estimation period. Thus, the final sample contains 30 events of issuing stocks. For every announcement, the daily stock returns were calculated for a period of 200 days surrounding the event. These 200 days are divided into two groups from t= -200 to t= -60, the estimation period and t= -60 to t= +20, the event period. The period of our research was characterized by the entrance of Euro as high volatility and other very significant events for the world economy, like the crisis in Yugoslavia, the increase of the oil price etc. All these factors affected the stock exchange trading and the use of the market model for our study. ASE firms showed different reactions on the three events tested in this paper, due to the period of the event, the business profile of the firm, the developments plans, the profits of the last year, the perspectives and their future projects. Table 2 presents the results for all the sample firms and the Cumulative Abnormal Returns calculated for each event. In the first category (issuing stocks with bonus), 29 enterprises have positive CARs, while 7 of them show falling returns. More specifically, the positive reaction of the 29 stock returns prove their leading

10

position in ASE, their powerful investment plans, the profits and the good perspectives of the Greek capital market at the end of the 90s. The negative CARs could be due to the immature initial period of the market at the beginning of the 98 or due to the bad firms profile. These results support the Positive Impact Hypothesis, where unexpected rise of capital leverage signals positive expectations and securities prices increase. On the contrary, when enterprises issue stocks with rights the reaction of the investors is mainly negative. The results show 37 negative CARs and only 14 positive ones. The positive reaction to these events is related with the booming period of ASE in 1998 and the trust of the investors to the firms future plans. Though, the majority of the announcements in this second category lead to the Negative Impact Hypothesis. Finally, in the last category (issuing stocks with bonus and rights-complex) presented 9 firms with positive CARs and only 7 with negative. Such a reaction was due to the event of issuing stocks with bonus and it was rather expected even though the contradicting results dont help to support one of the major signaling theories.

11

Table 2 Average Cumulative Abnormal Returns (CARs) distribution and statistical significance for the three events Types of issuing new stocks (events) (1) Negative BONUS (-7,281 CARs -1,120) Positive (0,103 CARs 17,836) Negative RIGHTS (-0,032 CARs -0,270) Positive (0,432 CARs 2,802) Negative COMPLEX (0,631 CARs 1,156 Positive (0,231 CARs 12,779)
ote: * The Cumulated Abnormal Returns are in logs ** Statistical significant CARs at 5% confidence interval

CARs*

Number of firms

t (CARs)

(2)

(3) 7 29 37 14 7 9

(4) (4,64 t ( CARs) 65,07)** (7,64 t ( CARs) 106,19) ** (1,37 t ( CARs) 15,31)** (2,14 t ( CARs) 44,32) ** (1,41 t ( CARs) 36,46** (13,03 t ( CARs) 36,46) **

Table 3 demonstrates the reaction of 56 firms tested in both announcements (bonus and rights) in order to be compared. 41 of them reacted on the expected way (positive) of issuing stocks with bonus, while only 15 had a negative sign. As far as the rights are concerned 32 firms had the expected negative reaction and 24 a positive one. The statistical significance of the results was tested with t-statistic and the

12

confidence intervals in all levels confirm the results as these came forth from the event study.

Table 3 Comparative results for 56 enterprises issuing new stocks with bonus and rights. Average Cumulative Abnormal Returns (CARs) distribution and statistical significance. Types of issuing new stocks (events) (1) (0,103 CARs 5,188) BONUS (0,103 CARs 1,753 (0,432 CARs 0,909) RIGHTS (-4,009 CARs -0,032)
ote: * The Cumulated Abnormal Returns are in logs ** Statistical significant CARs at 5% confidence interval

CARs*

Expectedunexpected signs

t (CARs)

(2)

(3)

(4) (4,64 t ( CARs) 65,07)** (12,31 t ( CARs) 46,58** (15,31 t ( CARs) 44,32)** (2,14 t ( CARs) 75,93) **

Positive 41 Negative 15 Positive 24 Negative 32

13

5.

Conclusions

This paper applies event study methodology to provide evidence on the EMH in Athens Stock Exchange during the period 1998-2003. More specifically the paper tests the reaction of the firms securities returns to the announcements of issuing new stocks (bonus, rights or complex). According to the empirical evidence of Athens Stock Exchange (ASE) the three events tested have both positive and negative price effects. The signaling process can not be securely characterized from the same impact. The announcement for issuing new stocks with bonus was found to mainly support the Positive Impact Hypothesis, while new stock issuances with rights mainly support the Negative Impact Hypothesis. When complex issues are concerned the results are contradicting. On the other hand, the signaling for raising the capital stock of a firm by issuing rights is negative. The nature of the information about increasing capital stock is consistent with the international evidence and gives support to specific hypotheses of capital structure changing. The basic conclusion is the rejection of semi-strong form efficiency in ASE, which is confirmed by the existence of non-zero cumulated abnormal returns around the announcements periods. Finally, the heterogeneity of the results shows the difficulties of the empirical investigation of the price effects for new stocks issuances.

14

Notes

1. For testing semi-strong form of efficiency in ASE, see Papaioannou (1984) Niarchos and Georgacopoulos (1986) and Tsangarakis (1993). 2. Each period of our sample was tested for stationarity, normality, serial correlation, heteroskedasticity and structural stability in order to achieve a better specification of the model. In the interests of brevity, test results are not presented here but they are available upon request.

References

Alexakis, P., & Petrakis, P (1991). Analysing stock market in a small capital market. Journal of Banking and Finance, 15, 471-83. Fama, E. (1970). Efficient capital markets: a review of the theory and empirical work. Journal of Finance, 25, 383-416. Fama, E. (1976). Efficient capital markets: reply. Journal of Finance, 3, 143-45. Heinkel, R. (1982). A theory of capital structure relevance under imperfect information. Journal of Finance, 37, 1141-1150. Jensen, M. C. & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs, and capital structure. Journal of Financial Economics, 3, 305-360. Leland, ., & Pyle, D. (1977). Information asymmetries, financial structure, and financial intermediation. Journal of Finance, 32, 371-388. Miller, M. H. & Rock, . (1985). Dividend policy under asymmetric information. Journal of Finance, 40, 1013-1051.

15

Myers, S. C. & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187-221. Niarchos, N & Alexakis, C. (1998). Stock market prices, causality and efficiency: evidence from the Athens stock exchange. Applied Financial Economics, 8,167-74. Niarchos, N. (1972). The Greek Stock Market. thens: Athens Stock Exchange Publications. Niarchos, N. & Georgacopoulos, M. (1986). The effect of annual corporate profit reports on the Athens Stock Exchange: an empirical investigation. Management International Review, 26, 64-72. Panas, E. (1990). The behavior of Athens stock prices. Applied Economics, 32, 17251727. Papaioannou, G. (1984). Informational efficiency tests in the ASE. In G. Hawawini and P. Michael (Eds.) European equity markets: risk, return and efficiency. New York, Garland Publishing, Inc. Ross, S. . (1977). The determination of financial structure: The incentive signaling approach. Bell Journal of Economics, 8, 23-40. Tsangarakis N. (1993). Seasoned equity issues in the Greek stock market. Stock market reaction on the ex-rights day: a test of semi-strong form market efficiency. Spoudai, 43, 32-41.

16