Full Text - Borsa İstanbul [PDF]

Associate Editors Board Academicians Prof. Dr. Alaattin TİLEYLİOĞLU, Orta Doğu Teknik University Prof. Dr. Asaf Savaş AKAT, Bilgi University Prof. Dr. Birol YEŞİLADA, Portland State University, USA Prof. Dr. Burç ULENGİN, İstanbul Teknik University Prof. Dr. Cengiz EROL, İzmir Ekonomi University Prof. Dr. Coşkun Can AKTAN, Dokuz Eylül University Prof. Dr. Doğan ALTUNER, Yeditepe University Prof. Dr. Erdoğan ALKİN, İstanbul Ticaret University Prof. Dr. Erol KATIRCIOĞLU, Bilgi University Prof. Dr. Gülnur MURADOĞLU, Cass Business School, England Prof. Dr. Halil KIYMAZ, Rollins College, MBA, USA Prof. Dr. Hatice DOĞUKANLI, Çukurova University Prof. Dr. İhsan ERSAN, İstanbul University Prof. Dr. Kürşat AYDOĞAN, Bilkent University Prof. Dr. Mahir FİSUNOĞLU, Çukurova University Prof. Dr. Mehmet ORYAN, İstanbul University Prof. Dr. Mehmet Şükrü TEKBAŞ, İstanbul University Prof. Dr. Nejat SEYHUN, University of Michigan, USA Prof. Dr. Nicholas M. KIEFER, cornell University, USA Prof. Dr. Niyazi BERK, Bahçeşehir University Prof. Dr. Nuran Cömert DOYRANGÖL, Marmara University Prof. Dr. Oral ERDOĞAN, Bilgi University Prof. Dr. Osman GÜRBÜZ, Marmara University Prof. Dr. Reena AGGARWAL, Georgetown University, USA Prof. Dr. Reşat KAYALI, Yeditepe University Prof. Dr. Rıdvan KARLUK, Anadolu University Prof. Dr. Robert JARROW, Cornell University, USA Prof. Dr. Robert ENGLE, NYU-Stern, ABD Prof. Dr. Targan ÜNAL, İstanbul University Prof. Dr. Taner BERKSOY, Bahçeşehir University Prof. Dr. Ümit EROL, Bahçeşehir University Prof. Dr. Ünal BOZKURT, İstanbul University Dr. Veysi SEVİĞ, Marmara University Prof. Dr. Zühtü AYTAÇ, Bilkent University

Professionals Adnan CEZAİRLİ Dr. Ahmet ERELÇİN Assoc. Prof. Ali İhsan KARACAN Dr. Atilla KÖKSAL Bedii ENSARİ Dr. Berra KILIÇ Prof. Dr. Bhaskaran SWAMINATHAN Assoc. Prof. B. J. CHRISTENSEN Cahit SÖNMEZ Çağlar MANAVGAT Çetin Ali DÖNMEZ Emin ÇATANA Erhan TOPAÇ Dr. Erik SIRRI Filiz KAYA Assoc. Prof. Hasan ERSEL Prof. Dr. Kenan MORTAN Mahfi EĞİLMEZ Dr. Meral VARIŞ KIEFER Muharrem KARSLI Assoc. Prof. Dr. Ömer ESENER Reha TANÖR Serdar ÇITAK Sezai BEKGÖZ Tolga SOMUNCUOĞLU Prof. Dr. Ünal TEKİNALP Prof. Dr. Vedat AKGİRAY

Online Access: ISE REVIEW, Quarterly economics and Finance review published by the Istanbul Stock Exchange. Starting with Volume 3 No: 10 issue (year 1999), full-text articles published in the ISE Review are now available through the Internet in pdf format. Abstracts: Abstracts of all articles published in the ISE Review are available through the ISE website. The database covers all abstracts of refereed journal articles published since 1997. (1) http://www.imkb.gov.tr/yayinlar-html (2) Select: ISE Review For further informaiton, comments and suggestions please contact. Tel: (90.212) 298 21 71 The ISE Review Price and Payment Information: Hard Copy US$ 7.5 per copy (Issue No. 47) Wire transfer to T. İş Bankası Borsa Branch Account No. 1125 4599 (US$ account) Please write the name of the publication and send us a copy of the receipt. Please do not send cash. Address: İMKB (ISE) Strategy Development and Research Department Reşitpaşa Mah., Tuncay Artun Cad. Emirgan 34467 Istunbul-TURKEY Tel: +90 212 298 21 71 Fax: +90 212 298 21 89

The ISE Review Volume: 12 No: 47

CONTENTS The Effect of Foreign Portfolio Investments in the ISE: An Index-Based Study .......................................................................................... 1 Bekir Elmas Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE) ................................................... 19 Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak A Game Theoretic Approach to Model Financal Markets: Guessing Game ................................................................................................... 43 Ü. Barış Urhan & Zafer Akın

Global Capital Markets .................................................................................... 61 ISE Market Indicators ...................................................................................... 71 ISE Publication List .......................................................................................... 75

_________________________________________________________________ The ISE Review is included in the “World Banking Abstracts” Index published by the Institute of European Finance (IEF) since 1997, in the Econlit (Jel on CD) Index published by the American Economic Association (AEA) as of July 2000, and in the TÜBİTAK-ULAKBİM Social Science Database since 2005.

The ISE Review Volume: 12 No: 47 ISSN 1301-1642 © ISE 1997

THE EFFECT OF FOREIGN PORTFOLIO INVESTMENTS ON THE ISE: AN INDEX-BASED STUDY Bekir ELMAS*

Abstract Foreign shares in the market of the developing countries increased to significant levels with the financial globalization. Istanbul Stock Exchange (ISE) is the only stock market of Turkey. It is seen that the foreign investors are generally institutional investors; their shares fluctuate around 65-70% and sometimes exceed over 70 %. Depending on these two factors, it is obvious that the foreign investors have a significant effect on the ISE indexes. This study aims to determine whether a change in the foreign share have an effect on the increases and decreases of the indexes calculated in the ISE and to determine the influence. The foreign share in the ISE and 12 index variables calculated in the ISE were adopted as the data set. Consequently, it was determined that the foreign investor movements were generally not the triggering factor in the increases and decreases in the ISE-100 index and the other indexes; on the contrary, the foreign investor increased or decreased their investments by following the movements (increase/decrease) in the other indexes. I. Introduction The notion of globalization evolved with different means in the historical course in parallel with the capitalist development. Elementarily, the globalization of commerce after the Industrial Revolution until the 2nd World War, the globalization of production until 1970s and the globalization of finance after 1970s occurred respectively (Öztürk, 2001). __________________________________________________________________________________________________________________________________ *

Assist. Prof. Dr. Bekir Elmas, Atatürk University, Faculty of Economics and Administrative Sciences, Department of Business Administration, Accounting and Financing. Tel: 90 (442) 231 14 63 E-mail: [email protected] Key Words: Foreign Portfolio Investments, Index, ISE JEL Classification: G-21, G-12, G-32.

2

Bekir Elmas

The globalization of finance means the liberalization of the international capital movements and transactions of supply and demands of funds and the participants of the financial market in more appropriate environments without geographical limitations (Eroğlu, 2002). In other words, financial globalization is the removal of the boundaries dividing the national financial markets and opening of the financial markets to the international competition by removing the controls and limitations on these markets, convertibility of the markets, letting the fluctuation in exchange rates, increase in the international capital flows, increase in the role of new kind of institutional investments such as investment funds and investment partnerships in the financial markets. (http://tr.wikipedia.org/wiki/Finansal_liberalizasyon). The main requirement of the financial globalization is the Financial Integration. The Financial Integration is the removal of the difference between the rate of returns with cost of funds in different financial markets (Eroğlu, 2002). Until mid-1970s, one of the common aspects in the economy policies applying in the developing countries was the intervention in the financial markets (Galbis, 1977). In the 1980s, most of the developing countries began the process of the financial globalization by removing the limitations against the foreign investors. The developments in communication and improvements in accessing information enabled the formation of a global commercial system and the integration of the financial markets in the world (Frenkel, 2003). In Turkey which has a developing economy, the financial globalization process continued with the announcement of the decree in 24th January 1980 and the other measures taken in 1980 (Yüce, 1997). The opening of the ISE in 1986 and the issuance of Decree No.32 in August 1989 which permitted the foreign investors to invest in all kinds of securities and transfer the proceeds was a part of the financial globalization process. With financial globalization, foreign investors’ entry process especially to the developing financial markets accelerated. Entry of the foreign investors in financial markets brings along some disadvantages as well as some benefits. The benefits of the entry of foreign investors to financial markets are (Adabag and Ornelas, 2004); The market investors spread to a wider base and the liquidity and risk sharing increases in the market. While market volatility decreases in the long term, the effectiveness of the market increases.

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

3

As a consequence of market regulators’ demand for more improvements of the rules and regulations, more better quality and sufficient information flows to the market. With the appearance of more transitional markets, better distribution of the sources and the healthier markets occur. The disadvantages of the entry of foreign investors to financial markets are (Adabag and Ornelas, 2004); Foreign portfolio investments want to gain profits faster compared to the direct investments. Consequently, the sudden entries and exits to the market increase. These sudden entries and exits have negative effects on the market. Foreign investors are occasionally accused of destabilizing the market. In the developing markets, trading volume of the foreign portfolio investments is generally high. This may lead to the price pressure in the markets because of low liquidity. As well as leading to temporary ineffectiveness of the market, price pressure is related to positive feedback trading and herding behaviors which are accepted as the destabilizing reason in the market. II. Foreign Portfolio Investments in the ISE Table 1: Number of Investors and Percentages according to Nationality and Portfolio Values (TL) and Percentages in the ISE on 06.12.2010 NATIONALITY DOMESTIC FOREIGN TOTAL

INVESTMENT NUMBER 9.615.430 80.999 9.696.429

PORTFOLIO INVESTMENT VALUE(TL) NUMBER (%) 326.715.824.354 99,16 724.232.225.057 1.050.948.049.411

0,84 100

PORTFOLIO VALUE (%) 31,09 68,91 100

Source: Central Registry Agency

As shown in the Table 1, there are 9.615.430 domestic investors and 80.999 foreign investors in the ISE on 06.12.2010. While 99.16 % of the total investors are domestic investors, 0.84 % are foreign investors. These results show how low the number of foreign investors compared to domestic investors. When the portfolio values of the investors are considered, however; the situation changes in favor of the foreign investors. While 68.91 % of the total portfolio values in the ISE belong to the foreign investors, only 31.09 % belong to the domestic

4

Bekir Elmas

investors. This differentiation in the results shows that the domestic investors use their savings for individual investments while the foreign investors prefer institutional portfolio investments. Table 2: Number of Investors According to Index Type, Portfolio Values (TL) and the Foreign Shares in Portfolio Values (TL) in the ISE on 06.12.2010 INDEX TYPE NATIONAL-ALL NATIONAL ALL-100 NATIONAL-100 NATIONAL 100-30 NATIONAL-50 NATIONAL-30 NATIONAL-FINANCE 10 BANK NAT.-INDUSTRIAL NAT.-SERVICES INST. MANAGEMENT NAT.-TECHNOLOGY

INVESTOR NUMBER 996.465 429.250 858.201 417.431 752.464 679.193 627.446 354.577 565.281 292.371 367.219 46.611

INVESTOR NUMBER (%) 100,0 43,08 86,12 41,89 75,51 68,16 62,97 35,58 56,73 29,34 36,85 4,68

PORTFOLIO VALUE (TL) 158.609.708.377 23.027.132.498 135.582.575.879 31.253.981.282 120.903.665.629 104.328.594.597 92.767.925.677 60.048.504.772 39.540.303.617 25.630.194.346 32.532.109.767 671.284.738

PORTFOLIO FOREIGN VALUE. (%) PD (%) 100,0 14,52 85,48 19,70 76,23 65,78 58,49 37,86 24,93 16,16 20,51 0,42

68,91 57,90 69,89 55,92 72,55 74,07 71,77 74,26 56,73 74,05 69,14 13,47

Source: Central Registry Agency

Table 2 shows the shares of foreign investors according to the index types calculated in the ISE on 06.12.2010. According to the table, it is seen that the foreign investor shares are over 70 % in 5 indexes. These indexes are respectively 10 BANKS (74.26 %), NATIONAL-30 (74.07 %), NATIONALSERVICES (74.05 %), NATIONAL-50 (72.55 %) and NATIONAL-FINANCE (71.77 %). While the share of the foreign investors in NATIONAL-100 is 68.89 %, it is 68.91 % in NATIONAL-ALL index. The low shares of the foreign investors in NATIONAL-INDUSTRIAL (56.73 %) and especially in NATIONAL-TECHNOLOGY (13.47 %) are remarkable. These results show that the foreign investors prefer banking and services which are the fields with short-term results instead of industry and technology where to gain benefits take time. At the same time, the rates of interest probably attract attention of the foreign investors. The high shares of the foreign investors sometimes lead to excessive volatility in the whole of ISE and especially in the stocks of financial sector and banking. In Graphic 1, the foreign share in the ISE and ISE-100 index changes between 28.11.2005–06.12.2010 are shown. In order to show both variables in

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

5

the same graphic, ISE-100 index data was divided to 1000. While the foreign share fluctuates between 62.1 % and 72.6 % in this period, ISE-100 index fluctuates between 21.228 points and 71.543 points. While the fluctuation in foreign shares is around 10 %, the fluctuation in ISE-100 index reaches over 200 %. Generally, when the index increases, the foreign shares increase and when the index decreases, the foreign shares also decrease. Another remarkable point in the graphic is that the foreign share does not increase considerably in the upward trend that starts in the second month of 2009 and the domestic investors make use of this trend substantially. Graphic 1: Foreign Share in the ISE and ISE-100 Index Changes between 28.11.2005 – 06.12.2010 80 70 60 50 40 30 20 2006

2007 NATIONAL-100

2008

2009

2010

FOREIGN SHARE

III. Literature Review When the literature is analyzed, we see that mostly the influences of the financial globalization on the economy and stock exchange markets have been researched in the studies about the foreign portfolio investments. Some of these studies are stated below. Henry (2000) found out that the private investments accelerated with the financial globalization, the investments increased in the first years of the financial liberalization in almost all of 11 countries which were subject to the research and this increase exceeded 22 %.

6

Bekir Elmas

Pavabutr (2004) analyzed the influence of the foreign flow on the horizontal sections of the asset prices, market liquidity and market volatility. Consequently it was ascertained that the volatility increased in the stocks in which the foreign investors had interest and the influence of the foreign flows had the main role in the liquidity increase especially in the stocks which were highly demanded by foreign investors. Nguyen (2005) researched the relation between the globalization of stock exchange markets and conditional volatility in the developing countries such as Argentine, Brazil, Chile, Colombia, Malaysia, Mexico, Thailand and Venezuela. Consequently, he found out that the financial globalization policies did not increase the volatility and the market became silent as a result of two reasons. Kit Yi (2006) states that while the research results show there may be some positive and negative effects of the financial liberalization, it does not mean that the liberalization should not be promoted and careful applications are more appropriate. In his study, Yüce (1997) researched the influence of the foreign investments in the ISE registered stocks in 11th August 1989 and the removal of limitations on Turkish investors to invest overseas markets. Consequently, any differentiation in profits was determined between before and after 11th August 1989. Dönmez, Karataş and Kiraz (2004) compared the monthly, annually and the total profits of foreign investors on the basis of all stock markets and the profits in ISE-30 index since 1997. In their study, it was observed that the foreigners did not show a better portfolio performance than the index and the timing of their entries and exits in the market were also not so successful. At the same time, a linear relation was determined between the positive index profits in the months when the foreigners were the net buyers in the market and the decrease in index when the foreigners were the net sellers in the market The results of Baklacı’s study (2009) shows that there is a strong mutual effect between the operations of the foreign investors and the profits of the stocks. In each stock in market analysis, it was noticed that the foreign investors followed the stock profits closely and determined their operation strategies according to the profits. Similarly, another result is that the foreign investors’ operations influence the stock profits and there is a price pressure influence. In addition, it was ascertained in the analyses of the study that the foreign investors changed position in many stocks in Turkish market, especially because there was no barrier to their exits from the markets.

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

7

According to the results of Gümüş’s study (2010), there is a long-term relation (cointegration relation) between the purchases and sales of the foreign investors and the trading volume in ISE-100 index and the market prices. The purchases of the foreign investors have a positive effect on the variables and the sales have a negative effect and their explanatory power is high. Due to the longterm positive relation between the purchases of the foreign investors in the stock market and ISE-100; as the purchases of the foreign investors increase, the prices of the stocks increase and the index increase accordingly. IV. Data Set and Method The aim of this study is to determine whether the change (increase/decrease) in the foreign share have an effect on the increases and decreases of the indexes calculated in Istanbul Stock Exchange Market and to determine this influence if it exists. The series added in the study and their periods are shown in Table 3 below. In the study in which the daily data was used, the data set was provided by the Central Registry Agency. Table 3: Series Added in the Study and their Periods SERIES FOREIGN SHARE NATIONAL-ALL NATIONAL-100 NATIONAL-50

PERIOD 28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010

NATIONAL-30 NAT.-FINANCE NAT.-INDUSTRIAL NAT.-SERVICES NAT.-TECHNOLOGY

28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010 28.11.2005 – 06.12.2010

28.11.2005 – 06.12.2010

SERIES NATIONAL ALL -100 NATIONAL 100-30 10 BANK INSTITUTIONAL MANAGEMENT

PERIOD 02.01.2009 - 06.12.2010 02.01.2009 - 06.12.2010 04.01.2010 - 06.12.2010 31.08.2007 - 06.12.2010

Note: As the data of the foreign shares has been recorded in CRA since 28.11.2005, the period of the application begin in this date. The indexes calculated after 28.11.2005 was added in the application since the date they began to be calculated.

In an econometric model estimated with time series data, variables must be stationary. An equation estimated with non-stationary time series might be causes spurious regression. Whether a regression reflects a true relationship or not is closely related to stationary of the variables (MacKinnon, 1991; Gujarati, 1995). The first test to research the existence of the unit root in a time series and the most accepted test in the literature is Dickey-Fuller (DF) test (1979, 1981)

8

Bekir Elmas

(Sevüktekin and Nargeleçekenler, 2005). Dickey-Fuller test is not used if the error terms include autocorrelation. The autocorrelation in the error term can be removed by using the delayed values of the time series. Dickey-Fuller developed a new test which adds the delayed value of the dependent variable to the model as an independent variable. This is the Augmented Dickey-Fuller (ADF) test which was used in this study (Enders, 1995; 225). In their studies, Philips-Perron developed a new unit root test with a non-panoramic approach by generalizing the unit root test suggested by Dickey-Fuller (Phillips and Perron 1988: 335-346; Sevüktekin and Nargeleçekenler, 2007: 4). However, when the autocorrelations of the error term are highly negative in this approach, Philips-Perron test encounters with a sample error of the error term. Anyhow if it is corrected, Philips-Perron test is stronger than ADF test (Schwert, 1989: 147-160; Sevüktekin and Nargeleçekenler, 2007: 4). In this study, ADF (Augmented Dickey-Fuller) and PP (Phillips-Perron) unit root tests were applied on the series to choose the right model. These variables should be cointegrated, in other words, they should move together in long term; so that the foreign share change and the index profits can be explained. If two instable time series are cointegrated in the same level, there may be a cointegration relation between them. The cointegration relation formed with instable series is meaningful and does not include spurious regression (Tarı, 2009). The cointegration relation between the series was brought to the literature by Engle-Granger. Although Engle-Granger (1987) test explains the cointegration relations, it has some deficiencies. In order to make up these deficiencies, Johansen (1988, 1995) and Johansen-Juselius (1990) developed new cointegration tests. In this study, Engle-Granger (1987) cointegration test and the cointegration tests developed by Johansen (1988, 1995) and JohansenJuselius (1990) were applied. Cointegration relation shows the long-term relation between the variables. In order to determine the short-term relation between the variables, Error Correction Model (ECM) is formed. With this model, it is determined how many deviants in the long-term series can be corrected in short term. In the study, the causal relation between ISE indexes and the foreign share and the direction of this relation was researched with the help of Granger causality test. Granger causality analysis tests whether there is a relation between the current and previous value of the variables and the direction of the relation if it exists.

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

9

In order to test Granger causality between the variables Gt and YPt, primarily VAR (Vector Autoregressive) model is estimated as shown below. T11

T12

i 1

j 1

G t   12    11 i G t  i    12 j YPt  j   12 t

YPt   22 

T 21

 i 1

21 i

YPt  j 

T22

 j 1

22 j

G t  i   12 t

(1)

(2)

Where, T selected lag length , α and β show the parameters to be estimated and εt shows white noisy error terms. In Equality (1), in order to test whether there is Granger causality from the variable YPt to Gt, the null hypothesis is defined as H0 : β12j = 0 and Wald and chi-square test is applied to this hypothesis. If the null hypothesis is rejected, namely at least one of the delayed parameters is different from zero, it is accepted that there is Granger causality from the variable YPt to Gt .The same method is also used in Equality (2), in order to test whether there is Granger causality from the variable Gt to YPt. Between the variables with cointegration relation, at least one way Granger causality relation is required. While the equations numbered (1) and (2) are used to determine Granger causality between the series without cointegration relation, Granger causality test (3) and equations based on the Error Correction Model (4) are used for the cointegrated variables. T11

T12

i 1

j 1

Gt  12   11i Gt i   12 j YPt  j  ECt 1   12t

T21

T22

i 1

j 1

YPt   22    21i YPt  j    22 j Gt i   EC t 1   12 t

(3)

(4)

In the regression equations numbered (3) and (4), Δ shows the first-degree difference operator. The coefficients of the delayed error terms (µ and λ) in the form of ECt-1 show the speed of short-term equilibration.

10

Bekir Elmas

V. Test Results Table 4: Definitive Statistics of the Series SERIES

Max.

Min.

Std. Deviation 2.44 10.177 12.696 10.410 16.632 10.339 13.136 17.340 15.564 7.336 6.492

Skewness

Kurtosis

Jarque-Bera (Prob)

FOREIGN SHARE 72.6 62.1 0.26 0.22 48 (0.00) NATIONAL-ALL 70.572 21.258 0.23 2.9 12 (0.00) NATIONAL ALL-100 70.728 25.825 -0.33 2.0 27 (0.00) NATIONAL-100 71.543 21.228 0.15 2.8 6.6 (0.04)* NATIONAL 100-30 81.741 25.739 -0.46 2.0 37 (0.00) NATIONAL-50 71.291 21.303 0.16 2.8 7.8 (0.02)* NATIONAL-30 91.249 27.062 0.22 2.8 12 (0.00) NAT.-FINANCE 111.849 29.355 0.08 2.8 2.8 (0.24)* 10 BANK 168.132 104.753 0.37 2.4 8.5 (0.012)* NAT.-INDUSTRIAL 51.333 16.845 -0.20 2.9 8.6 (0.015)* NAT.-SERVICES 44.884 16.937 0.49 2.4 71 (0.00) INSTITUTIONAL 59.826 19.131 11.196 -0.35 2.1 46 (0.00) MANAGEMENT NAT.-TECHNOLOGY 18.630 4.045 3.617 0.01 2.2 36 (0.00) Note: H0 = Series distributed normally. *According to the 1% significance level, H0 hypothesis cannot be rejected.

In daily language, volatility means the change in any event in time. In economy literature, on the other hand, volatility has a more official meaning and defines the change and the standard deviation in the random time series (K. Öztürk, 2010: 61). Financial time series are generally more volatile according to macroeconomic time series and their volatility change more in time. While rapid increases follow sharp decreases in these series, sharp decreases also follow the rapid increases. In Table 4; when the definitive statistics of ISE index series are analyzed, it is seen that the index with highest volatility is NATIONALFINANCE index, and it is followed respectively by NATIONAL 100-30, 10 BANK, NATIONAL ALL-100, INSTITUTIONAL MANAGEMENT, NATIONAL-100, NATIONAL-50, NATIONAL ALL, NATIONALINDUSTRIAL and NATIONAL-SERVICES. The index with lowest volatility is NATIONAL-TECHNOLOGY index. According to Jarque-Bera test statistics, NATIONAL-100, NATIONAL-50, NATIONAL-FINANCE, 10 BANK and NATIONAL-INDUSTRIAL indexes distribute normally at 1 % significance level. In these series, the coefficient of the volatility is near to 0 and the coefficient of the kurtosis is near to 3. FOREIGN SHARE series, on the other hand, does not distribute normally. This serial is a little kurtic and right aligned. Table 5 shows the results of ADF and PP unit root tests. In Table 5, it is seen that profits of all series are stable at 1% significance level in 2 unit root tests.

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

11

Table 5: Unit Root Test Results FOREIGN SHARE AND ISE INDEXES FOREIGN SHARE NATIONAL-ALL NATIONAL ALL -100 NATIONAL-ALL NATIONAL 100-30 NATIONAL-50 NATIONAL-30 NATIONAL-FINANCE 10 BANK NATIONAL--INDUSTRIAL NATIONAL--SERVICES INST. MANAGEMENT NAT.-TECHNOLOGY CRITICAL TEST VALUES a 1% significance level b 5% significance level c 10% significance level

ADF UNIT ROOT TEST LEVEL PROFIT -1.46 (1)[.84] -32.0 (0)[.00]a -0.94 (0)[.95] -34.4 (0)[.00]a -2.38 (0)[.39] -20.3 (0)[.00]a -1.00 (0)[.94] -34.4 (0)[.00]a -1.92 (0)[.64] -20.8 (0)[.00]a -1.06 (0)[.93] -34.5 (0)[.00]a -1.15 (0)[.92] -34.7 (0)[.00]a -1.06 (0)[.93] -34.6 (0)[.00]a -2.94 (0)[.15] -15.4 (0)[.00]a -0.67 (0)[.97] -32.9 (0)[.00]a -1.75 (0)[.73] -35.0 (0)[.00]a -0.81 (0)[.96] -26.3 (0)[.00]a -0.37 (1)[.99] -32.5 (0)[.00]a

PP UNIT ROOT TEST LEVEL -1.60 (13)[.79] -32.5 (12)[.00]a -1.02 (16)[.94] -34.4 (17)[.00]a -2.42 (5)[.37] -20.4 (9)[.00]a -1.06 (15)[.93] -34.5 (17)[.00]a -1.96 (6)[.62] -20.8 (8)[.00]a -1.11 (15)[.93] -34.5 (17)[.00]a -1.21 (17)[.91] -34.7 (17)[.00]a -1.11 (14)[.93] -34.6 (16)[.00]a -2.94 (0)[.15] -15.5 (4)[.00]a -0.76 (14)[.97] -32.9 (17)[.00]a -1.70 (18)[.75] -35.1 (21)[.00]a -0.82 (9)[.96] -26.2 (10)[.00]a -0.38 (6)[.99] -32.5 (9)[.00]a

-3.965363 -3.413390 -3.128731

Note: The numbers in parentheses show the delay distances chosen for ADF unit root test according to Schwarz Information Criteria and bandwidths determined according to Newey-West which uses Bartlett kernel for PP unit root test. The values in the square brackets show the p-values of ADF and PP statistics.

At this stage of the study, research was carried out whether there is a longterm relation between variables with Engle-Granger (1987) and Johansen (1988, 1995) and Johansen-Juselius (1990) cointegration tests. Engle-Granger cointegration test results are shown in Table 6. According to test results, it could not be rejected the null hypothesis that there is no any relationship between any variable pairs. These results show that there is no cointegration vector in all variable pairs. Secondly, the existence of the cointegration relationship was investigated with Johansen cointegration test. In the first stage of Johansen cointegration test, optimal lag length was determined according to Akaike Information Criterion. The results of cointegration test are shown in Table 7. The results of Johansen cointegration test support the results of EngleGranger cointegration test. By this way, the error correction model estimations are not necessary anymore, as there is no long-term cointegration relation between variable pairs.

12

Bekir Elmas

Table 6: Engle-Granger Cointegration Test Results VARIABLE PAIR NATIONAL-ALL FOREIGN SHARE NATIONAL-ALL-100 / FOREIGN SHARE NATIONAL-100/ FOREIGN SHARE NATIONAL-100-30/ FOREIGN SHARE NATIONAL-50/ FOREIGN SHARE NATIONAL-30/ FOREIGN SHARE NATIONAL-FINANCE/ FOREIGN SHARE NATIONAL-10 BANK/ FOREIGN SHARE NATIONAL-INDUSTRIAL/ FOREIGN SHARE NATIONAL-SERVICES/ FOREIGN SHARE INSTITUTIONAL MANAGEMENT/ FOREIGN SHARE NATIONAL-TECHNOLOGY/ FOREIGN SHARE

Engle–Granger T-Statistic -1.45 -2.67 -1.15 -2.73 -1.21 -1.30 -1.14 -2.88 -0.90 -2.10 -1.59 -0.38

Critical Test Values (10%) -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087 -3.087

Table 7: Johansen Cointegration Test Results VARIABLE PAIR

Path Testing

Probabilities

NATIONAL-ALL FOREIGN SHARE

6.50 1.22 11.2 1.14 6.48 1.24 11.5 1.09 6.53 1.35 6.42 1.46 7.40 1.91 7.18 0.80 7.95 1.33 3.08 8.71 10.8 2.02 4.74 0.72

0.64 0.27 0.20 0.29 0.64 0.27 0.18 0.30 0.63 0.24 0.65 0.23 0.53 0.17 0.56 0.37 0.47 0.25 0.96 0.74 0.22 0.16 0.83 0.40

NATIONAL-ALL-100 FOREIGN SHARE NATIONAL-100 FOREIGN SHARE NATIONAL-100-30 FOREIGN SHARE NATIONAL-50 FOREIGN SHARE NATIONAL- 30 FOREIGN SHARE NATIONAL- FINANCE FOREIGN SHARE NATIONAL- 10 BANK FOREIGN SHARE NATIONAL-INDUSTRIAL FOREIGN SHARE NATIONAL-SERVICES FOREIGN SHARE INSTITUTIONAL MANAGEMENT FOREIGN SHARE NATIONAL-TECHNOLOGY FOREIGN SHARE

Max. Eigen Value Test 5.28 1.22 10.1 1.14 5.24 1.24 10.4 1.09 5.17 1.35 4.96 1.46 5.49 1.91 6.38 0.80 6.62 1.33 2.97 8.71 8.78 2.02 4.02 0.72

Probabilities 0.71 0.27 0.21 0.29 0.71 0.27 0.19 0.30 0.72 0.24 0.75 0.23 0.68 0.17 0.57 0.37 0.53 0.25 0.95 0.74 0.30 0.16 0.86 0.40

Note: While H0: r = 0, r ≥ 1 hypothesis was tested with 1. row values, H0: r = 1, r ≥ 2 hypothesis was tested with 2. row values.

Granger causality analysis results which are determined on the basis of VAR (Vector Autoregressive) model are shown in Table 8. According to analysis results, the null hypothesis that there is no Granger causality relation from the indexes of NATIONAL-ALL, NATIONAL-ALL-100, NATIONAL100, NATIONAL-100-30, NATIONAL-50, NATIONAL-30, NATIONALFINANCE and NATIONAL-SERVICES to FOREIGN SHARE at the

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

13

significance level of 1% and from the indexes of NATIONAL-10 BANK, NATIONAL-INDUSTRIAL and INSTITUTIONAL MANAGEMENT at the significance level of 5%. The null hypothesis that there is no Granger causality relation at the significance level of 10% is from FOREIGN SHARE to NATIONAL-ALL, NATIONAL-INDUSTRIAL and INSTITUTIONAL MANAGEMENT indexes. This way, at least one way Granger causality relation was determined between all variable pairs (11 variable pairs) except NATIONAL-TECHNOLOGY/FOREIGN SHARE. While Granger causality relation is bi-directional in 3 variable pairs, it is unidirectional from index to the foreign share in 8 variable pairs. Table 8: Granger Causality Analysis Results VARIABLE PAIR NATIONAL-ALL FOREIGN SHARE NATIONAL-ALL-100 FOREIGN SHARE NATIONAL-100 FOREIGN SHARE NATIONAL-100-30 FOREIGN SHARE NATIONAL-50 FOREIGN SHARE NATIONAL-30 FOREIGN SHARE NATIONAL-FINANCE FOREIGN SHARE NATIONAL-10 BANK FOREIGN SHARE

p 4 1 5 1 5 5 5 1

NATIONALINDUSTRIAL FOREIGN SHARE

2

NATIONALSERVICES FOREIGN SHARE

5

INSTITUTIONAL MANAGEMENT FOREIGN SHARE

2

NATIONALTECHNOLOGY FOREIGN SHARE

2

Ho NATIONAL-ALL≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-ALL NATIONAL-ALL-100 ≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-ALL-100 NATIONAL-100 ≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-100 NATIONAL-100-30 ≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-100-30 NATIONAL-50 ≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-50 NATIONAL-30≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-30 NATIONAL-FINANCE≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-FINANCE NATIONAL-10 BANK≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONAL-10 BANK NATIONAL-INDUSTRIAL≠ > FOREIGN SHARE FOREIGN SHARE ≠ > NATIONALINDUSTRIAL NATIONAL-SERVICES≠ > FOREIGN SHARE

Test Results 22.9 (0.00)* 8.40 (0.08)*** 8.95 (0.00)* 0.39 (0.53) 29.8 (0.00)* 7.56 (0.18) 9.24 (0.00)* 0.43 (0.51) 30.9 (0.00)* 7.49 (0.19) 31.8 (0.00)* 7.35 (0.20) 31.64 (0.00)* 8.03 (0.15) 4.49 (0.03)** 0.15 (0.70)

FOREIGN SHARE ≠ > NATIONAL-SERVICES

2.57(0.76)

INSTITUTIONAL MANAGEMENT≠ > FOREIGN SHARE FOREIGN SHARE ≠ > INSTITUTIONAL MANAGEMENT NATIONAL-TECHNOLOGY≠ > FOREIGN SHARE YABANCI PAYI ≠ > ULUSAL-TEKNOLOJİ

7.11 (0.03)** 5.11(0.08)*** 26.9 (0.00)*

8.08 (0.02)** 5.65 (0.06)*** 3.16 (0.21) 4.07 (0.13)

Note: In VAR(p) model, delay distances are determined according to Akaiki Information Criteria. The numbers in parentheses show p value and meaningfulness * at the significance level of 1%, ** at the significance level of 5% , ***at the significance level of 10%.

14

Bekir Elmas

VI. Conclusion Since 1970s, especially developing countries started financial globalization process by removing the barriers against the international capital. With the start of this process, capital-rich investors began to invest in the markets with high profit rates getting beyond the boundaries. Due to the international capital flows, the capital pulled away from its national characteristics and acquired an international dimension. This led to the increase in direct and indirect investments especially in the developing countries where the profit potentials were high. Turkey, as a developing country, has also got its share from these changes and developments. Along with the financial globalization, foreign shares in the developing country markets have increased to significant levels. The shares of foreign investors in the Istanbul Stock Exchange Market fluctuate around 65-70 % and sometimes exceed 70 %. At the same time, the foreign investors in the ISE generally invest institutionally. Due to these two factors, it is obvious that the foreign investors have a significant influence on the ISE indexes. This study aimed to determine whether the change (increase/decrease) in the foreign share have an effect on the increases and decreases of the indexes calculated in the Istanbul Stock Exchange Market and to determine this influence if it existed. In the study, the foreign share in the ISE and 12 index variables calculated in the ISE was used as data set. The existence of the cointegration relation between the variables was researched both with Granger and Johansen cointegration tests. According to the results of both tests, any cointegration relation between the variables was discovered. Therefore, there is no long-term joint-movement between the foreign share and indexes. The foreign share generally increases with the increase in the exchange market and decreases with the decrease in the exchange market. However; while the fluctuation of the foreign share is around 10 %, the fluctuation in the index sometimes exceeds 200 %. According to Granger causality analysis between 12 indexes used in the study and the foreign share, a Granger causality relation was not determined between only one index and foreign share However, unidirectional Granger causality relation from 8 indexes to the foreign share and bi-directional Granger causality relation between 3 indexes and the foreign share were determined. These results show that the foreign investors’ movements generally do not direct the moves of the ISE-100 and other indexes; on the contrary the foreign investors increase or decrease their investments by following the movements

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

15

(increase/decrease) in the ISE-100 and other indexes. That is, foreign investors are in a position following the movements and not starting the movements. This way, the indexes increase more due to the purchases of the foreign investors during an upward trend, decrease more during an downward trend.

16

Bekir Elmas

References Adabag C.; J. R. H. Ornelas, “Behavior and Effects of Foreign Investors on Istanbul Stock Exchange”, 2004, ss. 3-4. Available at SSRN: http://ssrn.com/abstract=656442 (Erişim Tarihi: 15.11.2010). Brooks C., Introductory Econometrics for Finance, Cambridge University Press, Eighth Printing, 2007. Chau Kit Yi, “Do Foreign Investments Affect Stock Markets -The Case of Shanghai Stock Market-”, 2006. http://libproject.hkbu.edu.hk/ trsimage/hp/03017885.pdf Dickey, D. A.; W. A. Fuller, “Distribution of the Estimator for Autoregressive Time Series with a Unit Root”, Journal of the American Statistical Association, 74, 1979, ss. 427-431. Dickey, D. A.; W. A. Fuller, “Likelihood Ratio Statistics for Autoregressive Time Series with a Unit Root”, Econometrica, 49, 1981, ss. 1057-1072. Dönmez Ç. A.; A. Karataş; F. Kiraz, “İstanbul Menkul Kıymetler Borsası’ndaki Yabancı Yatırımcıların Performans Analizi”, İktisat İşletme ve Finans Dergisi, 19, 2004, ss. 95-104. Enders W., Applied Econometric Time Series, Jonh Wiles and Sons, Canada, 1995, ss. 225. Engle, R. F.; C. W. J. Granger, “Co-integration and Error Correction: Representation, Estimation, and Testing”, Econometrica, 55, 1987, ss. 251-276. Eroğlu N., “Finansal Küreselleşme: Devletin Düzenleyici Rolü Üzerine Etkileri içinde Küreselleşme: İktisadî Yönelimler ve Sosyo-politik Karşıtlıklar”, Derleyen: Alkan Soyak, Om Yayınevi, İstanbul, 2002, ss. 18. Frenkel R., “Globalization and Financial Crises in Latin America”, 18 (207), 2003, ss. 41-56. Galbis V., “Financial Intermediation and Economic Growth in Less Developed Countries: A Theoritical Approach”, Journal of Development Studies, 13, 1977, ss. 58-72. Gujarati D. N., Basic Econometrics, McGraw-Hill, Third Edition, New York, 1995. Henry, P. B., “Stock Market Liberalization, Economic Reform and Emerging Market Equity Prices”, Journal of Finance, 55, 2000, ss. 529-564. http://tr.wikipedia.org/wiki/Finansal_liberalizasyon (Erişim Tarihi: 15.11.2010) http://www.imkb.gov.tr/Data/StocksData.aspx (Erişim Tarihi: 06.12.2010) http://www.mkk.com.tr/MkkComTr/tr/bilgimerkezi/ist_g.jsp (Erişim Tarihi: 06.12.2010). Johansen S., Likelihood-Based Inference in Cointegrated Vector Autoregressive Models, Oxford University Press, New York, 1995. Johansen S.; K. Juselius, “Maximum Likelihood Estimation and Inference on Cointegration- with Applications to the Demand for Money,” Oxford Bulletin of Economics and Statistics, 52 (2), 1990, ss. 169–210.

The Effect of Foreign Portfolio Investments on the ISE: An Index-Based Study

17

Johansen, S., “Statistical Analysis of Cointegration Vectors”, Journal of Economic Dynamics and Control, 12, 1988, ss. 231-254. Lukauskas A.; S. Minushkin, “Explaining Styles of Financial Market Openning in Chile, Mexico, South Korea and Turkey”, International Studies Quarterly, 44, 2000, ss. 695-723. MacKinnon J., “Critical Values for Cointegration Tests”, In: Engle R. and C. Granger (eds) Long-Run Economic Relationships: Readings in Cointegration, Oxford University Press, Oxford, 1991, ss 267–276. Nguyen D. K., “Liberalization of Emerging Equity Markets And Volatility”, 2005. http://www.efmaefm.org/efma2005/papers/182-nguyen_paper .pdf (Erişim Tarihi: 15.11.2010) Öztürk K., “Döviz Kuru Oynaklığı ve Döviz Kuru Oynaklığının Faiz Oranı Oynaklığıyla Olan İlişkisi: Türkiye Örneği”, Uzmanlık Yeterlilik Tezi, Ankara, 2010. http://www3.tcmb.gov.tr/kutuphane/TURKCE/ tezler/kevserozturk.pdf (Erişim Tarihi: 15.11.2010) Öztürk N., “Finansal Küreselleşmenin İktisat Politikalarına Etkisi ve Türkiye”, (Basılmamış Doktora Tezi), Marmara Üniversitesi, Sosyal Bilimler Enstitüsü, 2001, ss. 26. Pavabutr P., “Foreign Portfolio Flows and Emerging Market Returns: Evidence from Thailand”, Dissertation, The University of Texas at Austin, August 2004. Phillips, P. C. B.; P. Pierre, “Testing for Unit Roots in Time Series Regression”, Biometrika, 75, 1988, ss. 335-346. Schwert, G. W., “Tests for Unit Roots: A Monte Carlo Investigation”, Journal of Business and Economic Statistics, 7, 1989, ss. 147-160. Sevüktekin M.; M. Nargeleçekenler, “Türkiye'de İMKB ve Döviz Kuru Arasındaki Dinamik İlişkinin Belirlenmesi”. http://web.inonu.edu.tr /~eisemp8/bildiri-pdf/sevutekin-nargelecekenler.pdf (Erişim Tarihi: 15.05.2010), s. 4. Sevüktekin, M.; M. Nargeleçekenler, Zaman Serileri Analizi, Nobel Yayın Dağıtım, Ankara 2005, ss. 304. Tarı R., Ekonometri, 6. Baskı, Kocaeli Üniversitesi Yayını, 2009. Yüce, A., “Türkiye’de Liberalizasyon Hareketlerinin Hisse Senedi Fiyatlarına Etkisi”, İMKB Dergisi, 1 (4), 1997, ss. 1-13.

The ISE Review Volume: 12 No: 47 ISSN 1301-1642 © ISE 1997

STOCK PRICE REACTIONS TO MERGER ANNOUNCEMENTS: EVIDENCE FROM ISTANBUL STOCK EXCHANGE (ISE) Berna KIRKULAK ULUDAĞ* Özlem DEMİRKAPLAN GÜLBUDAK**

Abstract The stock price performances of the ISE listed non-financial firms are examined before and after the merger announcements by employing cumulative average abnormal returns (CAARs) from 1997 to 2006. In Turkey, merger and acquisitions (M&As) intensified in particular after the 2001 financial crisis. Consistent with the previous studies, the findings suggest that the stock prices prior to merger announcements are likely to increase. However, this positive effect disappears following the merger announcements and the post-merger stocks underperform in the long-run. Investors’ sentiment (optimism) explains the post-merger underperformance anomaly. Furthermore, a benchmark of control firms is constructed according to size and growth potential. The findings reveal the fact that bidding firms perform better than non-merged firms in the long-run. I. Introduction Merger waves have been the subject of intense interest by many researchers. There are several factors that push firms to merge or to acquire. Mitchell and Mulherin (1996) stated that mergers are the outcomes of financial crises, changes in technology and economy and regulatory environment. In particular, the periods of financial crises are characterized by liquidity problems and this results in more mergers and acquisitions (M&As). __________________________________________________________________________________________________________________________________ *

Berna Kırkulak Uludağ, Department of Business Administration, Faculty of Business, Dokuz Eylül University, 35160, Buca, İzmir, Turkey. Tel: +90 (0) 232 412 8244 Fax:+90 (0)232 453 5062 E-mail:[email protected] ** Özlem Demirkaplan Gülbudak, Department of Business Administration, Faculty of Business, Dokuz Eylül University, Masters in Business Administration. E-mail:[email protected] JEL Classification: G12, G14. Keywords: merger, acquisition, price reaction, ISE.

20

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

A merger is assumed to enhance wealth if the value of the bidder and target firms increase following the merger announcement. Several researchers attempted to examine merger gains through stock returns. There is a growing concern among researchers about the pre-and post-merger stock performance of merger announcements. In the literature, some empirical studies document positive stock price reactions prior to merger announcements (Dennis and McConnell, l986). However, some studies show that positive short-run reactions to announcements are fully reversed in the post-merger period (Loughran and Vijh, 1997; Mitchell and Stafford, 2000). Overall, the large majority of the studies suggest that the positive effects of mergers disappear in the long-run. In this regard, some researchers attempted to explain overvaluation of stocks before and around the merger announcements. Among them, Shleifer and Vishny (2003) and Rhodes-Kropf and Viswanathan (2004) argued that stock markets misvaluations push bidding firms to make profits. The financial markets are inefficient and therefore some firms are mispriced. Bidding firms are rational and they take the advantage of market misvaluation until the market realizes. Another approach to stock market overvaluation is investor’s sentiment (optimism). The divergence of opinion between optimistic and pessimistic investors explains the low stock returns in the long-run. As the information about the true value of the firm increases over time, the long-run returns decrease. This approach has been first introduced for the long-run stock returns after the public offerings (Ritter, 1991; Loughran and Ritter; 1995). Rosen (2006) applied this view to merger markets and provided evidence that stock prices are more likely to increase when a merger is announced. However, long-run stock returns are negative for mergers announced. The negative post-merger stock returns are attributed to M&As being anticipated prior to announcements. This view was supported by Kaplan (2000) who argued that abnormal returns around and after the merger announcements may not accurately reflect the value gains. The reason is that M&As are largely anticipated so that the positive effects on firm value do not show themselves on the announcements dates. In particular, speculations and rumors concerning the forthcoming M&As become influential, hence pre-merger returns become positive. The Turkish economy experienced a large wave of M&As during 2000s. Most of these deals were different from the hostile takeovers. The consecutive economic crises of the 1990s and the ongoing deregulation of Turkish financial markets have motivated many changes in corporate structure. The recent

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

21

financial crises led to a broad decline in the equity prices and therefore stimulated larger M&A activities. Under the high inflation era of the 1980s and 1990s, investors had become accustomed to high returns on their investments due to high interest rates. After the implementation of the disinflationary program, the inflation rate declined sharply and this led numerous investors to seek new investment opportunities. M&A activities became more popular during the low inflation period. The purpose of this study is to measure the stock price reactions to M&A announcements. This paper examines the stock price performances both before and after the biding firms’ merger announcements. The cumulative average abnormal returns (CAAR) method is employed to 37 ISE non-financial firms from 1997 to 2006. The contribution of this paper is in two parts. First, the current study provides empirical evidence of stock price reactions to M&As before and after the merger announcements. As it is difficult to avoid M&As rumors, the stock price returns prior to the merger announcements are calculated. In order to mitigate the merger rumors on stock returns, stock price reactions are measured at different periods. In order to achieve this task, three time lines are constructed; M&A rumor, declaration and approval dates. The stock price reactions are examined 12 months prior to merger announcements and the postmerger stock performances are calculated 12 months after the merger announcements. Second, this paper compares M&A announcement effects on stock returns of merged firms versus non-merged firm In order to understand the impact of M&A announcements on stock returns, unmerged control firms are selected on the basis of size and growth potential. We compared short-term and long-term stock returns of merged firms with those of non-merged firms. This paper is one of the few studies available that examine Turkish mergers and acquisitions. To the authors’ knowledge, there has been yet no study using the benchmark methods to compare the stock returns of merged firms with those of non-merged firms. The evidence presented in this paper supports the investor sentiment (optimism) hypothesis. The findings suggest that post-merger stock returns are generally lower than pre-merger stock returns. The results show that investors earned significant capital gains during the pre-merger period. However, stock returns started to decline slightly after M&A announcements and these declines are statistically significant. High pre-merger stock returns posit that M&As are largely anticipated prior to official announcements and this situation is reflected in high pre-merger stock returns.

22

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

The remainder of this paper is structured as follows: Section 2 reviews the literature and gives information about M&As in Turkey. Section 3 describes sample selection and the limitations of the study. Section 4 presents methodology. Section 5 reports the empirical results and Section 6 concludes. II. Literature Some researchers have focused on the value creation effect of M&As and they tried to measure whether or not mergers create value. Most of the event studies examined the stock returns of bidding firms rather than target firms. In their study, Wong and Cheung (2009) examined the effect of M&A announcement on the stock returns of bidding and target firms in Hong Kong, Taiwan, China, Singapore, South Korea and Japan from 2000 through 2007. Their findings revealed the fact that while the stock price reaction to M&A announcement of bidding firms is positive, market reaction to M&A announcement of target firms is negative. Studies examining value creation effect of mergers have documented that short-run stock performances are positive surrounding the M&A announcements. Stock prices react quickly to new information and incorporate any changes in value. Keown and Pinkerton (1981) examined merger announcements in a sample of 194 US firms and found significant positive abnormal returns 12 days prior to takeover announcements. They attributed this finding to illegal trading on inside information prior to the takeover. Similarly, in their study, Dennis and McConnell (l986) documented significant abnormal returns prior to merger announcements. The second measure of value creation is the long-run performance of the bidding firms during the post-merger period. A growing body of stock performance studies following the merger announcements are contradictory. While Healy, Palepu and Ruback (1992), Ramaswamy and Waegelein (2003) found an increase in the performance of the firms involved in mergers, Ravenscraft and Scherer (1989), Dickerson, Gibson and Tsakalotos (1997) Agrawal and Jaffe (2003), Rosen (2006) found decrease in the stock performance after merger activities. The diversity of the results about post-merger stock returns can be explained by sample and time period selection and stock return calculation methods. The longer the post-merger period analyzed, the greater the probability of the other factors that affect long-run stock returns.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

23

Overall, post-merger studies suggest that firms with high pre-merger stock returns exhibit lower or negative long-run abnormal stock performances. Market misvaluation and investment sentiment approaches have gained significant importance over years while explaining the post-merger puzzle. Studies showed that market misvaluation has considerable explanatory power (Shleifer and Vishny, 2003; Rhodes-Kropf and Viswanathan ,2004; Dong et al., 2006). However, this approach does not take the hot market into account fully. A recent literature tries to solve this puzzle using behavioral finance approach (Rosen, 2006; Petmezas, 2008). During the hot merger markets, when optimism increases, stock returns are influenced by investors’ sentiment. However, prices are reversed in the long-run when the market realizes that mergers are announced in hot periods. Studies examining the M&As show that payment method of mergers affects the bidding firm’ stock prices (Travlos, 1987). This assumption supports the signaling hypothesis. Cash financing is associated with positive abnormal returns at merger announcements for bidding firms. However, the empirical results are conflicting. Loughran and Vijh (1997) selected control firms based on size and book-to-market equity and they compared the stock performances of merged firms with those of control firms. They found that cash mergers outperform stock mergers both in the short and in the long-run. Chang (1998) examined the payment method of mergers and showed that bidding firms are associated with more positive abnormal returns in equity mergers. There are few studies that comprehensively examine the long-run stock performance of the Turkish mergers. Among them, Citak and Yildiz (2006) examined the buy-and-hold abnormal returns (BHARs) and cumulative abnormal returns (CARs) of the ISE non-financial firms from 1997 to 2005. They found significant positive stock returns 1 month after the merger activities. They reported no significant impact of merger announcements on the stock returns in the long-run. Mandaci (2004) examined whether or not M&A announcements provide positive abnormal returns to the stockholders of the ISE listed firms during 19982003 period. Her findings show that statistically significant abnormal returns were observed in the first and second day preceding and in the first day following the merger announcement dates. In addition to this, the findings showed that the stock returns were more statistically significant before the announcements than those after the announcements. Mandaci (2005) also examined the effect of M&As on the financial structure and performance of the

24

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

ISE-listed firms. Her study included 14 mergers from 1998 through 2000. The findings of the study reveal that the liquidity ratios tended to decrease after the merger announcements. 2.1. Mergers and Acquisition in Turkey Turkey’s M&A activities are deeply associated with the M&A waves in the US and Europe. In order to understand M&A activities in Turkey, worldwide M&As should be taken into account. There are several general trends in merger waves. As summarized by MacCarthy and Schmidt (2006), during the 1980s, the M&A market was primarily driven by hostile takeovers and demergers of earlier formed conglomerates. Up through the late 1990s, the M&A market was characterized by strategic expansion and liquid capital markets. The focus was on technology and growth. The period between 2000 and 2005 was characterized by a focus on core competencies and profitability rather than growth and expansion. Structural changes in most sectors were driven by increasing globalization as well as by escalating technological development. In Turkey, M&A activities are primarily significant for attracting investments. In early 2000s, Turkey experienced a merger wave after the recovery period of the 2001 banking crisis. In order to understand M&A activities, it is better to have a close look into the Turkish economy. Turkey, as an emerging market, is shaped by economic crises and turbulences. Among them, the 1994 crisis and the 2001 crisis had extensive effects. The Turkish economy was put under global pressure during 1990s. The exchange rates increased dramatically, foreign investment left the country, problems of short-run debt payment increased, and the availability of internal credit decreased. At the end of the 1990s, the Asian crisis in 1997, the Russian crisis, the financial crises in 1999 and 2000 affected the economy negatively and these consecutive crises portended the 2001 financial crisis (Akyuz and Borotav, 2002; Onis and Alper, 2002; Yeldan 2002). The Turkish economy succumbed to hyperinflation and it kept being fragile over years. Hyperinflation reduced the size of the financial sector and gradually eroded the efficiency of the price system. The 2001 banking crisis has been a feature of the hyperinflation and the cause of the fragility in the market was high public sector borrowing requirement and the way it was financed. Turkey initiated a recovery program with the help of IMF and accelerated its privatization program in order to overcome the 2001 financial crisis. Pressures

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

25

weakened and the exchange rate recovered somewhat and long-run interest rates declined. The Turkish economy grew rapidly after the 2001 financial crisis. The recovery was impressive, annual inflation fell steadily, reaching single digits in 2004 for the first time in three decades. Both fiscal and monetary policies improved confidence in the markets and reduced risk premium. Turkey has become attractive to new investors both domestic and external. The increase of competition in all markets pushed the merger waves.

18000 16000 14000 12000 10000 8000 6000 4000 2000 0

1400000 1200000

Million $

1000000 800000 600000 400000 200000 0

Million $

Figure 1: Mergers and Acquisitions in Turkey

1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Year of Announcement

Turkey

Source: UNCTAD, World Investment Report, 2008. Figure 1 shows the M&A activities in Turkey and in the world from 1987 through 2006. There was a remarkable increase in the M&As in 2001 when the banking crisis occurred. The regulatory authorities such as the Banking Regulatory and Supervisory Agency (BRSA) encouraged bank consolidations. Following the crisis, bank mergers became a commonplace. In particular, the period including the years from 2002 to 2006 was characterized by a high number of deals and high volume of transactions. Compared to transactions in the 1990s, there was a remarkable leap on the merger values in the mid of 2000s (World Investment Report, 2008). The M&A activities intensified, during 2005 and 2006, following the recent developments in the Turkish economy. This is consistent with the M&As around the globe. The increase in the volume of M&As can be attributed to the sale of Turkish state-owned entities such as Turk

26

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

Telekom, Tupras (the leading state-owned oil refineries), and Erdemir (the stateowned iron-steel monopoly). III. Data The sample is constructed by examining the `Year Book of Companies` listed on the ISE from 1997 to 2006. Each merge activity was gathered manually and the dates of M&A registrations were noted carefully. Since the analysis of financial statements of financial firms need special treatment, the sample is restricted to the ISE non-financial firms. The financial statements were obtained from the ISE CD Rom financial database and daily stock price data for bidding firms was taken from the Analiz Software Co. Database. 3.1. Selection of the Control Firms Matching procedure is conducted based on the study of Barber and Lyon (1997). The matching firms were selected amongst the firms that have not announced any mergers or acquisitions within the last 2 years. A benchmark of control firms is constructed matched by size, book-to-market ratio. Thus, two groups of control firms are used in the benchmark methodology. Firms size is measured as the market equity of a firm and market-to-book ratio is used as proxy for growth potential. A sample size for each benchmark group consists of 36 control firms. Each merged firm was matched with non-merged firm and then stock return of merged firm was compared with that of non-merged firm. 3.2. Limitations of the Study One of the limitations of the paper is rumors during the pre-merger period. Merger announcements are usually poorly kept as secret. There are rumors of the potential M&As before they actually take place. Therefore, the stock price reactions around the merger announcements may not reflect the real values. There is a significant concern about the rumors of M&As and it is difficult to avoid the informed traders. The level of informed trading in stock markets is a crucial question. In order to avoid the effect of the rumor of a proposed alliance on stock returns, the stock price reactions prior to M&A activities are analyzed. The pre-merger period is extended up to 12 months prior to announcements. Another constrain is that in the sample, most of the target firms are not listed on the ISE. Since the sample does not allow us to examine the stock performances of target firms, stock performance analysis to only acquirer firms are constrained.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

27

IV. Methodology The majority of previous M&A studies have measured the stock price reactions to merger announcements by applying event study methodology. In this study, the market-adjusted abnormal returns of bidding firms before, around and after the merger announcement dates are examined. Both pre-and post-merger performances were calculated using 12-month time span. In order to measure stock price reactions, a standard event study methodology is applied (Ritter, 1991; Loughran and Ritter; 1995). One month consists of 20 trading days. The market-adjusted return for stock i in event month t is defined as below. In Equation 1, ari , t represents market-adjusted abnormal return, ri ,t represents abnormal return and rm ,t shows the market return for the ISE-100 index.

ar i , t  ri ,t  rm ,t

 

(1)

The average market-adjusted return on a portfolio of n stocks for event month t is the arithmetic average of the market-adjusted returns:

AAR t 

1 n  ari ,t n i 1

(2)

Cumulative average abnormal returns (CAARs) are used to evaluate the short-run and long-run performance of stock returns following the merger announcements. The short-run performance analysis covers 5 consecutive days before and after the merger announcements. The long-run performance analysis covers 12 consecutive months. The cumulative average abnormal return (CAAR) from month q to month s is defined as: s

CAAR q, s   AAR t tq

(3)

28

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

V. Empirical Results Figure 2 shows M&A activities in Turkey from 1995 through 2006. It is clear that merger activities intensified after 2001 when the financial crisis emerged. It is not a coincident that many firms tended to merge during and after the crisis period. Many firms were hit by a severe financial crisis and they tended to involve in M&A in order to raise funds. It is important to note that merger activities speeded up in the banking industry following the 2001 financial crisis. Since the crisis emerged in finance industry, most of the banks were acquired by foreign investors. However, the number of mergers and acquisitions increased in the other industries after 2003 when the economy was relatively stabilized. It is likely that in particular, manufacturing companies waited until the economy was stabilized. The year 2003 was a turning point that economic indicators exhibited a range of positive trends. The inflation decreased, macro-economic indicators improved, and the confidence in the economy was restored. Figure 2: M&A Activities Across Years 2006 2005 2004 2003 2002

Year 2001 2000 1999 1998 1997 1996 ACB AKC AKC ANA ARC ARC ASU AYG BAN BRS BSP BSP DERI DUR ENK GUB HUR KEN KOR KRT MAA MER MIG MIL OLM OTK OTK OYS PIN PTO SAB TIR TNS TOA TOF TUR YAS D NS NS* CM LK LK* ZU AZ VT AN RO RO* M OF AI RF GZ T DS EK LT KO RS YT KS AR AR* AC SU FS AH E AS SO AS CAS AS

Firms Firms

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

29

Table 1: Financial Performance of Firms Before and After M&A Ratios Tangibility Ratio Profitability Ratio Liquidity Ratio Equity Ratio Leverage Ratio M/B ST/TL ST/TA LT/TA

Number of Firms (N) 36 35 36 36 36 35 36 36 36

Rumor (Mean)

Approval (Mean)

t-test

0.420 0.057 1.788 0.428 0.569 8.006 0.691 0.399 0.170

0.482 0.042 1.819 0.461 0.539 5.853 0.679 0.366 0.172

-1.373 0.603 -0.116 -0.642 0.595 -0.066 0.285 0.748 -0.097

Note: Tangibility Ratio refers to total fixed assets divided by total assets. Profitability ratio refers to net income after tax divided by total assets. Liquidity ratio refers to current assets divided by current liabilities, Equity Ratio refers to equity capital divided by total asset. M/B refers to market to book ratio. ST/TL is the ratio of short-run liabilities to total liabilities. ST/TA is the ratio of short-run liabilities to total assets. LT/TA is the ratio long-run liabilities to total assets. Leverage Ratio refers to total liabilities divided by total assets.

Table 1 shows the financial performances of bidding firms before and after the M&A activities. It is assumed that firms merge in order to accelerate their growth and create synergy. Hence, the financial performances of the bidding firms are expected to be better after the M&A announcements. The column of Rumor denotes the date, 6 months prior to official declaration of the M&A of a firm. In this study, it is assumed that rumors about the M&A activities start prior to official declaration on average. The column of Approval denotes the official approval date of the M&A of a firm. The findings show that while the tangibility, liquidity, equity, and LT/TA ratios increased, the profitability, leverage, ST/TL, and ST/TA ratios and growth potential of the firms decreased at the date of M&A official approval. There is a slight difference observed between the ratios at the date of rumor and at the date of approval. In general, the reported financial ratios are higher after the M&As are officially approved. However, none of the findings is statistically significant. This suggests that there is no significant impact of official M&A approval announcement on the financial performance of the bidding firms.

30

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

Table 2: Short-run Pre-and Post-Merger Stock Returns for Bidding Firms Panel A: Short-run Pre-Merger Stock Returns Rumor

Declaration

Approval

Stock Returns

AAR

CAAR

t-values

AAR

CAAR

t-values

AAR

CAAR

t-values

- 5 days

0.699

0.699

1.76c

0.993

0.993

2.49b

0.418

0.418

1.289

- 4 days

0.574

1.273

1.43

1.187

2.180

2.54b

0.245

0.663

0.651

- 3 days

0.636

1.909

1.25

1.216

3.396

2.28b

0.177

0.840

0.391

- 2 days

1.195

3.104

2.07b

1.329

4.725

1.92c

-0.222

0.618

-0.430

- 1 day

1.668

4.772

1.99c

2.023

6.748

2.17b

-0.282

0.336

-0.519

Panel B: Short-run Post-Merger Stock Returns + 1 day

0.563

0.563

0.96

2.492

2.492

2.71a

0.597

0.597

0.73

+ 2 days

0.168

0.731

0.36

0.872

3.364

1.46

0.108

0.705

0.27

+ 3 days

0.183

0.914

0.42

1.011

4.375

2.07b

0.040

0.745

0.12

+ 4 days

0.170

1.084

0.43

0.670

5.045

1.81c

-0.019

0.726

-0.06

+ 5 days Number of Firms

-0.055

1.029

-0.15

0.648

5.693

2.10b

0.165

0.891

0.52

37

37

37

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Table 2 reports short-run pre-and post-merger stock returns from -5 days through +5 days. The findings show that the market reacted positively prior to official declaration announcement in the short-run. Table 2 presents positive and relatively high stock returns 1 week before and 1 week after merger declarations. Mandaci (2004) found similar findings that statistically significant abnormal returns were observed in the first and in the second day preceding and in the first day following the merger announcements. However, in general, the t-values of stock returns around the M&A approval date are not statistically significant.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

31

Table 3: Long-run Pre-and Post-Merger Stock Returns for Bidding Firms Panel A: Long-run Pre-Merger Stock Returns Rumor Stock Returns t-values AAR CAAR - 12 months 0.274 0.274 5.74a - 11 months 0.264 0.538 5.03a - 10 months 0.275 0.813 5.08a - 9 months 0.278 1.091 4.99a - 8 months 0.307 1.398 5.52a - 7 months 0.324 1.722 5.11a - 6 months 0.346 2.068 5.26a - 5 months 0.357 2.425 4.79a - 4 months 0.313 2.738 3.09a - 3 months 0.276 3.014 2.02b - 2 months 0.220 3.234 1.16 - 1 month 0.107 3.341 0.50 Panel B: Long-run Post-Merger Stock Returns + 1 month 0.193 0.193 0.95 + 2 months 0.053 0.246 0.46 + 3 months 0.142 0.388 1.43 + 4 months 0.158 0.546 1.76c + 5 months 0.231 0.777 3.17a + 6 months 0.253 1.030 4.18a + 7 months 0.294 1.324 5.36a + 8 months 0.269 1.593 5.07a + 9 months 0.268 1.861 5.68a + 10 months 0.227 2.088 5.71a + 11 months 0.249 2.337 7.08a + 12 months 0.222 2.559 5.98a Number of 37 Firms

AAR 0.317 0.293 0.284 0.262 0.243 0.265 0.288 0.320 0.379 0.396 0.454 0.554 0.406 0.284 0.232 0.188 0.204 0.158 0.173 0.192 0.215 0.239 0.251 0.248

Declaration t-values CAAR 0.317 6.76a 0.610 5.88a 0.894 5.25a 1.156 4.16a 1.399 3.41a 1.664 4.04a 1.952 4.35a 2.272 4.24a 2.651 4.50a 3.047 4.15a 3.501 3.03a 4.055 2.52b 0.406 0.690 0.922 1.110 1.314 1.472 1.645 1.837 2.052 2.291 2.542 2.790 37

3.70a 2.72a 2.37b 2.56b 3.51a 2.70a 3.31a 3.77a 3.69a 4.36a 4.69a 4.75a

AAR 0.261 0.283 0.296 0.284 0.289 0.317 0.335 0.326 0.328 0.329 0.338 0.332 0.350 0.126 0.191 0.146 0.158 0.172 0.164 0.142 0.140 0.146 0.156 0.145

Approval t-values CAAR 0.261 5.84a 0.544 6.21a 0.840 6.43a 1.124 5.77a 1.413 5.81a 1.730 5.64a 2.065 5.63a 2.391 5.05a 2.719 4.67a 3.048 3.66a 3.386 2.64b 3.718 2.05b 0.350 0.476 0.667 0.813 0.971 1.143 1.307 1.449 1.589 1.735 1.891 2.036

2.04b 1.49 2.92a 2.32b 2.86a 2.63b 2.65b 2.74a 3.02a 3.09a 3.53a 3.53a

37

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Table 3 reports long-run pre-and post-merger stock returns from -12 months through +12 months. The post-merger stock returns are positive and significant. This is consistent with the findings of Citak and Yildiz (2006) who found positive and significant 1 month stock returns following the M&A announcements. The findings in Table 3 suggest that post-merger stock returns are generally lower than pre-merger stock returns in the long-run. Following the M&A announcements, stock returns started to decline slightly and these declines are statistically significant. This finding is consistent with Mandaci (2004) who

32

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

found that the cumulative abnormal returns before the announcement dates were statistically more significant than those after the announcement dates. The findings of this paper are in line with the previous findings. The results suggest that bidding firms generate positive and superior abnormal returns prior to merger announcements. However, stock prices start to decline in the long-run. This may support the investor sentiment as a driving force. As the true value of the merged firm is revealed over time, the long-run returns tend to decrease. Although stock returns are positive after the merger announcements, they are relatively lower than pre-merger stock returns. This reveals the fact that investors are often overoptimistic about the value creation of mergers and acquisitions. In addition to this, superior pre-merger stock returns are the supportive of Kaplan (2000) who argued that positive and high abnormal returns occur prior to announcements, since the M&As are largely anticipated. Table 4: Short-run Stock Returns for Non-Merged Firms by Size (Market Equity) Panel A: Short-run Pre-Merger Stock Returns by Market Equity Stock Returns

Rumor

Declaration

Approval

AAR

CAAR

t-values

AAR

CAAR

t-values

AAR

CAAR

t-values

- 5 days

0.351

0.351

1.25

0.398

0.398

1.28

-0.187

-0.187

-0.52

- 4 days

0.130

0.482

0.41

0.251

0.649

0.83

-0.094

-0.281

-0.25

- 3 days

0.030

0.512

0.08

0.207

0.856

0.55

-0.240

-0.521

-0.58

- 2 days

0.443

0.956

1.02

0.402

1.259

0.93

-0.208

-0.730

-0.39

- 1 day

0.676

1.632

1.00

0.166

1.425

0.28

-0.212

-0.942

-0.29

Panel B: Short-run Post-Merger Stock Returns by Market Equity + 1 day

0.147

0.147

0.17

0.946

0.946

1.70c

0.856

0.856

1.22

+ 2 days

-1.784

-1.637

-1.29

0.821

1.768

1.88c

0.629

1.486

1.39

+ 3 days

-0.673

-2.310

-0.79

0.962

2.730

2.81a

0.437

1.923

1.32

+ 4 days

0.026

-2.284

0.03

0.858

3.589

3.52a

0.422

2.345

1.41

+ 5 days Number of Firms

0.152

-2.132

0.24

0.905

4.494

3.02a

0.681

3.027

2.29b

36

36

36

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

33

Table 4 shows the short-run stock returns of control firms which are gathered by matching market equity. Control firms are selected according to the closest ME of the bidding firms at the time of M&A declaration dates. It is important to note that control firms are not involved in M&A activities at the time of declaration. The benchmark model captures the time period from -5 days through +5 days. The results show that the stock returns of control firms are lower in the premerger period than those in the post-merger period. However, pre-merger stock returns are not statistically significant. In particular, post-merger stock returns of control firms are higher during the declaration and approval periods. It is important to note that while the stock returns of M&A firms are higher prior to M&A (see Table 2), the stock returns of control firms are higher after the M&A declaration. This suggests that stock performance of control firms are not affected from M&A activity, in particular, prior to M&A announcements.

34

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

Table 5: Long-run Stock Returns for Non-Merged Firms by Size (Market Equity) Panel A: Long-run Pre-Merger Stock Returns by Market Equity Rumor Declaration t-values t-values AAR CAAR AAR CAAR - 12 months 0.155 0.155 3.46a 0.104 0.104 2.39b a - 11 months 0.137 0.293 3.06 0.087 0.191 1.78c - 10 months 0.123 0.416 2.69b 0.076 0.268 1.43 - 9 months 0.130 0.546 2.62b 0.059 0.328 1.14 - 8 months 0.139 0.686 2.48b 0.051 0.379 0.83 - 7 months 0.128 0.814 2.06b 0.050 0.430 0.76 - 6 months 0.186 1.000 2.46b 0.040 0.470 0.50 - 5 months 0.148 1.149 1.89c 0.058 0.528 0.67 - 4 months 0.147 1.296 1.37 0.151 0.680 1.79c - 3 months 0.186 1.483 1.60 0.244 0.924 2.68b - 2 months 0.137 1.621 0.84 0.225 1.150 1.82c - 1 month -0.006 1.614 -0.03 0.217 1.368 1.41 Panel B: Long-run Post-Merger Stock Returns by Market Equity + 1 month 0.091 0.091 0.38 0.217 0.217 1.10 + 2 months -0.144 -0.053 -0.87 0.249 0.467 1.73c + 3 months -0.144 -0.197 -1.20 0.183 0.650 1.61 + 4 months -0.040 -0.238 -0.41 0.174 0.825 2.19b + 5 months 0.018 -0.220 0.22 0.216 1.042 2.86a + 6 months 0.027 -0.192 0.34 0.150 1.192 2.15b + 7 months 0.050 -0.141 0.76 0.127 1.319 2.06b + 8 months 0.068 -0.073 1.25 0.158 1.478 2.81a + 9 months 0.090 0.017 1.67 0.156 1.634 2.64b + 10 months 0.084 0.101 1.72c 0.176 1.810 2.83a + 11 months 0.105 0.207 2.42b 0.184 1.995 3.40a b + 12 months 0.086 0.293 2.15 0.183 2.179 3.24a Number 36 36 of Firms Stock Returns

AAR 0.063 0.077 0.083 0.074 0.081 0.096 0.117 0.105 0.058 0.105 0.104 -0.158 0.555 0.412 0.288 0.229 0.224 0.148 0.147 0.146 0.126 0.103 0.124 0.105

Approval t-values CAAR 0.063 1.21 0.141 1.42 0.224 1.40 0.298 1.27 0.379 1.27 0.476 1.43 0.593 1.66 0.698 1.36 0.757 0.65 0.862 1.12 0.967 0.83 0.808 -0.76 0.555 0.968 1.257 1.486 1.710 1.859 2.006 2.152 2.279 2.382 2.507 2.613

2.60b 3.59a 3.43a 3.13a 3.12a 1.90c 2.13b 2.52b 2.35b 2.31b 3.33a 2.93a

36

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Table 5 reports the long-run pre-and post-merger stock returns of control firms according to firm size (Market Equity). This benchmark model captures the time period from -12 months through +12 months. Non-merged firms have higher long-run stock returns after the declaration and approval dates than during the pre-merger period. The stock performances of control firms are positive and

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

35

statistically significant after the M&A declaration and approval. The cumulative abnormal returns following the merger announcements are high in the long-run. In comparison with long-run stock returns of merged firms, the results document that while merged firms had high stock returns prior to merger, non-merged firms had high stock returns after the merger. This evidence can be supportive of the probability of informed insider traders and investors. Table 6: Short-run Stock Returns for Non-Merged Firms by Growth Potential (Market-to-Book) Panel A: Short-run Pre-Merger Stock Returns by Market-to-Book Rumor

Stock Returns

AAR

CAAR

- 5 days

0.332

- 4 days

0.369

- 3 days

Declaration AAR

CAAR

0.332

tvalues 0.63

0.426

0.702

0.76

0.429

0.107

0.809

0.21

- 2 days

0.602

1.412

- 1 day

0.240

1.652

Approval AAR

CAAR

0.426

tvalues 1.40

tvalues 0.14

0.064

0.064

0.856

1.29

0.044

0.108

0.08

0.571

1.427

1.36

-0.129

-0.020

-0.25

1.07

0.917

2.344

1.84c

0.092

0.071

0.16

0.35

0.701

3.046

1.34

-0.186

-0.114

-0.22 1.48

Panel B: Short-run Post-Merger Stock Returns by Market-to-Book + 1 day

-0.210

-0.210

-0.23

-0.603

-0.603

-1.20

1.055

1.055

+ 2 days

-0.615

-0.825

-1.00

-0.384

-0.988

-0.91

0.509

1.564

0.93

+ 3 days

-0.236

-1.061

-0.42

0.162

-0.826

0.52

0.394

1.958

0.98

+ 4 days

0.192

-0.869

0.43

0.402

-0.423

1.07

0.162

2.121

0.41

+ 5 days Numbe r of Firms

0.130

-0.739

0.33

0.460

0.036

0.99

0.160

2.281

0.49

36

36

36

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Table 6 reports short-run pre-and post-merger stock returns of non-merged firms which are gathered according to the closest value of Market-to-Book (M/B) ratio. This approach matches firms according to their growth potential. In the literature, M/B benchmark is commonly used. In Table 6, this benchmark model captures the time period from -5 days through +5 days.

36

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

The findings show that pre-merger stock returns of control firms according to size and growth potential benchmarks are consistent. Stock returns of nonmerged firms, matched by market-to-book ratio, are higher in the pre-merger period than those in the post-merger period. The stock returns are fully reversed following the merger announcements. The stock prices of non-merged firms turned negative after the announcements. Table 6 reports that stock returns of matched firms according to M/B are not statistically significant. Table 7: Long-run Stock Returns for Non-Merged Firms by Growth (Market-to-Book) Panel A: Long-run Pre-Merger Stock Returns by Market-to-Book Rumor Declaration AAR CAAR t-values AAR CAAR t-values - 12 months 0.132 0.132 2.38b 0.175 0.175 3.54a - 11 months 0.108 0.240 1.79c 0.152 0.328 2.99a c - 10 months 0.101 0.342 1.70 0.151 0.479 2.80a - 9 months 0.108 0.451 1.62 0.112 0.591 1.86c - 8 months 0.108 0.559 1.52 0.092 0.684 1.28 - 7 months 0.152 0.711 2.08b 0.140 0.824 1.75c - 6 months 0.234 0.945 2.86a 0.168 0.992 2.10b c - 5 months 0.175 1.121 1.72 0.166 1.159 1.59 - 4 months 0.192 1.313 1.51 0.261 1.421 2.29b - 3 months 0.124 1.437 0.84 0.354 1.775 3.36a - 2 months 0.002 1.440 0.01 0.352 2.128 2.91a - 1 month -0.059 1.381 -0.19 0.291 2.419 1.45 Panel B: Long-run Post-Merger Stock Returns by Market-to-Book + 1 month 0.280 0.280 1.18 0.293 0.293 1.33 + 2 months 0.018 0.299 0.14 0.097 0.391 0.58 + 3 months -0.003 0.295 -0.03 0.020 0.412 0.16 + 4 months 0.046 0.341 0.41 0.013 0.425 0.13 + 5 months 0.152 0.494 1.69c 0.025 0.450 0.24 + 6 months 0.136 0.630 1.64 -0.018 0.432 -0.20 + 7 months 0.175 0.806 2.70b -0.012 0.419 -0.17 + 8 months 0.146 0.952 2.40b -0.007 0.412 -0.10 + 9 months 0.120 1.073 1.94c 0.022 0.434 0.34 + 10 months 0.098 1.171 1.71c 0.059 0.493 1.03 + 11 months 0.103 1.275 1.80c 0.063 0.557 1.21 + 12 months 0.056 1.331 1.03 0.060 0.618 1.16 Number of 36 36 Firms Stock Returns

AAR 0.092 0.104 0.113 0.113 0.120 0.120 0.115 0.138 0.123 0.138 0.112 0.098 0.327 0.061 0.157 0.175 0.164 0.091 0.077 0.070 0.085 0.068 0.074 0.081

Approval t-values CAAR 0.092 1.85c 0.197 1.90c 0.310 1.98c 0.424 1.98c 0.544 1.97c 0.664 1.71c 0.780 1.61 0.918 1.64 1.042 1.08 1.180 0.97 1.293 0.64 1.391 0.48 0.327 0.389 0.546 0.721 0.886 0.978 1.055 1.125 1.211 1.279 1.354 1.435

1.66 0.40 1.47 2.19b 2.17b 1.25 1.13 1.19 1.78c 1.51 1.91c 2.27b

36

Note: AAR indicates Average Abnormal Returns and CAAR indicate Cumulative Average Abnormal Returns. Numbers are mean values. t–statistics are given with a null hypothesis of equal means. a1% significant level, b5% significant level, c10 % significant level. There are 3 time lines. First, rumor column represents the time lag of 6 months prior to official M&A declaration date. It is assumed that the rumors of the M&A activities started or intensified almost 6 months before the official declaration date on average. Second, declaration represents official announcement date of M&As and the last column of approval represents official M&As approval date.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

37

Table 7 reports long-run pre-and post-merger stock returns of non-merged firms which are matched according to market-to-book ratio (MB). This benchmark model captures the time period from -12 months through +12 months. The findings show that in general, pre-merger long-run returns for nonmerged firms are higher than those of post-merger returns. The results suggest that while pre-merger long-run returns are statically significant, post-merger long-run returns are not significant. Merger announcements have no effect on the long-run stock returns of non-merged firms which are selected according to market to book ratios. While ME-matched firms have high and significant postmerger performance in the long-run, M/B-matched firms have high and significant pre-merger returns in the long-run. VI. Conclusion The main purpose of this study is to measure the stock price reactions of bidding firms f to merger announcements of the ISE listed non-financial firms from 1997 through 2006. This paper provides empirical evidence of stock returns before and after merger announcements. The findings of the paper are not limited to stock price reactions of bidding firms but also stock return performances of nonmerged firms are calculated around the time event of M&As. While doing so, control firms are selected according to size (market equity) and growth potential (market-to-book). The findings of the paper show that merger activities increased, in particular, after the period of the 2001 financial crisis. It is important to note that manufacturing firms waited to engage in M&A activities until the economy was relatively stabilized. When the Turkish economy started to recover in 2003, the merger and acquisition activities intensified. The stock returns are calculated using cumulative average abnormal returns (CAAR). Since one of the major limitations of M&A studies is informed investors, stock returns prior to M&A declaration date are calculated. It is assumed that investors might get M&A news before the declaration and approval dates. Thus, the pre-merger stock returns are calculated 12 months prior to M&A. The findings present that the operating performances of bidding firms do not change significantly in the post-merger period. Although liquidity, tangibility, equity and leverage, ratios of the firms increased after the M&A

38

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

announcements, these financial performance improvements were not statistically significant. The empirical evidence on the stock return performance shows that premerger stock returns are higher than post-merger stock returns in the short-run. This is consistent with the previous findings in the literature showing the fact that investors’ optimism was dominant during the pre-merger period. Stock returns decreased in the long-run, when the information about true value of the firms increased over time. High pre-merger stock returns are supportive of Kaplan (2000)’s findings. The reason is that M&As are largely anticipated. The market, in particular in the short-run, reacts positively prior to official declaration announcement. Further, the current paper compared the merger announcement effects of merged firms versus non-merged firms on the basis of size and growth potential. The findings show that while ME-matched firms had higher post-merger stock returns than pre-merger stock returns in the short-run, In addition to this, M/Bmatched firms had higher pre-merger stock returns than post-merger returns in the short-run. However, the findings show mixed results for the matching firms in the long-run. While the post-merger stock returns are higher than pre-merger stock returns for ME-matched firms in the long run, the pre-merger stock returns are higher than post-merger stock returns for M/B-matched firms in the long run. In general, the results suggest that merged firms had high stock returns prior to merger. This evidence can be supportive of the probability of informed traders and investors. The stock returns of merged firms are higher than those of nonmerged firms both in the short-run and in the long-run. In the further research, it will be interesting to investigate whether insider trading has significant effect on the pre-merger stock returns or not. This paper is a first attempt to apply market equity and growth potential benchmarks that provide useful insight in interpreting post-merger stock returns in the long run.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

39

References Agrawal, A.; Jaffe, J., “Do Takeover Targets Underperform? Evidence from Operating and Stock Returns”, Journal of Financial and Quantitative Analysis, 38(2003): 2003, ss. 721–746. Akyuz, Y.; Boratav, K., “The Making of the Turkish Crisis”, UNCTAD, Working Paper, 2002. Chang, S., “Takeover of Privately Held Targets, Methods of Payment and Bidder Returns”, Journal of Finance, 2, 1998, ss. 773-784. Çıtak, L.; Yıldız, K., “Devralmanın Devralan İşletmelerin Hisse Senedi Getirileri Üzerindeki Etkisi: Sermaye Piyasası Kurulu İzni İle Gerçekleşen Devralmaların Devralan İşletmelerin Hisse Senedi Getiri Oranları Üzerindeki Etkisinin İncelenmesi”, 10. Ulusal Finans Sempozyumu, Dokuz Eylül Üniversitesi, İzmir, 2006. Dennis D. K.; McConnell J. J., “Corporate Mergers and Security Returns”, Journal of Financial Economics, 16, 1986, ss. 143–187. Dickerson, A. P.; Gibson, H. D.; Tsakalotos, E., “The Impact of Acquisitions on Company Performance: Evidence from a Large Panel of UK Firms”, Oxford Economic Papers, 49(3): 1997, ss. 344-361. Dong, M.; Hirshleifer, D.; Richardson, S.; Teoh, S. H., “Does Investor Misvaluation Drive the Takeover Market?”, Journal of Finance, 61, 2006, ss. 725-762. Healy, P. M.; Palepu, K. G.; Ruback, R. S., “Does Corporate Performance Improve after Mergers?”, Journal of Financial Economics, 31(2): 1992, ss. 135-175. Kaplan, N. S., Mergers and Productivity, University of Chicago Press, 2000. Keown, A. J.; Pinkerton, J. M., “Merger Announcements and Insider Trading Activity: An Empirical Investigation”, Journal of Finance, 36, 1981, ss. 855-869. Loughran, T.; Vijh, M.A., “Do Long-term Shareholders Benefit from Corporate Acquisitions”, Journal of Finance, 52, 1997, ss. 1765–1790. Loughran, T.; Ritter, J., “The New Issues Puzzle”, Journal of Finance, 50, 1995, ss. 23-52. MacCarthy C., Schmidt W., Denmark Limited-Global by Design, Narayana Press: Denmark, 2006. Mandaci, P. E., “Sirketlerin Birleşme ve Satın Alma Duyurularının Hisse Senedi Fiyatları Üzerine Etkileri”, İktisat İşletme ve Finans, 19(225): 2004, ss. 118–124. Mandaci, P. E., “Şirketlerin Birleşmelerinin Finansal Yapı ve Performanslarına Etkileri”, İktisat İşletme ve Finans, 20(233): 2005, ss. 60–67. Mitchell, M.; Mulherin, J. H., “The Impact of Industry Shocks on Takeover and Restructuring Activity”, Journal of Financial Economics, 41, 1996, ss. 193–229. Mitchell, M. L.; Stafford E., “Managerial Decisions and Long-Term Stock Price Performance”, Journal of Business, 73: 2000, ss. 287–320.

40

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

Öniş, Z.; Alper, E., “Emerging Market Crises and the IMF: Rethinking the Role of the IMF in the Light of the 2000–2001 Financial Crises in Turkey”, 2002, Araştırma VI. METU Conference on Economics’ta Sunulmuştur, Eylül 11–15, Ankara. Petmezas, D., “What Drives Acquisitions?: Market Valuations and Bidder Performance”, Journal of Multinational Financial Management, 19(1), 2008, ss. 54-74. Ramaswamy, K. P.; Waegelein, J. F., “Firm Financial Performance Following Mergers”, Review of Quantitative Finance and Accounting, 20(2), 2003, ss. 115–126. Ravenscraft, D.; Scherer, F. M., “The Profitability of Mergers,” International Journal of Industrial Organization, 7, 1989, ss. 101–116. Rhodes-Kropf, M.; Viswanathan, S., “Market Valuation and Merger Waves”, Journal of Finance, 59, 2004, ss. 2685–2718. Ritter, J., “The Long-run Performance of Initial Public Offerings”, Journal of Finance, 42, 1991, ss. 365–394. Rosen, R., “Merger Momentum and Investor Sentiment: the Stock Market Reaction to Merger Announcements”, Journal of Business, 79, 2006, ss. 987–1017. Shleifer, A.; Vishny, W. R., “Stock Market Driven Acquisitions”, Journal of Financial Economics, 70, 2003, ss. 295–311. Travlos, N. G.; “Corporate Takeover Bids, Methods of Payment, and Bidding Firms’ Stock Returns”, Journal of Finance, 42, 1987, ss. 943–963. UNCTAD, World Investment Report, 2008. Wong, A.; Cheung, Y. K., “The Effects of Mergers and Acquisition Announcements on the Security Price of Bidding Firms and Target Firms in Asia”, International Journal of Economics and Finance, 1(1): 2009, ss. 274-283. World Investment Report (WIR), World Investment Report - Transnational Corporations and the Infrastructure Challenge. New York ve Genova, 2008. Yeldan, E., “Behind the 2000-2001 Turkish Crisis: Stability, Credibility, and Governance, for Whom?”, Bilkent University, Working Paper, 2002.

Stock Price Reactions to Merger Announcements: Evidence from Istanbul Stock Exchange (ISE)

41

APPENDIX In Turkey, the majority of the firms are family-owned. There is a lack of institutional structure and professional management system. Generally, CEOduality is the common case among Turkish firms. Therefore, it is hard to improve the process of mergers and acquisitions. This is one of the reasons why Turkish firms tend to implement horizontal mergers. The tables given in the Appendix show the industrial classification of bidder and target firms. Most of the M&As are in manufacturing industry and they are in the form of horizontal mergers. Table 8: Control Firms Which Used in Benchmark Model Declaration Year of Merger 1997 1998 1999 1998 1998 1998 1998 1999 1999 1999 1999 2000 2000 2000 2001 2001 2001 2001 2002 2002 2000 2002 2002 2001 2002 2002 2002 2002 2000 2002 2003 2004 2003 2004 2004 2005 2005

Bidder Firm + Target Firm → Result TOFAS + Opar Otomotiv → TOFAS AKCNS + Betonsa Beton → AKCNS TURCS + Tabaş → TURCS ARCLK + ARDEM → ARCLK OTKAR + Otokar → OTKAR SABAH + BUGUN → SABAH KORDS + Dusa Endüstriyel → KORDS ARCLK + Türk Elektrik&Atılım&Gelişim → ARCLK MERKO + Sultanköy&Frumiks → MERKO HURGZ + Gerçek → HURGZ ANACM + Topkapı → ANACM AYGAZ + Gazal → AYGAZ BANVT + Tadpi → BANVT TOASO + TOFAS → TOASO DERIM + Has Deri → DERIM BSPRO + BSH Grünberg&BSH Ev&Profilo → BSPRO OLMKS + Olmuksa → OLMKS YASAS + BYRBY → YASAS AKCNS + Agregasa → AKCNS ENKAI + ENKA → ENKAI TNSAS + ATI Dış Ticaret → TNSAS TIRE + Bomsaş → TIRE GUBRF + Gübretaş → GUBRF OTKAR + İstanbul Fruehauf → OTKAR KRTEK + Konfeksiyon Sanayi → KRTEK PTOFS + İş Doğan → PTOFS ASUZU + Otopar → ASUZU NIGDE + Oysa İskenderun → NIGDE KENT + Birlik → KENT MILYT + Simge → MILYT MAALT + Tütaş → MAALT BRSAN + Mannesmann → BRSAN BSPRO + BSH PEG Beyaz Eşya → BSPRO ACIBD + Acıbadem Bursa&Acıbadem Kanser → ACIBD DUROF + Duran Makine&Doğan Matbaacılık → DUROF PINSU + Marmara → PINSU MIGRS + TNSAS → MIGRS

Market Equity VESTL VESTL TATKS AYGAZ BOYNR YATAS AKSA

Market-toBook MIGRS DENCM KOTKS HZNDR TUPRS TOFAS EGGUB

THYAO

TOASO

BRFEN ADANA BRSAN BSPRO GOODY TNSAS KLBMO

ANACM DITAS KRTEK KERVT GOODY BURCE UNYEC

CIMSA

ALCTL

BAGFS OYSAC BSPRO TUPRS AKCNS FMIZP HEKTS KIPA HEKTS ARCLK ALCAR ACIBD BUCIM SARKY UKIM BSPRO SISE

LIOYS KENT PTOFS TCELL CBSBO DOKTS DOBUR ZOREN HEKTS LINK OTKAR ALKIM FROTO ALKIM MUTLU TNSAS FMIZP

KIPA

TOASO

n.a.

n.a.

HZNDR TOASO

ECILC FROTO

42

Berna Kırkulak Uludağ & Özlem Demirkaplan Gülbudak

Table 9: Industry Classification of M&A Activities Bidder Firm + Target Firm → Result

Industry

TOFAS + Opar Otomotiv → TOFAS

Manufacturing

AKCNS + Betonsa Beton → AKCNS TURCS + Tabaş → TURCS

Manufacturing Manufacturing

ARCLK + ARDEM → ARCLK

Manufacturing

OTKAR + Otokar → OTKAR

Manufacturing

SABAH + BUGUN → SABAH KORDS + Dusa Endüstriyel → KORDS ARCLK + Türk Elektrik&Atılım&Gelişim → ARCLK MERKO + Sultanköy&Frumiks → MERKO HURGZ + Gerçek → HURGZ ANACM + Topkapı → ANACM AYGAZ + Gazal → AYGAZ BANVT + Tadpi → BANVT

Manufacturing Manufacturing Manufacturing Manufacturing

Food and Beverage and Tobacco Products

Manufacturing Manufacturing Mining Manufacturing

Printing and Related Support Activities Nonmetallic Materials Oil and Gas Extraction Food and Beverage and Tobacco Products Motor Vehicles, Bodies and Trailers, and Parts Apparel and Leather and Allied Products Motor Vehicles, Bodies and Trailers, and Parts Paper Products Chemical Products Nonmetallic Materials Nonmetallic Materials n.a. Paper Products Chemical Products Motor Vehicles, Bodies and Trailers, and Parts Textile Mills and Textile Product Mills Petroleum and Coal Products Motor Vehicles, Bodies and Trailers, and Parts Nonmetallic Materials Food and Beverage and Tobacco Products Printing and Related Support Activities

TOASO + TOFAS → TOASO

Manufacturing

DERIM + Has Deri → DERIM BSPRO + BSH Grünberg&BSH Ev&Profilo → BSPRO OLMKS + Olmuksa → OLMKS YASAS + BYRBY → YASAS AKCNS + Agregasa → AKCNS ENKAI + ENKA → ENKAI TNSAS + ATI Dış Ticaret → TNSAS TIRE + Bomsaş → TIRE GUBRF + Gübretaş → GUBRF

Manufacturing Manufacturing Manufacturing Manufacturing Manufacturing Manufacturing Retail Trade Manufacturing Manufacturing

OTKAR + İstanbul Fruehauf → OTKAR

Manufacturing

KRTEK + Konfeksiyon Sanayi → KRTEK PTOFS + İş Doğan → PTOFS

Manufacturing Manufacturing

ASUZU + Otopar → ASUZU

Manufacturing

NIGDE + Oysa İskenderun → NIGDE KENT + Birlik → KENT MILYT + Simge → MILYT MAALT + Tütaş → MAALT BRSAN + Mannesmann → BRSAN BSPRO + BSH PEG Beyaz Eşya → BSPRO ACIBD + Acıbadem Bursa&Acıbadem Kanser → ACIBD DUROF + Duran Makine&Doğan Matbaacılık → DUROF PINSU + Marmara → PINSU MIGRS + TNSAS → MIGRS

Sub-Industry Motor Vehicles, Bodies and Trailers, and Parts Nonmetallic Materials Petroleum and Coal Products Motor Vehicles, Bodies and Trailers, and Parts Motor Vehicles, Bodies and Trailers, and Parts Printing and Related Support Activities Textile Mills and Textile Product Mills Motor Vehicles, Bodies and Trailers, and Parts

Manufacturing Manufacturing Manufacturing Educational Services, Health Care, and Social Assistance Manufacturing Manufacturing Educational Services, Health Care, and Social Assistance

Social Assistance Plastic and Rubber Products Motor Vehicles, Bodies and Trailers, and Parts Hospitals and Nursing and Residential Care Facilities

Manufacturing

Paper Products

Manufacturing Retail Trade

Food and Beverage and Tobacco Products n.a.

The ISE Review Volume: 12 No: 47 ISSN 1301-1642 © ISE 1997

A GAME THEORETIC APPROACH TO MODEL FINANCIAL MARKETS: GUESSING GAME Ü. Barış URHAN* Zafer AKIN**

Abstract How deep our reasoning and strategic thinking is and how strategic we think are crucial in making financial investment decisions because we need to guess how other people will behave. This paper studies Guessing Game (Beauty contest game) that helps us understand how people make decisions in environments where depth of reasoning plays a key role. We mainly summarize the findings of experimental studies focusing on this game and then link these findings to financial markets, especially to stock market players’ behavior. We also offer some extensions that may further contribute to produce better and more realistic models of financial markets.

I. Introduction A financial market is a mechanism or a medium in which people buy and sell financial instruments and commodities. Financial markets basically match the ones who own capital to the ones who need them. In other words, it allocates available savings in the most productive way by eliminating informational frictions between borrowers and lenders of fungible items such as stocks, bonds, agricultural goods and oil. By nature, financial markets are the fundamental channels that transmit changes in one part of the economy to other parts since transactions can occur instantly as opposed to other markets in which transactions may take a long time to occur. __________________________________________________________________________________________________________________________________ *

**

U. Baris Urhan, Economic Policy Research Foundation of Turkey (TEPAV), Sogutozu Caddesi, No: 43, Sogutozu, Ankara, 06560, Turkey. Tel: (90-312) 292 5521 Fax: (90-312) 292 5555 E-mail:[email protected] Web: www.barisurhan.com Assist. Prof. Dr. Zafer Akin, TOBB University of Economics and Technology, Department of Economics, No: 43, Sogutozu, Ankara, 06560, Turkey. Tel: (90-312) 292 4216 Fax: (90-312) 292 4104 E-mail:[email protected] Web: http://zafer.akin.etu.edu.tr/ Key Words: Guessing game; experiment; team decision-making; individual decision-making JEL Classification: C72; C91, C92

44

Ü. Barış Urhan & Zafer Akın

A broad look at the structure of financial markets indicates the following ingredients: traders as players, strategies of traders, and outcomes and implied payoffs for each trader determined by their collective strategies under some trading rules. Since any situation that involves parties with possibly conflicting interests, their strategies and payoffs emerging from joint strategies can be named as a game, this essentially brings us to a point where we see a financial market as a game. Then, in order to understand and investigate this very complex game, we can utilize the tools of Game Theory. Game theory, a branch of applied mathematics, is simply a science of decision making in strategic environments. It primarily tries to understand and model the behavior of agents who interact in different strategic situations. It has numerous applications in economics and finance as well as in other sciences such as biology, political science, international relations and computer science. Game theory generates predictions on how a certain game should be played under some rationality and informational assumptions. Since we are interested in financial markets and try to understand their dynamics, we need a simple model that will give insights about how people behave in financial markets. At this point, Keynes gives us a clue in understanding the dynamics of financial markets, more broadly of any investment decision1 by making an analogy between a professional investment and a newspaper contest: “...professional investment may be likened to those newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those faces which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the other competitors, all of whom are looking at the problem from the same point of view. It is not a case of choosing those which, to the best of one's judgment, are really the prettiest, nor even those which average opinion genuinely thinks the prettiest. We have reached the third degree where we devote our intelligences to anticipating what average opinion expects the average opinion to be. And there are some, I believe, who practice the fourth, fifth and higher degrees.2”

__________________________________________________________________________________________________________________________________ 1 2

Colin Camerer, Behavioral Game Theory, Princeton University Press, 2003, p.16. John Maynard Keynes, The General Theory of Employment, Interest and Money, London: MacMillan, 1947, p.156.

A Game Theoretic Approach to Model Financial Markets: Guessing Game

45

This implies that in any investment decision3, it is important what you believe about what other people will do and to act in accordance with these beliefs. Moreover, as mentioned, it is not enough to form beliefs about what others will do, it is also crucial to consider what others think about what others will do. Perfectly rational players must consider this recursively and act accordingly. However in real life, most people do not take this consideration into account and they fail to make a profit out of their investments. Inspiring from this quote, a game called guessing game (also known as pbeauty contest game) is introduced. The guessing game incorporates the mentioned reasoning in a very simple game format. When we employ game theoretical tools to predict the outcome of this game, it gives a unique and trivial solution. However, since there is no clear cut answer of whether human decisions occur in the same way as game theory predicts4 and since there is a vast amount of evidence that there are significant differences between traditional game theoretical results and the actual behavior of people, we need to test the theory’s predictions. The most apparent and directly informative method to test predictions of game theory is to conduct experiments in controlled environments. It turns out that guessing game is a striking example because although it is very simple and understandable by everyone, experimental results clearly show that people do not employ such reasoning in their decision making process and that standard theory’s predictions simply fail. The aim of this paper is to analyze experimental studies on guessing game that is an extremely simple model of financial markets. To fix the ideas, we give a real life example from stock traders’ behavior. Especially inexperienced traders make decisions cursorily and use some rule of thumbs. One example of this “irrational” behavior is selling winning investments too early and selling losing investments too late. The reason for this behavior is the lack of recursive thinking in the sense that one should sell a winning stock whenever he thinks that other traders thinks that others will sell it, not at the time that he thinks that others will sell it (This is only the third level of thinking and actually it should proceed further, if necessary). There is a fairly recently emerging field called “behavioral finance” investigating the interaction of psychological characteristics people have with the financial decision making __________________________________________________________________________________________________________________________________ 3

4

This can be thought as any investment opportunity that one encounters such as investing in a land, a real estate, a stock and so on. Even an academician, who believes a specific area will become popular soon and invests his/her time and effort to that area, can be considered as an investor. Colin Camerer, Behavioral Game Theory, Princeton University Press, 2003, p.20.

46

Ü. Barış Urhan & Zafer Akın

processes (for an extensive survey of Behavioral finance see Thaler and Barberis, 2002; Olsen, 1998; and Shleifer, 2000). One of the main reasons of the observed anomalies in financial markets, especially in stock markets is the mentioned lack of mentioned reasoning. This paper aims to increase awareness of this phenomenon that has attracted academic and professional attention recently by focusing on the guessing game which intuitively reflects this lack of reasoning. The remainder of the paper is organized as follows. Section 2 introduces the guessing game and theoretical solution of its basic version. Section 3 presents the experimental studies with their contributions and critiques to the theory. Section 4 presents some extensions and Section 5 concludes with a brief discussion. II. Guessing Game and Its Game Theoretical Solution In this section, the logic and basic structure of the guessing game and its game theoretical solution will be presented. In a typical guessing game, there are N

( N  2, N   ) players and each of them chooses simultaneously a number (ni, i = 1, 2, 3, …, N) from a closed interval, frequently [0, 100]. The winner of a fixed and known prize is the player whose number is close to the target number (T) that is calculated as follows: N

T  [( ni / N )  d ]* p i 1

where d is a constant. In its simplest form5 0 < p < 1 and d is zero. p and d are common knowledge. In case of a tie, prize it is divided among the players who guess the target number. Other players receive nothing. In other words, players guess a number from a certain interval and the one whose guess is the closest to a certain fraction (p) of the average number wins. Guessing game has been mainly used to analyze the players’ reasoning process and the availability of boundedly rational behavior (see Güth, Kocher and Sutter, 2002). In other words, it can be said that “The structure of the game is favorable for investigating whether and how a player’s mental process incorporates the behavior of the other players in conscious reasoning.6” __________________________________________________________________________________________________________________________________ 5

6

In the case where p=1, all numbers can be the equilibrium. This kind of game is called “coordination game”. Rosemarie Nagel, “Unraveling in guessing games: an experimental study”, American Economic Review 85, 1995, p.1313.

A Game Theoretic Approach to Model Financial Markets: Guessing Game

47

We now consider a strategic game of N ( N  2) players where each 

player chooses a number between 0 and X ( X  , X  R ) , [0, X]. For

d  0 and 0 0). It means subjects do not understand how to play the game.

54

Ü. Barış Urhan & Zafer Akın

IV. What is left? Discussion for the Future Experiments Although it seems that many critique questions about guessing games have been answered since 1995, there are still some “blur” parts that need further research. Some ideas for future experiments are as follows: Cultural Effect: Kachelmeier and Shehata (1992) study the cultural effects on a market situation where subjects are either buyers or sellers. This effect has never been studied in a guessing game. A possible set up can be as follows: There can be a group of 4 subjects (1 East + 3 West) against a group of 4 subjects (3 East + 1 West). Only the information about the winner’s identity and the winning number can be given at the end of each period as in Patrizia (2008). In this kind of setup, it can be investigated whether people follow only the people from their own culture or they do not follow anyone at all when they are not the winner of that period. It is also possible to see whether the source of information (coming from the subjects of their own culture or from others) affects the behavior of subjects. IQ Level Differences: Some studies show that in heterogeneous groups, mathematically talented subjects converge to the equilibrium faster than others (Kovac et al., 2007). To see the effect of “being smart” can be studied broadly by using subjects of different IQ levels. For a clear overview, it may also be useful to have subjects of different ages. Mainly, there might be 2 different groups: in one group 6-8 years old subjects who have not studied mathematic in a systematic way (i.e. in the school) can play against another subject group which is a group of students at least 16-18 years old who have enough practice of mathematics. Using cross-level24 setup, it would be possible to see the effect of mathematical practicing and the level of IQ. Asymmetric Players: In every stock market, there are different types (spectrum) of players in terms of their market power. Strong/big players are the ones who have the financial power to affect the price of an asset more than smaller players whereas small players are the most common players in stock markets. When relatively small players make transactions, in addition to following the market movements, they also follow the big players’ or their advisors’ moves. Until now, none of the guessing game experiments examined the game from this perspective. In a design where one player has power to choose more than one number, it is easy to mimic the real life in this sense. In __________________________________________________________________________________________________________________________________ 24

6-8 years old subjects with higher IQ against to the same ages and higher age subjects with lower IQ levels and reverse.

A Game Theoretic Approach to Model Financial Markets: Guessing Game

55

this environment, whether big players have a significant effect on relatively small players’ choices and being big means always winning can be investigated. The Power of History: Studies show that announcing relevant information, e.g. the winning number, has an effect on people’s guesses. When we closely look at the stock market, it is obvious that both the quality and the quantity of information are huge and its update-speed is enormous. Therefore, in a simple guessing game setup, it would be good to engage subjects with full information, i.e. winner’s number, target number, guesses of all players in all previous periods, to see whether there is a significant correlation between the quantity of information they get and the quality of guesses they make. V. Discussion and Conclusion Guessing game has been extensively studied for more than 10 years by many researchers. This game is very simple in terms of both conducting an experiment and controlling the environment. Moreover, it is very useful to understand the reasoning process of players in financial markets. General findings of the literature are: i) subjects learn by playing repeatedly, ii) the speed of convergence to the equilibrium depends on the information they have and the environment, and iii) Experience has no significant effect on the depth of reasoning level but has a positive effect on earned payoffs. There is no doubt that the behavior of the players i.e. investors and the mechanism which adjusts prices in stock markets are more complex than in this game. Guessing game is just an attempt to simply model how do players guess how the other players will behave and the interaction between players in stock markets. This model does not incorporate any information related to the stocks or their real value. This game focuses on how people’s beliefs and guesses, given the information, range, p values etc., affect the outcome. It is assumed that the prices fully reflect all the relevant information in stock markets in which the information complexity and diversity is so high that both acquiring and processing information is much more difficult than what the guessing game models. While only the person who has the closest guess to the target number is awarded in the game (a discrete type reward scheme), it is enough to guess the direction of the price correctly in stock exchanges and all the ones who guess the direction right are awarded (a continuous type reward structure).25 In spite of this, the incentives to guess the target / price direction correctly are similar. __________________________________________________________________________________________________________________________________ 25

Costa-Gomes and Crawford (2006), Güth et al. (2002) and Kocher and Sutter (2006) employed this continuous reward structure such that they reward all participants based upon the absolute distance between their guess and the target number. The experiment conducted by Güth et al. (2002) showed that subjects try to avoid extreme choices and in games with interior equilibria, convergence is faster than in games with boundary equilibria.

56

Ü. Barış Urhan & Zafer Akın

The points mentioned above can be seen as a lack of comprehension of the game in modeling real behavior. However, the findings of the research on guessing games somewhat shed light on how prices are formed and trader behavior in stock exchanges. The finding that subjects learn by playing repeatedly implies that as players buy and sell in the stock market and positive and negative payoffs are realized, they may train themselves in terms of rational reasoning. Repeated play helps people comprehend the rationale that the price is a function of all players’ actions and that the recursive thinking is critical26. The second finding implies that the information through feedback that the players are given plays a critical role in the speed of convergence to the equilibrium. What this may mean for the stock market is that naive players who do not base their decisions on available information are disadvantageous relative to the ones who seek information on the player structure; the financial status of the public corporation (firm) whose shares you trade, managerial situation and risks of the firm, and who base their trades on the analysis of this kind of information. Nonetheless, the most important factors that should determine how to make trade are the information and beliefs about the other players and about how they take positions. The reason why the information about firm’s financial status etc. and analysis of this information are important is that especially the institutional and big investors intensely incorporate this information into their decision making process and this is publicly known. The last point emphasizes the importance of experience and states that experience has no significant effect on the depth of reasoning level but makes you earn more. There may be different reasons of this result one of which can be that people do not fully understand the strategic motive that lies behind the game they play. The other reason, related but possibly more appealing one, is that people may think that the others do not fully understand the game. Given this, the best response is to behave according to what you believe about how others will behave. Actually, the related result that if any experienced subjects are replaced with the inexperienced ones, experienced players stop to take decisions towards equilibrium is a good indicator of this way of thinking. This is similar to what happens in the stock markets as well. If you think people will behave in a certain way, then it is optimal to respond in a particular way (e.g., if you believe people think that price a certain stock/index will skyrocket and they will intensively buy it soon, then you should buy it as well). __________________________________________________________________________________________________________________________________ 26

Of course, here we assume almost a homogenous distribution of players in terms of size. Relatively big players may engage in speculative behavior. Even in this case, main strategy is still the same and based on how the rest of the players including big ones will behave.

A Game Theoretic Approach to Model Financial Markets: Guessing Game

57

Consequently, guessing game is not only a very simple, trivial and understandable game but also rich in terms of implications and applications to real life. Its standard version can especially be used to model markets whose players focus more on dividend yield and value stocks27. Although it is not being studied as intensive as before, its current results and implications shed light on the mechanisms of decision making in strategic environments and help us understand the behavior of investors in real life.

__________________________________________________________________________________________________________________________________ 27

Guessing games conducted in different environments such as median and maximum game as in Duffy and Nagel (1997) can be interpreted as giving more or less weight to fundamental or speculative behavior. For example, the median game environment can be viewed as a game where players focus relatively more on market fundamentals such as the fraction p or the previous period’s market average, and are not interested in speculating about the future actions of other players. On the other hand, the maximum game can be considered as a game where players are relatively more concerned with speculating about the future actions of other players who choose high numbers and they do not focus on market fundamentals much.

58

Ü. Barış Urhan & Zafer Akın

References Barberis, Nicholas; Thaler, Richard, “A Survey of Behavioral Finance in Advances in Behavioral Finance, Volume II”, (Ed.), Barberis, Nicholas and Thaler, Richard, Princeton University Press, 2005, pp. 1-74. Bosch-Domènech Antoni; García-Montalvo, Jose; Nagel, Rosemarie; Antoni, Satorra Albert, “One, Two, (Three), Infinity: Newspaper and Lab Beauty-Contest Experiments”, The American Economic Review, Vol. 92, No 5, 2002, pp. 1687-1701. Camerer, Colin F., Behavioral Game Theory, Princeton University Press, 2003. Costa-Gomes, M. A.; Crawford, V. P., “Cognition and Behavior in Two-Person Guessing Games: An Experimental Study”, American Economic Review, 96, 2006, pp. 1737-1768. Chou, Eileen; McConnell, Margaret; Nagel, Rosemarie; Plott, Charles R., “The Control of Game Form Recognition in Experiment: Understanding Dominant Strategy Failures in a Simple Two Person ’Guessing’ Game”, Experimental Economics, 12, 2009, pp. 159-179. Dasgupta, Amil, “Order Beliefs in Applied Game Theory”, [Online]; available from http://fmg.lse.ac.uk/~amil/research.html, accessed 03.03.2010. Dickinson, David L.; McElroy, Todd, "Rationality Around the Clock. Sleep and Time-Of-Day Effects on Guessing Game Response", Working Paper, 2009. Duffy, John; Nagel, Rosemarie, "On the Robustness of Behavior in Experimental 'Beauty Contest' Games", The Economic Journal, Vol. 107, November 1997, pp.1684-1700. Gibbons, Robert, Primer in Game Theory, Prentice Hall, 2002. Grosskopf, Brit; Nagel, Rosemarie, “The Two-Person Beauty Contest”, Games and Economic Behavior, 62 (1), 2008, pp. 93-99. Grosskopf, Brit; Nagel, Rosemarie, “Unraveling the Two-Person Beauty Contest”, Working Paper, November 2009. Guth, Werner; Kocher, Martin; Sutter, Matthias, “Experimental ‘Beauty Contests’ With Homogeneous and Heterogeneous Players and With Interior and Boundary Equilibria”, Economics Letters, Vol. 74, 2002, pp. 219-228. Ho, Teck; Camerer, Colin; Keith, Weigelt, “Iterated Dominance and Iterated Best Response in Experimental ‘P-Beauty-contests”, American Economic Review, Vol. 88, 1998, pp. 947-969. Kachelmeier, J. Steven; Shehata, Mohamed, “Culture and Competition: A Laboratory Market Comparison Between China and the West”, Journal of Economic Behavior and Organization, Vol. 19, 1992, pp. 145-168. Kahneman, Daniel; Tversky, Amor, "Rational Choice and the Framing of Decisions", Journal of Business, Vol. 59, 1986, pp. 251-278. Keynes, John Maynard, “The General Theory of Employment, Interest and Money”, London: MacMillan, 1947.

A Game Theoretic Approach to Model Financial Markets: Guessing Game

59

Kocher, Martin G.; Sutter, Matthias, “Time is Money—Time Pressure, Incentives, and the Quality of Decision-Making”, Journal of Economic Behavior & Organization, Vol. 61, 2006, pp. 375-392. Kocher, Martin; Straub, Sabine; Sutter, Matthias, “Individual or Team DecisionMaking -Causes and Consequences of Self-Selection”, Games and Economic Behavior, Vol. 56, 2006, pp. 259-270. Kocher, Martin; Sutter, Matthias; Wakolbinger, Florian, “The Impact of Naive Advice and Observational Learning in Beauty-Contest Games”, Tinbergen Institute Discussion Paper, January 24, 2007. Kovac, Eugen; Ortmann, Andreas; Vojtek, Martin, "Comparing Guessing Games with Homogeneous and Heterogeneous Players: Experimental Results and a CH Explanation", Economics Institute of the Czech Academy of Sciences Working Paper, March 2007. Mas-Colell, Andreu; Whinston, Michael D.; Green, Jerry R., “Microeconomic Theory”, Oxford University Press, 1995. Morone, Andrea; Morone, Piergiuseppe, “Guessing Games and People Behaviors: What Can We Learn?” , S.E.R.I.E.S. Working Paper No:15, 2006. Moulin, Hervé, “Game Theory for Social Sciences”, New York University Press, 1986. Nagel, Rosemarie, “Unraveling in Guessing Games: an Experimental Study”, American Economic Review, Vol. 85, 1995, pp. 1313-1326. Ohtsubo, Yohsuke; Rapoport, Amnon, “Depth of Reasoning in Strategic Form Games”, The Journal of Socio-Economics, Vol. 35, 2006, pp. 31-47. Olsen, Robert A., “Behavioral Finance and its Implications for Stock-Price Volatility” , Financial Analysts Journal, Vol. 54, No. 2, March-April, 1998, pp. 10-18. Sbriglia, Patrizia, “Revealing the Depth of Reasoning in p-Beauty Contest Games”, Experimental Economics, Volume 11, No. 2, 2008, pp. 107121. Shleifer, Andrei, “Inefficient markets: An introduction to Behavioral Finance”, Oxford University Press, April 2000. Slonim, R. L., “Competing Against Experienced and Inexperienced Players” Experimental Economics, Vol. 8, 2005, pp. 55-75. Sonnemans, Joep; Tuinstra, Jan, "Positive Expectations Feedback Experiments and Number Guessing Games as Models of Financial Markets", Tinbergen Institute Discussion Paper, August 22, 2008. Stahl, Dale O., “Boundedly Rational Rule Learning in a Guessing Game”, Games and Economic Behavior, Vol. 16, 1996, pp. 303-330. Sutter, Matthias, “Are Four Heads Better than Two? An Experimental BeautyContest Game with Teams of Different Size," Economics Letters, Vol. 88 (1), 2005, pp. 41-46. Tyran, Jean-Robert, “Unpublished Lecture Notes”, MA Course-Behavioral and Experimental Economics Fall 2006.

60

Ü. Barış Urhan & Zafer Akın

Weber, Roberto A., “Learning With No Feedback in a Competitive Guessing Game”, Games and Economic Behavior, Vol. 44, 2003, pp. 131-144. Woodworth, R.S., “Accuracy of Voluntary Movements”, Psychological Review, 1899, 3, pp. 1–101.

The ISE Review Volume: 12 No: 47 ISSN 1301-1642 © ISE 1997

GLOBAL CAPITAL MARKETS The global activity expanded at an annualized rate of over 3 ½ percent in the third quarter of 2010. In advanced countries, economic activity increased slower than the previous quarter. Growth is subdued, unemployment is still high and renewed stresses in the Euro area periphery are contributing to the downside risks. Real estate markets and household income were still weak in some major advanced economies (e.g. United States). The third-quarter global growth rate was better than forecast in the October 2010 IMF World Economic Outlook, due to stronger-than-expected consumption in the United States and Japan. Growth in emerging and developing economies remained strong in the third quarter due to strong private demand, still-accommodative policy stances and resurgent capital inflows. Most developing countries, particularly in sub-Saharan Africa, are also growing strongly. Global output is projected to expand by 4 ½ percent in 2011. During the second half of 2010, global financial conditions broadly improved, amid lingering vulnerabilities. Equity markets in advanced and emerging market countries rose and bank lending conditions in major advanced economies became less tight, even for small and medium sized-firms. The performances of some developed stock markets with respect to indices indicated that DJIA, FTSE-100, Nikkei-225 and DAX changed by 3.9%, 0.7%, 0.9% and -0.6%, respectively, at Sept. 29th, 2010 in comparison with the December 31, 2009. When US $ based returns of some emerging markets are compared in the same period, the best performer markets were: Indonesia (45.0 %), Thailand (44.5 %), Colombia (43.7 %), Chile (40.3 %) and Turkey (27.7 %) and in the same period, the lowest return markets were: Greece (-36.7 %), Hungary (-0.8 %) and Czech Rep. (2.1 %), and the performances of emerging markets with respect to P/E ratios as of end of Sept. 2010 indicated that the highest rates were obtained in Czech Rep. (20.6), S. Africa (20.2), Indonesia (18.4), and Mexico (17.8) and the lowest rates in Russia (8.2), Philippines (9.0), Turkey (11.7) and Chile (11.7).

62

ISE Review

Market Capitalization (USD $ Million, 1986-2009) Global

Developed Markets

Emerging Markets

6,514,199 6,275,582 238,617 1986 7,830,778 7,511,072 319,706 1987 9,728,493 9,245,358 483,135 1988 11,712,673 10,967,395 745,278 1989 9,398,391 8,784,770 613,621 1990 11,342,089 10,434,218 907,871 1991 10,923,343 9,923,024 1,000,319 1992 14,016,023 12,327,242 1,688,781 1993 15,124,051 13,210,778 1,913,273 1994 17,788,071 15,859,021 1,929,050 1995 20,412,135 17,982,088 2,272,184 1996 23,087,006 20,923,911 2,163,095 1997 26,964,463 25,065,373 1,899,090 1998 36,030,810 32,956,939 3,073,871 1999 32,260,433 29,520,707 2,691,452 2000 27,818,618 25,246,554 2,572,064 2001 23,391,914 20,955,876 2,436,038 2002 31,947,703 28,290,981 3,656,722 2003 38,904,018 34,173,600 4,730,418 2004 43,642,048 36,538,248 7,103,800 2005 54,194,991 43,736,409 10,458,582 2006 64,563,414 46,300,864 18,262,550 2007 35,811,160 26,533,854 9,277,306 2008 48,713,724 34,907,166 13,806,558 2009 Source: Standard & Poor’s Global Stock Markets Factbook, 2010.

ISE 938 3,125 1,128 6,756 18,737 15,564 9,922 37,824 21,785 20,782 30,797 61,348 33,473 112,276 69,659 47,150 33,958 68,379 98,299 161,537 162,399 286,572 117,930 235,996

Comparison of Average Market Capitalization Per Company (USD Million, Sept 2010) 8000 7000 6000 5000 4000 3000 2000 1000 0

Source: www.world-exchanges.org

964.1

Global Capital Markets

63

Worldwide Share of Emerging Capital Markets (1986-2009) 60% Market Capitalization (%) Trading Volume (%)

50%

Number of Companies (%)

40%

30%

20%

10%

0%

Source: Standard & Poor’s Global Stock Markets Factbook, 2010.

Share of ISE’s Market Capitalization in World Markets (1986-2009) 4.0%

0.80%

3.5%

0.70%

3.0%

0.60%

2.5%

0.50%

2.0%

0.40%

1.5%

0.30%

1.0%

0.20%

0.5%

0.10%

0.0%

0.00%

S h ar e i n Em e rg i ng M a rk et s

S h ar e i n De ve l o p e d M a r ke t s

Source: Standard & Poor’s Global Stock Markets Factbook, 2010.

64

ISE Review

Main Indicators of Capital Markets (Sept 2010) Market

Monthly Turnover Velocity (Sept 2010) (%)

1 2 3 4 5 6

Shenzhen SE NASDAQ OMX Shanghai SE Korea Exchange Taiwan SE Corp. Istanbul SE

336.1% 277.4% 170.4% 149.4% 146.9% 123.0%

7

NYSE Euronext (US)

119.1%

8 9

Deutsche Börse Tokyo SE

115.3% 101.6%

10 Oslo Børs

90.7%

11 Australian SE BME Spanish 12 Exchanges 13 TSX Group 14 Budapest SE 15 SIX Swiss Exchange

85.7%

16 London SE Group

71.5%

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

NYSE Euronext (Europe) Osaka SE Hong Kong Exchanges National Stock Exchange India Singapore Exchange Athens Exchange MICEX Colombo SE Warsaw SE Egyptian Exchange Tel Aviv SE Tehran SE Johannesburg SE Wiener Börse Bursa Malaysia Philippine SE Mexican Exchange Santiago SE Bombay SE Colombia SE Irish SE Cyprus SE Ljubljana SE Mauritius SE Buenos Aires SE Lima SE Bermuda SE Malta SE Luxembourg SE

85.3% 82.4% 74.3% 73.2%

Market

Value of Share Trading (millions, US$) Up to Year Total (2010/12010/9)

Market

Market Cap. of Share of Domestic Companies (millions US$) Sept 2010

NYSE Euronext (US) 13,714,901 NYSE Euronext (US) 12,278,418.6 NASDAQ OMX 9,998,008 NASDAQ OMX 3,480,718.2 Shanghai SE 2,858,264 London SE Group 3,467,570.2 Tokyo SE 2,815,578 Tokyo SE 3,423,744.6 Shenzhen SE 2,287,852 NYSE Euronext (Europe) 2,839,828.0 London SE Group 2,123,286 Hong Kong Exchanges 2,535,956.4 NYSE Euronext 1,557,428 Shanghai SE 2,468,793.1 (Europe) Deutsche Börse 1,283,937 TSX Group 1,867,609.0 Korea Exchange 1,150,581 Bombay SE 1,585,803.3 Hong Kong National Stock Exchange 1,027,601 1,548,577.6 Exchanges India TSX Group 993,223 Australian SE 1,318,559.3 BME Spanish 966,754 Deutsche Börse 1,292,630.3 Exchanges Australian SE 786,130 BME Spanish Exchanges 1,205,620.1 Taiwan SE Corp. 643,835 SIX Swiss Exchange 1,115,854.1 SIX Swiss Exchange 601,346 Shenzhen SE 1,099,182.2 National Stock 575,107 Korea Exchange 987,651.7 Exchange India

69.8% MICEX

304,428 MICEX

768,226.7

65.9% Istanbul SE

291,440 Johannesburg SE

748,113.3

61.6% Johannesburg SE

249,556 Taiwan SE Corp.

684,977.1

56.7% Singapore Exchange

204,571 Singapore Exchange

597,332.3

54.0% 46.9% 45.4% 45.3% 45.2% 39.1% 39.0% 35.8% 33.8% 32.7% 32.6% 27.0% 22.8% 18.1% 17.6% 15.0% 13.7% 11.2% 5.0% 4.6% 4.4% 3.8% 1.5% 0.8% 0.2%

195,423 191,011 137,289 86,683 76,844 74,293 49,297 36,400 35,536 34,634 30,467 21,039 18,219 14,821 14,589 6,893 3,533 2,523 2,336 585 341 278 162 66 33

388,570.8 368,023.8 320,219.1 319,124.8 243,857.4 215,585.7 209,316.2 206,930.1 183,584.2 130,085.5 110,196.8 89,164.8 85,947.9 84,872.4 78,773.6 73,025.4 57,692.5 48,543.3 30,199.3 20,623.1 9,692.1 8,114.0 7,151.1 3,820.0 1,688.8

Oslo Børs Bombay SE Osaka SE Mexican Exchange Bursa Malaysia Tel Aviv SE Warsaw SE Wiener Börse Athens Exchange Santiago SE Egyptian Exchange Budapest SE Colombia SE Philippine SE Tehran SE Irish SE Colombo SE Lima SE Buenos Aires SE Cyprus SE Ljubljana SE Mauritius SE Luxembourg SE Bermuda SE Malta SE

Source: www.world-exchanges.org

Mexican Exchange Bursa Malaysia Santiago SE Istanbul SE Osaka SE Oslo Børs Colombia SE Tel Aviv SE Warsaw SE Philippine SE Wiener Börse Luxembourg SE Tehran SE Lima SE Egyptian Exchange Athens Exchange Irish SE Buenos Aires SE Budapest SE Colombo SE Ljubljana SE Cyprus SE Mauritius SE Malta SE Bermuda SE

Global Capital Markets

65

Trading Volume (USD Milions, 1986-2009) Global

Developed

Emerging

ISE

Emerging/Global (%) ISE/Emerging (%)

3,573,570 3,490,718 82,852 13 1986 5,846,864 5,682,143 164,721 118 1987 5,997,321 5,588,694 408,627 115 1988 7,467,997 6,298,778 1,169,219 773 1989 5,514,706 4,614,786 899,920 5,854 1990 5,019,596 4,403,631 615,965 8,502 1991 4,782,850 4,151,662 631,188 8,567 1992 7,194,675 6,090,929 1,103,746 21,770 1993 8,821,845 7,156,704 1,665,141 23,203 1994 52,357 1995 10,218,748 9,176,451 1,042,297 37,737 1996 13,616,070 12,105,541 1,510,529 59,105 1997 19,484,814 16,818,167 2,666,647 68,646 1998 22,874,320 20,917,462 1,909,510 81,277 1999 31,021,065 28,154,198 2,866,867 2000 47,869,886 43,817,893 3,967,806 179,209 77,937 2001 42,076,862 39,676,018 5,604,092 70,667 2002 38,645,472 36,098,731 8,226,944 99,611 2003 29,639,297 26,743,153 2,896,144 2004 39,309,589 35,341,782 3,967,806 147,426 2005 47,319,584 41,715,492 5,604,092 201,258 2006 67,912,153 59,685,209 8,226,944 227,615 2007 98,816,305 82,455,174 16,361,131 302,402 2008 80,516,822 67,795,950 12,720,872 239,713 2009 80,418,059 64,458,380 15,959,679 316,326 Source: Standard & Poor’s Global Stock Markets Factbook, 2010.

2.32 2.82 6.81 15.66 16.32 12.27 13.20 15.34 18.88 10.20 11.09 13.69 8.55 9.24 8.46 5.71 6.59 9.77 10.09 11.84 12.11 16.56 15.80 19.85

0.02 0.07 0.03 0.07 0.65 1.38 1.36 1.97 1.39 5.02 2.50 2.18 3.60 2.86 4.42 3.25 2.77 3.44 3.72 3.59 2.77 1.85 1.88 1.98

Number of Trading Companies (1986-2009) Global

Developed Markets

Emerging Markets

ISE

28,173 18,555 9,618 80 1986 29,278 18,265 11,013 82 1987 29,270 17,805 11,465 79 1988 25,925 17,216 8,709 76 1989 25,424 16,323 9,101 110 1990 26,093 16,239 9,854 134 1991 27,706 16,976 10,730 145 1992 28,895 17,012 11,883 160 1993 33,473 18,505 14,968 176 1994 36,602 18,648 17,954 205 1995 40,191 20,242 19,949 228 1996 40,880 20,805 20,075 258 1997 47,465 21,111 26,354 277 1998 48,557 22,277 26,280 285 1999 49,933 23,996 25,937 315 2000 48,220 23,340 24,880 310 2001 48,375 24,099 24,276 288 2002 49,855 24,414 25,441 284 2003 48,806 24,824 23,982 296 2004 49,946 25,337 24,609 302 2005 50,212 25,954 24,258 314 2006 51,322 26,251 25,071 319 2007 49,138 26,375 22,763 284 2008 48,561 24,635 23,926 267 2009 Source: Standard & Poor’s Global Stock Markets Factbook, 2010.

Emerging/Global (%)

34.14 37.62 39.17 33.59 35.80 37.76 38.73 41.12 44.72 49.05 49.64 49.11 55.52 54.12 51.94 51.60 50.18 51.03 49.14 49.27 48.31 48.85 46.32 49.27

ISE/Emerging (%)

0.83 0.74 0.69 0.87 1.21 1.36 1.35 1.35 1.18 1.14 1.14 1.29 1.05 1.08 1.21 1.25 1.19 1.12 1.23 1.23 1.29 1.27 1.25 1.12

66

ISE Review

Comparison of P/E Ratios Performances Argentina Brazil Chile China Czech Rep. Hungary India Indonesia Jordan Korea Malaysia Mexico Pakistan Peru Philippines Poland Russia S.Africa Taiwan Thailand Turkey

0.0

10.0

20.0

2010/6

30.0

40.0

2009

50.0

60.0

2008

Source: Standard & Poor’s, Bloomberg.

Price-Earnings Ratios in Emerging Markets 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010/9 32.6 -1.4 21.1 27.7 11.1 18.0 13.6 3.4 13.0 12.3 Argentina 8.8 13.5 10.0 10.6 10.7 12.7 16.6 5.3 21.7 15.4 Brazil 16.2 16.3 24.8 17.2 15.7 24.2 22.3 11.5 18.9 20.2 Chile 22.2 21.6 28.6 19.1 13.9 24.6 50.5 8.6 22.8 17.5 China 5.8 11.2 10.8 25.0 21.1 20.0 26.5 10.5 14.6 11.7 Czech Rep. 13.4 14.6 12.3 16.6 13.5 13.4 14.0 4.2 12.2 13.2 Hungary 12.8 15.0 20.9 18.1 19.4 20.1 31.6 8.6 23.2 18.4 India -7.7 22.0 39.5 13.3 12.6 20.1 31.7 7.0 27.3 20.6 Indonesia 18.8 11.4 20.7 30.4 6.2 20.8 28.0 10.9 34.8 N/A Jordan 28.7 21.6 30.2 13.5 20.8 12.8 16.4 6.4 15.9 13.2 Korea 50.6 21.3 30.1 22.4 15 21.7 20.1 4.2 22.6 17.8 Malaysia 13.7 15.4 17.6 15.9 14.2 18.6 17.2 0.3 18.3 15.8 Mexico 7.5 10.0 9.5 9.9 13.1 10.8 15.3 3.0 11.2 9.0 Pakistan 21.3 12.8 13.7 10.7 12.0 15.7 20.9 7.7 28.6 N/A Peru 45.9 21.8 21.1 14.6 15.7 14.4 17.7 8.2 13.4 13.3 Philippines 6.1 88.6 -353.0 39.9 11.7 13.9 15.6 6.4 23.0 15.7 Poland 5.6 12.4 19.9 10.8 24.1 16.6 18.4 3.4 14.3 8.2 Russia 11.7 10.1 11.5 16.2 12.8 16.6 18.7 7.5 18.2 15.0 S.Africa 29.4 20.0 55.7 21.2 21.9 25.6 27.9 7.2 17.1 14.4 Taiwan 163.8 16.4 16.6 12.8 10.0 8.7 11.7 7.5 11.9 14.3 Thailand 72.5 37.9 14.9 12.5 16.2 17.2 25.2 3.2 11.4 11.7 Turkey Source: IFC Factbook, 2004-2007; Standard & Poor’s & Bloomberg. Note: Figures are taken from S&P/IFCI Index Profile.

Global Capital Markets

67

Comparison of Market Returns in USD (31/12/2009-29/09/2010) 45.0 44.5 43.7 40.3 27.7 27.5 18.3 14.3 13.1 11.0 10.3 9.7 9.0 8.1 4.4 4.4 3.9 3.5 3.1 3.0 3.0 2.3 2.1 -0.8 -36.7

-50.0

-40.0

-30.0

-20.0

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

nd

Source: The Economist, Oct. 2 2010.

Market Value/Book Value Ratios 2001

2002

2003

2004

2005

2006

0.6 0.8 2.0 2.2 2.5 4.1 Argentina 1.2 1.3 1.8 1.9 2.2 2.7 Brazil 1.4 1.3 1.9 0.6 1.9 2.4 Chile 2.3 1.9 2.6 2.0 1.8 3.1 China 0.8 0.8 1.0 1.6 2.4 2.4 Czech Rep. 1.8 1.8 2.0 2.8 3.1 3.1 Hungary 1.9 2.0 3.5 3.3 5.2 4.9 India 1.7 1.0 1.6 2.8 2.5 3.4 Indonesia 1.5 1.3 2.1 3.0 2.2 3.3 Jordan 1.2 1.1 1.6 1.3 2.0 1.7 Korea 1.2 1.3 1.7 1.9 1.7 2.1 Malaysia 1.7 1.5 2.0 2.5 2.9 3.8 Mexico 0.9 1.9 2.3 2.6 3.5 3.2 Pakistan 1.4 1.2 1.8 1.6 2.2 3.5 Peru 0.9 0.8 1.1 1.4 1.7 1.9 Philippines 1.4 1.3 1.8 2.0 2.5 2.5 Poland 1.1 0.9 1.2 1.2 2.2 2.5 Russia 2.1 1.9 2.1 2.5 3.0 3.8 S.Africa 2.1 1.6 2.2 1.9 1.9 2.4 Taiwan 1.3 1.5 2.8 2.0 2.1 1.9 Thailand 3.8 2.8 2.6 1.7 2.1 2.0 Turkey Source: IFC Factbook, 2004-2007; Standard & Poor’s & Bloomberg Note: Figures are taken from S&P/IFCI Index Profile.

2007 3.2 3.3 2.5 6.3 3.1 3.2 7.9 5.6 4.4 2.2 2.5 3.6 4.7 6.0 2.8 2.8 2.8 4.4 2.6 2.5 2.8

2008 0.8 1.0 1.4 1.9 2.0 0.9 1.7 1.6 1.3 0.8 0.7 1.0 0.8 2.7 1.3 1.1 0.1 1.6 1.0 1.0 0.7

2009 1.5 2.2 2.4 3.3 1.4 1.5 3.5 2.7 1.3 1.2 2.3 2.7 1.6 5.4 2.0 1.5 1.0 2.2 2.1 1.5 1.6

2010/9 1.3 1.9 2.7 2.6 1.4 1.5 3.4 3.9 N/A 1.3 2.3 2.4 1.5 N/A 2.4 1.6 1.1 2.3 1.9 1.9 1.8

68

ISE Review

Value of Bond Trading (Million USD, Jan. 2009-Sept 2010) 7,102,547 2,894,977 1,668,778 836,688 400,466 344,900 339,170 166,331 142,920 130,992 128,135 116,031 100,852 73,724 52,507 23,441 19,689 16,616 9,363 7,819 4,275 3,300 3,131 973 792 531 514 461 452 381 129 112 62 26 16 1

Source: www.world-exchanges.org

10

100

1,000

10,000 100,000 1,000,000 10,000,000

Jan/04 Feb/04 Mar/04 Apr/04 May/04 Jun/04 Jul/04 Aug/04 Sep/04 Oct/04 Nov/04 Dec/04 Jan/05 Feb/05 Mar/05 Apr/05 May/05 Jun/05 Jul/05 Aug/05 Sep/05 Oct/05 Nov/05 Dec/05 Jan/06 Feb/06 Mar/06 Apr/06 May/06 Jun/06 Jul/06 Aug/06 Sep/06 Oct/06 Nov/06 Dec/06 Jan/07 Feb/07 Mar/07 Apr/07 May/07 Jun/07 Jul/07 Aug/07 Sep/07 Oct/07 Nov/07 Dec/07 Jan/08 Feb/08 Mar/08 Apr/08 May/08 Jun/08 Jul/08 Aug/08 Sep/08 Oct/08 Nov/08 Dec/08 Jan/09 Feb/09 Mar/09 Apr/09 May/09 Jun/09 Jul/09 Aug/09 Sep/09 Oct/09 Nov/09 Dec/09 Jan/10 Feb/10 Mar/10 Apr/10 May/10 Jun/10 Jul/10 Aug/10 Sep/10

%

Global Capital Markets

450

400

Source: Bloomberg

69

Foreigners’ Share in the Trading Volume of the ISE (Jan. 1998-Sept 2010)

35

30

25

20

15

10

5

0

Source: ISE Data.

Comparison of Market Indices (31 Jan. 2004 =100)

KOREA SINGAPORE PORTUGAL AUSTRIA

RUSIA GREECE MALAYSIA TURKEY

350

300

250

200

150

100

50

0

The ISE Review Volume: 12 No: 47 ISSN 1301-1642 © ISE 1997

ISE Market Indicators

Traded Value Total

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

80 82 79 76 110 134 145 160 176 205 228 258 277 285 315 310 288 285 297 304 316 319 317 315 331 317 325 331

TL Million 0,01 0,10 0,15 2 15 35 56 255 651 2.374 3.031 9.049 18.030 36.877 111.165 93.119 106.302 146.645 208.423 269.931 325.131 387.777 332.605 482.534 452.788 175.589 152.791 124.408

US$ Million 13 118 115 773 5.854 8.502 8.567 21.770 23.203 52.357 37.737 58.104 70.396 84.034 181.934 80.400 70.756 100.165 147.755 201.763 229.642 300.842 261.274 316.326 300.184 117.179 100.211 82.794

Market Value

Dividend Yield

Number of Comp.

STOCK MARKET P/E Ratios

Daily Average TL Million ------0,01 0,06 0,14 0,22 1 3 9 12 36 73 156 452 375 422 596 837 1.063 1.301 1.539 1.325 1.915 2.396 2.787 2.425 1.975

US$ Million ------3 24 34 34 88 92 209 153 231 284 356 740 324 281 407 593 794 919 1.194 1.041 1.255 1.588 1.860 1.591 1.314

TL Million 0,71 3 2 16 55 79 85 546 836 1.265 3.275 12.654 10.612 61.137 46.692 68.603 56.370 96.073 132.556 218.318 230.038 335.948 182.025 350.761 461.614 388.063 388.000 461.614

US$ Million 938 3.125 1.128 6.756 18.737 15.564 9.922 37.824 21.785 20.782 30.797 61.879 33.975 114.271 69.507 47.689 34.402 69.003 98.073 162.814 163.775 289.986 119.698 235.996 320.032 256.215 246.725 320.032

(%)

TL(1)

TL(2)

US$

9,15 2,82 10,48 3,44 2,62 3,95 6,43 1,65 2,78 3,56 2,87 1,56 3,37 0,72 1,29 0,95 1,20 0,94 1,37 1,71 2,10 1,90 4,93 2,37 1,62 2,41 1,92 1,62

5,07 15,86 4,97 15,74 23,97 15,88 11,39 25,75 24,83 9,23 12,15 24,39 8,84 37,52 16,82 108,33 195,92 14,54 14,18 17,19 22,02 12,16 5,55 17,89 13,45 13,85 12,92 13,45

--------------20,72 16,7 7,67 10,86 19,45 8,11 34,08 16,11 824,42 26,98 12,29 13,27 19,38 14,86 11,97 5,76 16,83 13,45 13,65 11,08 13,45

--------------14,86 10,97 5,48 7,72 13,28 6,36 24,95 14,05 411,64 23,78 13,19 13,96 19,33 15,32 13,48 4,63 17,34 14,11 13,70 10,59 14,11

Q: Quarter Note: Between 1986-1992, the price earnings ratios were calculated on the basis of the companies’ previous year-end net profits. As from 1993, TL(1) = Total Market Capitalization / Sum of Last two six-month profits TL(2) = Total Market Capitalization / Sum of last four three-month profits. US$ = based Total Market Capitalization / Sum of last four US$ based three-month profits. - Companies which are temporarily de-listed and will be traded off the Exchange under the decision of ISE’s Executive Council are not included in the calculations. - EFT’s data are taken into account only in the calculation of Traded Value.

72

ISE Review

Closing Values of the ISE Price Indices TL Based ISE 100 (Jan. 1986=1)

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

1,71 6,73 3,74 22,18 32,56 43,69 40,04 206,83 272,57 400,25 975,89 3.451,00 2.597,91 15.208,78 9.437,21 13.782,76 10.369,92 18.625,02 24.971,68 39.777,70 39.117,46 55.538,13 26.864,07 52.825,02 65.774,37 56.538,37 54.839,46 65.774,37

ISE CORPORATE GOVERNANCE

(Aug.29,2007 =48,082.17)

------------------------------------------55.406,17 21.974,49 42.669,96 55.958,30 46.860,89 46.405,00 55.958,30

ISE INDUSTRIALS (Dec.31, 90 =33)

----------49,63 49,15 222,88 304,74 462,47 1.045,91 2.660,-1.943,67 9.945,75 6.954,99 11.413,44 9.888,71 16.299,23 20.885,47 31.140,59 30.896,67 40.567,17 19.781,26 37.899,01 48.925,93 42.360,56 41.039,66 48.925,93

ISE SERVICES (Dec.27, 96 =1046)

ISE FINANCIALS

(Dec. 31, 90 =33)

----------------------3.593,-3.697,10 13.194,40 7.224,01 9.261,82 6.897,30 9.923,02 13.914,12 18.085,71 22.211,77 34.204,74 22.169,30 36.134,16 40.541,35 35.927,74 34.027,90 40.541,35

ISE TECHNOLOGY (June, 30,2000 =14.466,12)

----------33,55 24,34 191,90 229,64 300,04 914,47 4.522,-3.269,58 21.180,77 12.837,92 18.234,65 12.902,34 25.594,77 35.487,77 62.800,64 60.168,41 83.822,29 38.054,32 79.763,23 102.173,74 87.233,97 85.381,67 102.173,74

ISE INVESTMENT TRUSTS (Dec 27,1996 =976)

----------------------------10.586,58 9.236,16 7.260,84 8.368,72 7.539,16 13.669,97 10.341,85 10.490,51 4.858,62 14.335,01 15.928,38 15.851,41 14.337,38 15.928,38

ISE SECOND NATIONAL (Dec.27,1996 =976)

----------------------2.934,-1.579,24 6.812,65 6.219,00 7.943,60 5.452,10 10.897,76 17.114,91 23.037,86 16.910,76 16.428,59 8.655,55 18.215,26 20.644,38 20.461,16 18.531,10 20.644,38

----------------------2.761,-5.390,43 13.450,36 15.718,65 20.664,11 28.305,78 32.521,26 23.415,86 28.474,96 23.969,99 27.283,78 8.645,09 25.764,15 35.726,11 30.093,58 33.152,31 35.726,11

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

131,53 384,57 119,82 560,57 642,63 501,50 272,61 833,28 413,27 382,62 534,01 981,99 484,01 1.654,17 817,49 557,52 368,26 778,43 1.075,12 1.726,23 1.620,59 2.789,66 1.027,98 2.068,18 2.653,55 2.172,21 2.029,23 2.653,55

Q: Quarter

ISE CORPORATE GOVERNANCE

(Aug.29,2007 =2,114.37)

------------------------------------------2.783,03 840,87 1.670,60 2.257,54 1.800,40 1.717,13 2.257,54

ISE INDUSTRIALS

(Dec.31, 90 =643)

----------569,63 334,59 897,96 462,03 442,11 572,33 756,91 362,12 1.081,74 602,47 461,68 351,17 681,22 899,19 1.351,41 1.280,01 2.037,67 756,95 1.483,81 1.973,83 1.627,49 1.518,59 1.973,83

ISE SERVICES (Dec.27, 96 =572))

----------------------1.022,40 688,79 1.435,08 625,78 374,65 244,94 414,73 599,05 784,87 920,21 1.718,09 848,33 1.414,71 1.635,57 1.380,35 1.259,14 1.635,57

ISE

ISE

FINANCIALS

TECHNOLOGY

(Dec. 31, 90 =643)

(June 30,2000 =1.360,92)

----------385,14 165,68 773,13 348,18 286,83 500,40 2.287,-609,14 2.303,71 1.112,08 737,61 458,20 1.069,73 1.527,87 2.725,36 2.492,71 4.210,36 1.456,18 3.122,86 4.122,01 3.351,53 3.159,38 4.122,01

------------------------------------39.240,73 29.820,90 20.395,84 32.879,36 14.889,37 25.795,58 59.857,56 30.105,36 55.292,78 59.857,56

Euro Based

US $ Based ISE 100 (Jan. 1986=100)

ISE NEW ECONOMY (Sept. 02, 2004 =20525,92)

----------------------------917,06 373,61 257,85 349,77 324,59 593,24 428,45 526,93 185,92 561,24 642,60 609,01 530,53 642,60

ISE INVESTMENT TRUSTS

(Dec. 27, 96 =534)

----------------------835,-294,22 740,97 538,72 321,33 193,62 455,47 736,86 999,77 700,59 825,20 331,21 713,16 832,86 786,12 685,71 832,86

ISE SECOND NATIONAL (Dec. 27, 96 =534)

ISE NEW ECONOMY (Sept. 02, 2004 =796,46)

----------------------786,-1.004,27 1.462,92 1.361,62 835,88 1.005,21 1.359,22 1.008,13 1.235,73 993,05 1.370,45 330,81 1.008,71 1.441,30 1.156,20 1.226,74 1.441,30

------------------------------------1.689,45 1.294,14 844,98 1.651,52 569,76 1.009,94 2.414,84 1.156,65 2.046,00 2.414,84

ISE 100 (Dec. 31, 98 =484)

--------------------------1.912,46 1.045,57 741,24 411,72 723,25 924,87 1.710,04 1.441,89 2.221,77 859,46 1.682,53 2.279,47 1.890,97 1.940,04 2.279,47

ISE Market Indicators

73

BONDS AND BILLS MARKET Traded Value Outright Purchases and Sales Market Total TL Million 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

US$ Million

1 18 123 270 740 2.711 5.504 17.996 35.430 166.336 39.777 102.095 213.098 372.670 480.723 381.772 363.949 300.995 417.052 336.469 128.175 111.494 96.800

Daily Average TL Million US$ Million

312 2.406 10.728 8.832 16.509 32.737 35.472 68.399 83.842 262.941 37.297 67.256 144.422 262.596 359.371 270.183 278.873 239.367 269.977 222.367 85.410 72.715 64.242

0,01 0,07 0,50 1 3 11 22 72 143 663 158 404 852 1.479 1.893 1.521 1.444 1.199 1.655 1.780 2.035 1.770 1.537

2 10 44 35 66 130 141 274 338 1.048 149 266 578 1.042 1.415 1.076 1.107 954 1.071 1.177 1.356 1.154 1.020

Repo-Reverse Repo Market Repo-Reverse Repo Market Total TL Million 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

Q: Quarter

59 757 5.782 18.340 58.192 97.278 250.724 554.121 696.339 736.426 1.040.533 1.551.410 1.859.714 2.538.802 2.571.169 2.935.317 2.982.531 2.360.440 806.180 794.051 760.209

US$ Million 4.794 23.704 123.254 221.405 374.384 372.201 589.267 886.732 627.244 480.725 701.545 1.090.476 1.387.221 1.770.337 1.993.283 2.274.077 1.929.031 1.562.104 538.058 519.702 504.344

Daily Average TL Million US$ Million 0,28 3 23 73 231 389 1.011 2.208 2.774 2.911 4.162 6.156 7.322 10.114,75 10.203 11.694 11.835 12.489 12.796,51 12.603,99 12.066,82

22 94 489 879 1.486 1.489 2.376 3.533 2.499 1.900 2.806 4.327 5.461 7.053 7.910 9.060 7.655 8.265 8.541 8.249 8.005

74

ISE Review

ISE GDS Price Indices (January 02, 2001 = 100) TL Based 3 Months (91 Days) 102,87 105,69 110,42 112,03 113,14 111,97 112,67 112,56 114,96 114,94 115,03 114,87 114,94

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

6 Months (182 Days) 101,49 106,91 118,04 121,24 123,96 121,14 122,83 122,69 127,78 127,83 127,93 127,60 127,83

9 Months (273 Days) 97,37 104,87 123,22 127,86 132,67 127,77 130,72 130,63 138,50 138,76 138,75 138,24 138,76

12 Months (365 Days) 91,61 100,57 126,33 132,22 139,50 132,16 136,58 136,65 147,29 147,91 147,65 146,96 147,91

15 Months (456 Days) 85,16 95,00 127,63 134,48 144,47 134,48 140,49 140,81 154,03 155,14 154,52 153,66 155,14

General 101,49 104,62 121,77 122,70 129,14 121,17 128,23 128,03 131,08 136,07 133,79 132,61 136,07

ISE GDS Performance Indices (January 02, 2001 = 100) TL Based 3 Months (91 Days)

6 Months (182 Days)

9 Months (273 Days)

12 Months (365 Days)

179,24 305,57 457,60 574,60 670,54 771,08 916,30 1.083,04 1.227,87 1.299,43 1.251,79 1.273,23 1.299,43

190,48 347,66 558,19 712,26 839,82 956,21 1.146,36 1.369,76 1.558,64 1.649,67 1.592,06 1.619,36 1.649,67

159,05 276,59 438,13 552,85 665,76 760,07 917,23 1.070,37 1.247,88 1.325,73 1.277,03 1.299,00 1.325,73

195,18 314,24 450,50 555,45 644,37 751,03 887,85 1.047,38 1.165,91 1.229,21 1.186,30 1.206,63 1.229,21

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

15 Months (456 Days) 150,00 255,90 464,98 610,42 735,10 829,61 1.008,52 1.241,27 1.421,58 1.510,27 1.454,80 1.479,82 1.510,27

ISE GDS Portfolio Performance Indices (December 31, 2003 = 100) TL Based Equal Weighted Indices 2004 2005 2006 2007 2008 2009 2010 2010/Q1 2010/Q2 2010/Q3

EQ180125,81 147,29 171,02 203,09 240,13 270,34 285,72 275,28 280,19 285,72

EQ180+ 130,40 160,29 180,05 221,63 264,15 318,15 341,56 326,12 332,32 341,56

EQ Composite 128,11 153,55 175,39 211,76 251,95 293,06 312,27 299,35 304,84 312,27

Q: Quarter GDS: Government Debt securities

Market Value Weighted Indices MV180125,91 147,51 170,84 202,27 239,21 268,84 284,32 273,73 278,71 284,32

MV180+ 130,25 160,36 179,00 221,13 263,57 317,82 341,29 325,80 332,13 341,29

MVComposite 128,09 154,25 174,82 212,42 252,36 295,43 315,10 301,95 307,53 315,10

REPO 118,86 133,63 152,90 177,00 203,07 219,59 228,86 222,50 225,61 228,86

ISE PUBLICATIONS I- PERIODICALS

ISSN/ISBN

ISE Review*

ISSN 1301-1642 ISSN 1301-1650

DATE

ISE Finance Award Series Volume 4*

ISBN 975-6450-12-6

2005

Corporate Governance Compliance Rating – Hakan Güçlü

978-975-6450-43-7

2010

The Impact of the 1929 Depression on the Istanbul Stocks and BondsForeign Currency Exchange – Dr. Korkmaz Ergun

978-975-6450-42-0

2010

The Impact of Trading Statements by Principal Shareholders and Managers on Their Own Company Share Prices in the Istanbul Stock Exchange.– Selma Kurtay

978-975-6450-24-6

2009

The Role of Financial Markets on Inflation Targeting: Analysis of Correlation between Stock Returns and Inflation – Dr. Cahit Sönmez

978-975-6450-17-8

2007

Fixed Income Decision Analysis with Excel/VBA Models* – Prof. Dr. A. Gültekin Karaşin

978-975-6450-20-8

2007

Performance Evaluation of Real Estate Investment Trusts: The Case of Turkey – Dr. Feyzullah Yetgin

975-6450-14-2

2006

Conjunctive Fluctuations and Capital Markets – The Case of Turkey – Dr. Eralp Polat

975-6450-10-X

2005

Cross-sectional Anomalies in Stock Markets and a Research on the ISE – M. Volkan Öztürkatalay

975-6450-11-8

2005

Role of Financial Market Imperfections in Firm Level Investment: Panel Data Evidence from Turkish Corporations* – Bahşayiş Temir-Fıratoğlu

975-6450-08-8

2004

Depository Certificates within Turkish Law – Dr. Korkut Özkorkut

975-6450-06-1

2003

Developments in International Financial Markets and in Turkey Assoc. Prof. Ali Alp

ISBN 975-6450-03-7

2002

Evaluation of Mutual Fund Performance in Turkey - Saim Kılıç

ISBN 975-6450-00-2

2002

Duty of Loyalty of the Shareholder in Corporate Law, in particular, in the Incorporated Companies - Dr. Murat Yusuf Akın

ISBN 975-8027-99-9

2002

Political Economy of Natural Disasters - Assoc. Prof. Enver Alper Güvel

ISBN 975-8027-91-3

2001

An Analysis of Factors Influencing Accounting Disclosure in Turkey* Dr. Turgut Çürük

ISBN 975-8027-89-1

2001

Fund Management in the Insurance Sector - Prof. Dr. Niyazi Berk

ISBN 975-8027-86-7

2001

The Changing Role of the Central Bank of Turkey and Monetary Policy Implementation - Dr. Mehmet Günal

ISBN 975-8027-85-9

2001

Financial Asset Valuation Models and Testing of Arbitrage Pricing Model on the ISE - Nevin Yörük

ISBN 975-8027-77-8

2000

Stationary Portfolio Analysis and its Implementation on ISE Data İbrahim Engin Üstünel

ISBN 975-8027-76-X

2000

Seasonalities in Stock Markets and an Empirical Study on the Istanbul Stock Exchange – Dr. Recep Bildik

ISBN 975-8027-73-5

2000

Real Estate Financing and Valuation - Dr. Ali Alp, M. Ufuk Yılmaz

ISBN 975-8027-72-7

2000

South Asian Crisis: The Effects on Turkish Economy and the ISE Research Department

ISBN 975-8027-44-1

1998

Instutional Investors in the Capital Markets - Dr. Oral Erdoğan, Levent Özer

ISBN 975-8027-51-4

1998

II- RESEARCH PUBLICATIONS

ISE PUBLICATIONS What Type of Monetary System? Monetary Discipline and Alternative Resolutions for Monetary Stability - Prof. Dr. Coşkun Can Aktan, Dr. Utku Utkulu, Dr. Selahattin Togay

ISBN 975-8027-47-6

1998

Analysis of Return Volatility in the Context of Macroeconomic Conjuncture - Prof. Dr. Hurşit Güneş, Dr. Burak Saltoğlu

ISBN 975-8027-32-8

1998

Private Pension Funds: Chilean Example - Çağatay Ergenekon

ISBN 975-8027-43-3

1998

Equity Options and Trading on the ISE - Dr. Mustafa Kemal Yılmaz

ISBN 975-8027-45-X

1998

Resolution of Small and Medium Size Enterprises' Financial Needs Through Capital Markets - R. Ali Küçükçolak

ISBN 975-8027-39-5

1998

Regulations Related to Capital Market Operations - Vural Günal

ISBN 975-8027-34-4

1997

Strategic Entrepreneurship: Basic Techniques for Growth and Access to Foreign Markets for Turkish Companies - Ömer Esener

ISBN 975-8027-28-X

1997

The Integration of European Capital Markets and Turkish Capital Market Dr. Sadi Uzunoğlu- Dr. Kerem Alkin, Dr. Can Fuat Gürlesel

ISBN 975-8027-24-7

1997

Insider Trading and Market Manipulations* - Dr. Meral Varış Tezcanlı

ISBN 975-8027-17-4 & ISBN 975-8027-18-2

1996

European Union and Turkey - Prof. Dr. Rıdvan Karluk

ISBN 975-8027-04-2

1996

Repo and Reverse Repo Transactions - Dr. Nuran Cömert Doyrangöl

ISBN 975-8027-12-3

1996

Fortunes Made Fortunes Lost* - Abdurrahman Yıldırım

ISBN 975-7143-10-3

1996

Personnel Administration - Şebnem Ergül

ISBN 975-8027-07-7

1996

Research Studies on Capital Markets and ISE

ISBN 975-7869-04-X

1996

The Integration of Euoropean Union Capital Markets and Istanbul Stock Exchange - Dr. Meral Varış Tezcanlı, Dr. Oral Erdoğan

ISBN 975-8027-05-0

1996

Institutional Investors in the Developing Stock Exchanges: Turkish Example, Problems and Proposed Solutions - Dr. Targan Ünal

1995

International Capital Movements and their Macroeconomic Effects on the Third World Countries and Turkey - Dr. Sadi Uzunoğlu, Dr. Kerem Alkin, Dr. Can Fuat Gürlesel

1995

Modern Developments in Investment Theory and Some Evaluations and Observations at ISE - Dr. Berna Ç. Kocaman

1995

Linkage with International Markets (ADR-GDR) and Alternative Solutions to the Turkish Capital Market - Kudret Vurgun

1994

Portfolio Investments in International Capital Movements and Turkey ISE Research Department

1994

International Portfolio Investment Analysis and Pricing Model - Oral Erdoğan

1994

Taxation of Capital Market Instruments in Turkey Sibel Kumbasar Bayraktar

1994

ISE PUBLICATIONS RESEARCH ON DERIVATITES MARKET Some Basic Strategies on Securities Market Derived from Future Transactions and Options - Mustafa Kemal Yılmaz

1996

Derivatives Market - Theory and Practice - Prof. Dr. Ümit Erol

ISBN 975-8027-58-1

1999

Pricing of Future and Options Contracts Based on Capital Market Instruments - ISE Derivatives Market Department

ISBN 975-8027-62-X

1999

Interest Rate Futures - ISE Derivatives Market Department

ISBN 975-8027-61-1

1999

SECTORAL RESEARCH Automotive Sector - Sibel Kumbasar Bayraktar

1995

Textile Sector (Cotton) - Efser Uytun

1995

Food Sector - Ebru Tan

1995

Glass Sector - Özlem Özdemir

1995

Insurance Sector - Çağatay Ergenekon

1995

Tourism Sector - Oral Erdoğan

1995

Manifactural Paper and Paper Product Sector - Çağatay Ergenekon

ISBN 975-8027-09-3

1996

Textile Sector (Artificial-Synthetic, Woolen, Manufacturer Clothing Leather and Leather Goods) - Efser Uytun

ISBN 975-8027-10-7

1996

Food Sector (Vegetable Oil, Meat, Fruit, Dairy Products. Sugar, Flavor Products, Animal Feed) - Research Depertment

ISBN 975-8027-19-0

1997

Turkish Financial History from the Ottoman Empire to the Present*

975-7104-24-8

1999

Istanbul Stock Exchange in a Historical Perspective*

975-8027-00-X

1995

The Story of Ottoman Tiles and Ceramics*

975-7104-11-6

1997

CULTURE PUBLICATIONS

Turkey Timeless Culture* III. BOOKLETS Istanbul Stock Exchange For Investors Public Disclosure Platform The Emerging Companies Market Exchange Traded Funds Warrants ISE By Figures