US, European and Asian Investors in the Japanese Stock Market

Transcription

1 US, European and Asian Investors in the Japanese Stock Market Akiko Kamesaka* Abstract This paper investigates aggregate buying and selling by foreign investors, subdivided into US, European and Asian investors, in the Japanese stock market over the period 1981 to The results indicate that in the late 1990s US investors began to take more active positions than other foreign investors, and traded with good timing from the middle of the 1990s. US investors were also generally better than other investors when completing net purchases. While European and Asian investors also traded with good timing, other foreign investors generally did not net purchase or sell with good timing. JEL classification code: G15 Keywords: Investor Behavior, Foreign investors, Japan * School of Management, Aoyama Gakuin University, , Shibuya, Shibuya-ku, Tokyo , Japan. akiko@busi.aoyama.ac.jp Phone: URL: 1

2 1. Introduction The financial world is increasingly becoming complicated and integrated. From 1997, most Asian economies suddenly fell into a serious crisis, and the world has experienced several other financial crises in the most recent decade. We now know that the main cause of the Asian crisis was due to unhedged borrowing in foreign currencies. However, we still do not know very much about how foreign investors were speculating in currency, stock or other financial markets of the region. While Japan s financial market was already in crisis from the early 1990s, following the beginning of the Asian crisis several large Japanese financial institutions, including Yamaichi Securities Companies, Sanyo Securities Co. and the Hokkaido Takushoku Bank, began to experience financial difficulties. 1 It goes without saying that foreign investors play an important role in the Japanese stock market. Monthly surveys conducted by Quick Co., a Japanese research company, indicated that from June 1996 to October 1999 more than half of all professional investors in Japan continued to answer that foreign investors were the investor group that drew the most attention 2 (Wakasugi, Ohta and Asano 2001). The Fact Book of the Tokyo Stock Exchange reports that foreign investor purchases and sales in 2003 in the three largest Japanese markets (Tokyo, Osaka and Nagoya) amounted to 137 trillion yen, comprising more than 30 percent of the total purchases and sales of 435 million yen. The Fact Book also indicates that 89.3 percent of all trades in Japanese listed stocks were executed at the Tokyo Stock Exchange, 4.6 percent at the Osaka Stock Exchange and 0.2 percent at the Nagoya Stock Exchange. 3 At balance date 2003, foreign investors owned 17.7 percent of all listed stocks, based on the yen value of stock prices at the end of March This grew from only 6.4 percent in 1981 and 10 percent in 1995, with further increases in the flow and balance of foreign holdings following the end of the Asian crisis. The Tokyo Stock Exchange collects information on the yen value of client purchases and sales from securities firms each month. All client orders are categorized into trading by securities companies, banks, insurance companies, investment trusts, nonfinancial corporations, individuals, foreigners etc. Foreigner trading is further categorized into US, European, South-East Asian (hereafter, abbreviated as Asian) and other foreign investors. Kamesaka, Nofsinger and Kawakita (2003) found that foreign investors and securities companies trade with good market timing, while domestic individual investors and nonfinancial companies performed relatively poorly. 5 However, the trades of foreign investors across different regions have never been analyzed statistically. 6 In this paper, I focus on the behavior of four types of foreign investors in Japan: US, European, Asian and other foreigners. The sample period is from January 2

3 1981 to June It covers the bubble period of the latter half of 1980s, followed by the crash in the early 1990s, and the internet bubble period around I analyze foreign investor behavior before, during and after the crash by dividing the data into three subperiods: from January 1981 to December 1989 (before the crash or bubble period), from January 1990 to December 1998 (during the crash), and from January 1999 to June 2004 (after the crash). Figure 1 shows the value of the TOPIX index in local currency, and the Yen/Dollar (JPY/USD) exchange rate from 1972 to As shown, general stock prices rose sharply from the middle of 1980s and reached a historical maximum of in December 1989, falling sharply thereafter. JPY has gradually increased its value relative to USD. In this paper I employ the TOPIX to investigate and evaluate foreigners investment flows, rather than the Nikkei Index, the other major Japanese stock price index. This is because the TOPIX is based on all listed stocks in the First Section of the Tokyo Stock Exchange. As at year-end 2003, there were 1,533 listed companies on the Tokyo Stock Exchange s First Section, all of which are included in the TOPIX index, whereas the Nikkei index includes only 225 major stocks. Figure 1 about here II. Evidence Figure 2 graphs the annual net investment flow of US, European, Asian and other foreign investors in Japan. As shown, US investors began to take more active positions from the late 1990s; they more actively changed their position in annual terms, and sold massively in They continued to be net purchasers from On the other hand, European investors played the largest role until the middle of the 1990s. European pension funds aimed to make gains in the long term, and tried to realize profits before the crash was expected. From the annual net investment flow, we can see that the trade flows of all investor groups tend to move in the same direction: that is, when European investors are net purchasers, US, Asian and foreigners of other regions also tend to be net purchasers. Figure 2 about here 3

4 Table 1 lists the mean, standard deviation, minimum, maximum, median, first quarter and third quarter of monthly buying and selling of fund flows for each foreign investor group. The statistics are calculated by dividing the dataset into three subperiods: 1981 to 1989, when the Japanese market experienced continuous stock appreciation; 1990 to 1998, when the market experienced a sharp decline and the Asian crisis; and 1999 to 2004, when the market experienced the internet bubble and afterwards. The Net Purchases (NP) and Net Purchases Ratio (NPR) of foreign investor group i during month t is defined as follows: NP it = ( Purchasing Value it - Selling Value it ) (1) NPR it = ( Purchasing ( Purchasing Value Value it it - Selling Value + Selling Value it it ) ) (2) As indicated, both NP and NPR will be positive (negative) when the foreign investor group i buys more (less) equities than it sells, and will equal zero when purchases equal sales. NPR will equal one when all trades of foreign investor group i are buying, and will equal minus one when all trades of foreign investor group i are selling. Table 1 about here Panel A of Table 1 shows that in the 1980s more than half of all trades by foreign investors were conducted by European investors. The panel also shows that Asian investors took a larger role than US investors for the same period. The maximum and minimum values are at their most extreme for European investors. Panel B of the table shows that European investors continued to take the largest share of foreigners purchases and sales in the 1990s. However, US investors then began to overtake European investors, and took on the highest values for purchases, net purchases and the net purchases ratio. Purchases and Sales in Panels A to C show that the aggregate trades of US, European and Asian investors, respectively, increased from the bubble period to during the crash, irrespective of the fall in stock prices, and from during the crash period to after the crash period. Panel C indicates that US investors began to play the largest role of all four groups 4

5 after this period. Table 2 presents the correlation coefficients between the flows of funds of each investor group across the three different subperiods. The results show that US, European and Asian investors tended to increase or decrease purchases, sales and net purchases at the same time, before, during and after the crash. However, while their purchases and net purchases increased when stock prices were increasing, sales do not appear to have any relation with the stock return. Therefore, there was an asymmetry between their purchases and sales and the stock return, and the asymmetry continued throughout the period under investigation. The investment flows, whether they are in gross or in net terms, also do not appear to display any relation with the JPY/USD exchange rate. Table 2 about here Table 3 examines the causality relationships between the net purchases of each foreign investor group and the index return using VAR analysis. While the Akaike information criterion and Schwarz Bayes information criterion suggest relatively long lags should be specified for the VAR, since most of the coefficients become insignificant after just a few months, the results of three-month lag is specified. Panel A of the table indicates that before the crash period, European, Asian and other foreign investors, respectively, tended to be the net purchasers after their own net purchases: that is, after the month they net purchase, we expect net purchasing in the following month. The same panel also shows that US, European and Asian investors tended to be net purchasers a few months after a price fall. Stock returns are shown to have increased following European and Asian investors net purchases during 1980s. Panel B shows the results of the long crash period in the 1990s. During this subperiod, US investors were net purchasers after their own net purchases of the previous month. European and Asian investors were also net purchasers after their own net purchases. US, European and Asian investors tended to be net purchasers after a few months stock price decline, after observing the symptoms of price recovery. Panel C shows that US investors continued to be net purchasers after their own net purchases of the previous month from the late 1990s, during and after the internet bubbles. However, the relation between the net purchases and that of the previous month disappeared in this period for each of the other foreign investor groups, respectively. Asian investors tend to buy after the US and European investors net purchases during this period. Considering 5

6 the increase in the size of the US investor trade after the middle of the 1990s (Table 1), US investors then began to exert the largest influence on the Japanese stock market. 7 Table 3 about here In order to evaluate which foreign investor groups traded with good timing, I estimate the aggregate one-month following return after foreigners net purchases or net sales based on each foreign investor group s net investment flow (Grinblatt and Titman (1993) and Kamesaka, Nofsinger and Kawakita (2003)): 8 AggregateFollowingOne Month Return s t = = (Purchasing Valueis Selling Valueis ) I Rs+ - 1 whereas, I denote an indicator function that takes either one or zero under the conditions defined later in this section. it (3) Figure 3 about here Figure 3 presents the performance measure of each of the four groups of foreign investors after their net purchase, by imposing a condition to equation (3) that I takes a value of one when (Purchasing Value is Selling Value is ) > 0, and zero, otherwise. Earlier studies do not distinguish between positive returns after net purchase (net purchase before stock price increase) from negative returns after net sales (net sales before stock price falls). However, here they are separated by the inclusion of the indicator function. In this figure, foreign investor i s trade is evaluated by the following day s return only when positive net purchases by investor i are observed. Equation (3) adds this monthly evaluation measure up to time t, such that Figure 3 ensures the stability of each foreigner s trade performance throughout the period. The figure clearly shows that US investors were performing best in the Japanese stock market. Their superior performance was achieved mainly in 1999, during the period of the internet bubble. While US investors suffered large losses after the Asian crisis, they made much larger gains during the internet bubble. The performance of European and Asian investors show similar patterns to US investors. That is, after losses following the Asian crisis, recovery 6

7 was achieved. The differences in performance among US, European and Asian investors are most apparent in 1999, with the apparent superiority of US investors mainly brought about by active purchasing. Figure 4 about here Figure 4 captures each foreign investor group s performance after their net sales, by imposing a condition to equation (3), that I takes a value of one when (Purchasing Value is Selling Value is ) < 0, and zero, otherwise. In Figure 4, foreign investor i s trade is evaluated by the following month s return only when positive net sales by investor i are observed. When we observe net sales by foreign investor i, (Purchasing Value is Selling Value is ) will be negative, and if we observe negative stock return after they sell, it means that they have net sold before a stock price fall, which can be used to suggest that they have successfully sold before the stock price fall. If this is indeed the case, we may expect to observe positive performance values as defined in Equation (3) for Figure 4. The results indicate that the performance of US, European and Asian investors is at the same level when the data period from January 1981 to June 2004 is aggregated. One difference is that European and Asian investors sold with good timing, especially in the early 1990s, whereas US investors sold with good timing only after 1997, and after the Asian crisis had begun. Figure 4 shows that European investors were selling with bad timing from the middle of the 1980s to the late 1980s, however these sales were mainly from stocks invested in before the beginning of the data period; that is, they started to invest in the 1970s, and sold after capital gains in the stock price bubble of the mid 1980s. The dataset employed in this manuscript is only available after 1980 and there appears to be no possible means of evaluating European investors net sales in the 1980s. It is, however, likely that they sold at a higher price than they had previously purchased, and gained from their long-term investments in Japan. Summing up Figures 3 and 4, US, European and Asian investors were buying and selling with good timing. US investors performed relatively better, especially after the internet bubbles in the late 1990s, and they were also particularly better when net purchasing. 9 European and Asian investors also performed well; however, other foreign investors did not make profit or losses from investing in Japan. Previous studies have also shown that foreign investors are the market winners in Japan, while in this study it has 7

8 been shown that European and Asian investors were performing relatively better until the early 1990s, while US investors began to perform better after the middle of the 1990s. III. Conclusions This paper investigates over twenty years aggregate trade performance of US, European and Asian investors in the Japanese stock market. While the Tokyo Stock Exchange collects the yen value of aggregate purchases and sales each month, the trades of foreign investors of different regions have never been analyzed statistically in any previous work. The sample period employed is from January 1981 to June 2004, and covers the bubble period of the second half of the 1980s, followed by the crash in the early 1990s. The study examines the behavior of foreign investors in different regions, before, during and after the crash by dividing the data into three subperiods: from January 1981 to December 1989 (before the crash or bubble period), from January 1990 to December 1998 (during the crash), and from January 1999 to December 2004 (after the crash). The results before, during and after the crash indicate that all foreign investors increased their purchases, sales and net purchases (purchases less sales) at the same time. They increased their purchases and net purchases while stock prices were increasing, however, no relation was found between sales and stock returns. Therefore, an asymmetry exists between purchases, sales and stock price returns, and this asymmetry continues to exist regardless of the changes in the Japanese stock market. Foreign investors of each region tended to increase their net purchases after each investor s own net purchases of the previous month and after a few months stock price fall before and during the crash period. However, these tended to disappear after the crash period. The stock prices tended to increase after European and Asian investors net purchases during the 1980s, however the relation disappears in the 1990s and thereafter. By evaluating the one-month following return, I also observe positive stock returns following each foreign investor s net purchases, and negative stock returns after net sales. US investors performed extremely well by net purchasing during the internet bubble around The overall performance of net sales did not differ among US, European and Asian investors. However, I should also note that the European investors sold after the stock price appreciation from the middle of the 1980s, and it is likely that they gained from investing in Japan by selling stocks they bought in the 1970s. 8

9 Acknowledgements The author would like to thank Mikiyo Niizeki, Jun Shirota, Yoshiro Tsutsui and other participants at the Japanese Economic Association and the Japan Society of Monetary Economics meetings for helpful comments on earlier versions of this paper. Financial support from the Japan Society for the Promotion of Science, Grants-in-Aid for Young Scientists (B ) and a competitive grant from the Ryukoku University Research Institute for Social Sciences is also gratefully acknowledged. 9

10 References Choe, H., Kho, B., and Stulz, R. (1999) Do foreign investors destabilize stock markets? The Korean experience in 1997, Journal of Financial Economics, Choe, H., Kho, B., and Stulz, R. (2000) Do domestic investors have more valuable information about individual stocks than foreign investors? Ohio State University Working Paper. Grinblatt, M., and Titman, S. (1993) Performance measurement without benchmarks: An examination of mutual fund returns. Journal of Business, Hamao, Y., and Mei, J. (2001) Living with the enemy : An analysis of foreign investment in the Japanese equity market. Journal of International Money and Finance, Ito, T. (2000) The stagnant Japanese economy in the 1990s: The need for financial supervision to restore sustained growth, in Hoshi, T. and Patrick, H. eds, Crisis and Change in the Japanese Financial System, Kluwer Academic Publishers, Kamesaka, A. (Murase, A.) (1999) Stock investment performance of main investor groups in Japanese market, in Hiroji Baba, ed., Myth and Reality of Japanese Firms, Daito Bunka University Institute of Business Research Book Series No. 16. Reprinted in Review of Monetary and Financial Studies, (2001) (in Japanese). Kamesaka, A. (2003) Investor Groups in Japan, in Hayashi, T. Matsuura, K., and Yonezawa, Y. eds Financial Problems in Japan, Nippon Hyouron Sha (in Japanese). Kamesaka, A., Nofsinger, J., and Kawakita, H. (2003) Investment patterns and performance of investor groups in Japan, Pacific Basin Finance Journal, Kamesaka, A., and Wang, J. (2001) Foreign and domestic investors in Indonesia, Ryukoku University and University of New South Wales Working Paper, presented at the Ministry of Finance, Indonesia. Kamesaka, A., and Wang, J. (2005) The Asian Crisis and Investor Behavior in Thailand s Equity Market, presented at the American Economic Association meetings 2005 in Philadelphia. Kim, W., (2000) Do foreign investors perform better than locals? Information asymmetry 10

11 versus investor sophistication, KDI School of Public Policy and Management Working Paper. Kusano, T. (2001) QSS stock research and behavior of foreign investors in Wakasugi, T., Ohta, H., and Asano, Y. eds., Formation of Investors Expectation and Market Price Movement (Tousika no Yosoukeisei to Souba doukou) Nikkei BP Planning Inc. (in Japanese). Shirota, J. (2002) Foreign Investors in the Japanese Stock Market, Toyo Keizai Inc. (in Japanese). Wakasugi, T., Ohta, H., and Asano, Y. eds (2001) A supplement to Formation of Investors Expectation and Market Price Movement (Tousika no Yosoukeisei to Souba doukou) Nikkei BP Planning Inc. (in Japanese). 11

12 Table 1. Summary statistics of the fund flows of different foreign investment types Mean Std Dev Minimum Maximum Median 1st Qrt 3rd Qrt Panel A Purchases in Billion Yen US Europe Asia Others Sales in Billion Yen US Europe Asia Others Net Purchases in Billion Yen US Europe Asia Others Net Purchases Ratio US Europe Asia Others Return Yen/Dollar Panel B Purchases in Billion Yen US Europe Asia Others Sales in Billion Yen US Europe Asia Others Net Purchases in Billion Yen US Europe Asia Others Net Purchases Ratio US Europe Asia Others Return Yen/Dollar Panel C Purchases in Billion Yen US Europe Asia Others Sales in Billion Yen US Europe Asia Others Net Purchases in Billion Yen US Europe Asia Others Net Purchases Ratio US Europe Asia Others Return Yen/Dollar

13 Table 2. Correlation Matrix Panel A US Europe Asia Others Return Yen/Dollar Purchases US Europe Asia Others Return Yen/Dollar Sales US Europe Asia Others Return Yen/Dollar Net Purchases US Europe Asia Others Return Yen/Dollar Net Purchases Ratio US Europe Asia Others Return Yen/Dollar

14 Panel B US Europe Asia Others Return Yen/Dollar Purchases US Europe Asia Others Return Yen/Dollar Sales US Europe Asia Others Return Yen/Dollar Net Purchases US Europe Asia Others Return Yen/Dollar Net Purchases Ratio US Europe Asia Others Return Yen/Dollar

15 Panel C US Europe Asia Others Return Yen/Dollar Purchases US Europe Asia Others Return Yen/Dollar Sales US Europe Asia Others Return Yen/Dollar Net Purchases US Europe Asia Others Return Yen/Dollar Net Purchases Ratio US Europe Asia Others Return Yen/Dollar

16 Table 3. VAR statistics Panel A US Europe Asia Others Return Yen/Dollar US * * ** * Europe * * ** ** * Asia ** ** ** ** ** Others * * * Return * * ** ** * Yen/Dollar Constant * *** Adjusted R sq F-test * *** ** ** ** Significance of the coefficients at 10%, 5%, and 1% is denoted by *, ** and ***, respectively. 16

17 Panel B US Europe Asia Others Return Yen/Dollar US ** Europe ** ** Asia *** * Others *** *** Return * * Yen/Dollar ** Constant ** Adjusted R sq F-test Significance of the coefficients at 10%, 5%, and 1% is denoted by *, ** and ***, respectively. 17

18 Panel C US Europe Asia Others Return Yen/Dollar US *** * *** * *** *** Europe *** * *** * Asia ** *** ** * *** Others *** Return * Yen/Dollar * Constant * * Adjusted R sq F-test ** *** *** Significance of the coefficients at 10%, 5%, and 1% is denoted by *, ** and ***, respectively. 18

19 Figure 1. Stock Index and Yen/Dollar Exchange Rate TOPIX Yen/Dollar TOPIX Yen/Dollar Year Note: Closing Price of the TOPIX and Yen/Dollar exchange rate shown from fiscal years 1972 to

20 Figure 2. Net Investment Flow of US, European, Asian and Other Foreign Investors bil Yen US Europe Asia Others Year Note: Net investment flow of US, European, Asian and other foreign regions are shown in billion yen from 1972 to

21 Figure.3 Following One Month Stock Returns After Foreign Net Purchase 200 Billion Yen US Europe Asia Others Year Note: Following one month stock returns after net purchase of US, European, Asian and other foreign investors are shown using monthly dataset for the period 1981 to

22 Figure 4. Following One Month Stock Returns After Foreing Net Selling Billion Yen US Europe Asia Others Year Notes: Following one month stock returns after net selling of US, European, Asian and other foreign investors are shown using monthly dataset for the period 1981 to

23 1 Ito (2000) details the failures of Japanese financial institutions before and after the Asian crisis. 2 The Quick Co. sends questionnaires to over three hundred professional stock investors each month, with the response rate exceeding fifty percent since the commencement of the surveys. 3 There was also 5.9 percent of off-the-board trade in the year The Tokyo Stock Exchange reports investor groups stock holding based on closing prices in March, since the accounting year ends in March for most Japanese firms. 5 Kamesaka (Murase) (1999, 2003) also find foreigners trading with good market timing in the Japanese market. Hamao and Mei (2001) also analyze the Japanese market. Choe Kho and Stulz (1999, 2000) find domestic investors have better information than foreigners. Kim (2000) finds that foreigners out-perform domestic individuals in large stocks, however, under-perform in small stocks. Kamesaka and Wang (2001) find that foreign purchases from domestic investors outperform domestic investors purchases from foreign investors in Indonesia before the Asian crisis. Kamesaka and Wang (2005) analyze Thailand s stock market and find that foreigners consistently traded with good timing before, during and after the Asian crisis, regardless of the large changes in market conditions. 6 Shirota (2002) documents related statistics and a calendar of events regarding foreign investors in Japan. 7 In spite of these results, Japanese investors tend to gain an understanding of the total purchases and sales of the four groups of foreign investors by understanding the positions of European investors (Kusano (2001)). They therefore tend to underestimate the effect of trading by US investors. 8 Kamesaka (Murase) (1999, 2003) also use this measure to evaluate investment performance of main investor groups in Japan using weekly and monthly data, respectively. 9 US investors superiority in overall market timing remains, even when their trade performance is evaluated using a three-month following returns or US dollar-based stock returns. 23