LONG RUN PRICE PERFORMANCE OF IPO STOCKS IN BANGLADESH. M. Sadiqul Islam 1 Mahfuja Malik Mohammad Riaz Uddin

Transcription

1 Journal of Finance and Banking Volume 9, Number 2 December 2011 LONG RUN PRICE PERFORMANCE OF IPO STOCKS IN BANGLADESH M. Sadiqul Islam 1 Mahfuja Malik Mohammad Riaz Uddin Abstract: This study was conducted to analyze the short and long run price performance of IPOs in Bangladesh. Based on a sample of 163 IPOs that were issued during the period between 1992 and 2006, this study documents short run out-performance and long run underperformance of IPOs in the secondary market. The short run out-performance of IPOs peaks in the first month of trading in the secondary market. In the long run IPOs under perform the industry and the market in general. But the underperformance to industry is much severe than underperformance to market. Further analyses reveal that the long run underperformance is much higher for IPOs issued in few years, particularly the hot issue periods. IPOs belonging to Leather, Engineering, Paper, Ceramic and Food industries underperform their respective industry and the market severely in the long run. On the other hand, IPOs belonging to Banking and Non-bank Financial Institutions industries outperformed the industry and the market significantly. Keywords: IPO, Stock Return, Primary Market, Price Performance. INTRODUCTION The behavior of IPO stock price had been widely researched in many countries over the last two decades. The issue that puzzles the financial economists is why these anomalies in IPO price behavior occur in markets that are considered to be competitive. Researchers have documented three major anomalies in the IPO price behavior. First, IPOs provide abnormal initial return. That means investors buying IPOs from the primary market can sell at a significantly higher price in the initial secondary market. Second, IPO stocks in general outperformed the market and industry in the short run aftermarket. Third, IPO stocks underperform the market and their industry counterpart in the long run. It is perplexing in deed why these inefficiencies continue to exist in markets which are considered to be relatively efficient. Many financial economists view these as initial overreaction followed by a 1 The authors are respectively Professor and Lecturers, Department of Finance, University of Dhaka Electronic copy available at:

2 2 Islam, Malik and Uddin correction in the long run aftermarket. But there is no logical explanation why investors would consistently make abnormal initial return or befooled in the long run aftermarket. In Bangladesh as well IPOs produced initial abnormal return (Islam, 2001). However there is hardly any study on the long run price performance of IPO stocks in Bangladesh. The anomalies in the price performance have policy implications for the investors, companies, regulators, policy markers and researchers. In this backdrop, in this study an attempt has been made to analyze the long run price performance of IPO stocks in Bangladesh. LEGAL AND INSTITUTIONAL FRAMEWORK FOR IPOs The legal and institutional framework for the securities market in Bangladesh had been reshaped in the 1990s. The reform in the securities market started with the enactment of the Securities and Commission Act and subsequent formation of the Securities and Exchange Commission (SEC) in With this enactment, the SEC had been empowered to frames rules and regulations, and to take actions against any party for non-compliance with the rules. Within next few years, a number of rules and regulations were framed to develop the market and to protect the interest of investors. The major rules relate to merchant bankers and portfolio managers, insider trading, mutual funds, IPOs, right offering, stock broker and dealers, credit rating and substantial acquisition. Many of these rules were amended several times to match the needs of the market. In Bangladesh, firms intending to make public offering of shares have to prepare and submit draft prospectus to the SEC. The prospectus must include separate due diligence certificates from issue manager and underwriter. All public issues must be underwritten on firm commitment basis. The Securities and Exchange Commission (Public Issue) Rules, 2006 define the contents of a prospectus. The prospectus must include the project, use of issue proceeds, financial statements, risk factors of the firm, directors relationships and related party transactions, executive compensation and executive options. There is restriction on the use of issue proceeds. The issuing firms cannot use the issue proceeds for any purpose other than those mentioned in the prospectus. The shares held by the prepublic owners are subject to a lock in period of three years. This means that sponsors and directors cannot take advantage of a hot issue market for their personal benefit. Also, all shares purchased from the issuing firm through private placement during the immediate preceding two years shall be subject to a lock-in period of one year. Formally there is no restriction on the issue price, but the issuers have to provide a justification for the issue price by the price-earning multiple or price to book-value ratio of similar firms. In practice, issuers cannot take advantage of market bubbles to charge a high price of issued shares. The final offer price of IPO must be announced in the prospectus. In 2010, public issue of securities through book building method was introduced. Under this method, an issuing firm can discover the issue price through book building with participation of institutional investors. This creates Electronic copy available at:

3 Journal of Finace and Banking 3 greater possibility to maximize the issue proceeds of the company and reduce their cost of capital. Of the IPO issue, there is 10 percent quota for the non-resident Bangladeshi investors and there is another 10 percent quota for mutual funds. In Bangladesh, there are presently 31 merchant banks, but their activities are mostly limited to primary market making activities. Many of them have yet to develop professionals to manage portfolios. The market for issue management and underwriting is extremely competitive primarily because of limited number of IPO issues. There is full separation between merchant banking and stock broking and dealing. Lately, merchant banking had been separated from commercial banking. But commercial banks can conduct merchant banking through their subsidiaries. In Bangladesh there is absence of retail outlets for distribution of shares. Subscription of all public issues is usually made through the designated branches of commercial banks. There are two stock exchanges in Bangladesh Dhaka Stock Exchange and Chittagong Stock Exchange. The Dhaka Stock Exchange is the largest stock exchange of the country. Most of the stocks listed with the Chittagong Stock Exchange are also co-listed with the Dhaka Stock Exchange. Only recently the process of demutualization of both exchanges has started. The Central Depository Bangladesh Limited (CDBL) facilitates paperless trading of securities. There had been only two credit rating agencies, but lately six more credit rating agencies started their operations or are in the process of starting their operations. LITERATURE REVIEW The anomalies of IPO returns have been widely researched. These studies focused on the abnormal initial return of IPO stocks, price performance in the aftermarket and hot issue markets. Financial economists developed different models to explain the abnormal initial return of IPOs. Some of them consider information asymmetry as the major reason for abnormal initial return of IPOs. Baron (1982) attributes the IPO initial return to information asymmetry between the issuer and the investment banker. But Muscarella and Vetsupens (1989) contradict this hypothesis and document that self marketed IPOs of securities firms produced similar initial return. Also Bary, Muscarella, Peavy and Vetsuypens (1990) provide similar findings. Rock (1986) develops the winner s curse hypothesis and attributes the abnormal initial return to information asymmetry between informed and uninformed investors. Levis (1990) provides evidences to support the winner s curse problem in the UK market. Koh and Walter (1989) and Keloharju (1993) also find evidences to support Rock s model. However, Ruud (1993) and Welch (1989) contradict the validity of Rock s model. Beatty and Ritter (1986) extended Rock s model and argue that uninformed investors will demand more money left on the table because of uncertainty of the initial price of IPO share in the secondary market. Benveniste and Spindt (1989) argue that underpricing arises as a cost of compensating investors with positive information.

4 4 Islam, Malik and Uddin Allen and Faulhaber (1989) argue that only good firms can signal their type by underpricing of IPOs and firms having no information asymmetry do not underprice. Welch (1989) also develops a signaling model where high quality firms signal their superior information to investors by underpricing their IPOs. Jagadeesh, Weinstein and Welch (1993) provide evidence to support the signaling hypothesis. Booth and Chua (1996) have attributed the IPO underpricing to the issuer s demand for ownership dispersion. Brennan and Franks (1995) also have a similar hypothesis. Welch (1992) developed the cascade theory of IPO underpricing which assumes that later potential investors imitate the actions of earlier investors, which leads to a cascade that may succeed or fail rapidly. Issuers underprice their IPOs to avoid issue failure. Some financial economists attempted to explain the hot issue markets of IPOs. These hot issues provide more than the average rate of return in the aftermarket. Economists document that there are periods when the average performance of new issues is abnormally high. These periods are known as hot issue periods (Ibbotson and Jaffe, 1975). Hot issue market has been a periodic phenomenon almost in every market throughout the world. Ritter (1984) develops the changing risk composition hypothesis as a plausible explanation for the hot issue market. Riskier firms are more underpriced than less risky firms and it is very likely that periods of risky issues will experience higher average initial return. Allen and Faulhaber (1989) argue that exogenous shocks in some industries substantially improve the profitability and this may create a hot issue market. According to their hypothesis, hot issue market continues until the number of firms in the industry adjusts to new conditions. There is a wide body of literature on the aftermarket price performance of IPOs in different markets. These studies document that IPO stocks, in general, underperform the market or industry in the long run aftermarket. Many studies provide evidences to indicate that higher underpricing in the initial market is, in some way or other, related to the underperformance in the aftermarket. Krigman, Shaw and Womack (1999) provide evidence that first day winner IPOs continue to be winners in the short run aftermarket and first day dogs continue to be relative dogs. Only exceptions are extra-hot IPOs which provide the worst aftermarket performance. Mcdonald and Fisher (1972) found significant negative return during the first year after the issue. Ibbotson (1975) finds that there is generally positive performance during the first year, negative performance during the next three years and generally positive performance during the last (fifth) year. Ritter (1991) documents underperformance of US IPOs relative to their matching firms. Also, IPO stocks with high adjusted initial returns tend to have the worst aftermarket performance. In addition, smaller issues and young firms had the worst aftermarket performance.

5 Journal of Finace and Banking 5 Loughran and Ritter (1995) document that investors had to invest 44 percent more money to have the same wealth five years after the issue. Firms issuing equity during years of little issuing activity do not underperform much at all. They view that the market appears to overweight the recent improvement and underweight long-term, mean-reverting tendencies in the operating performance measures. Brav and Gompers (1997) find that the venture backed firms outperform the non-venture backed IPOs over a period of five years. The underperformance of non-venture backed IPOs is primarily driven by small issues. Teoh, Welch and Wong (1998) examine the relation between the long run post-ipo return underperformance and the IPO firms earnings management. Using discretionary current asset accruals as the proxy for earnings management, they document that issuers with higher discretionary accruals have poorer stock return performance in the subsequent three years. Chaney and Lewis (1998) provide evidence of a positive association between a proxy for income smoothing and the IPO performance. They document that IPO firms performing well tend to report earnings with less variability relative to operating cash flow, while firms that perform poorly tend to report earnings with greater variability relative to operating cash flow. Carter, Dark and Singh (1998) find that IPO stocks underperformed the market during next three years. They document that the long run underperformance is less severe for IPO stocks having more prestigious underwriter. Michaely and Shaw (1994) also find that IPOs managed by high prestige investment bankers tend to have smaller initial return and less negative long run returns than do IPOs handled by lower reputation underwriters. Fields (1995) finds that in the long run, IPOs having larger institutional shareholding significantly outperform those with smaller institutional shareholdings. Dong, Michel and Pandes (2011) also examined the relationship between underwriter quality and the long run performance of IPOs. They find that IPOs with underwriters with better quality and reputation perform better than others. Rajan and Servaes (1995) develop a model to explain the three anomalies associated with the IPO market underpricing of IPOs, windows of opportunities for new issues and long term underperformance of IPOs. Their empirical results indicate that common factors appear to partially explain all three phenomena and with these, they suggest a disciplined way of investigating potential irregularities in the IPO market. Similar to the US market, many other markets experienced long run underperformance of IPO stocks. Levis (1993) reveals that British IPOs underperformed significantly in the long run. He documents that small issues, which had high initial returns, experienced the worst aftermarket performance in the long run. He provides evidences that are consistent with the proposition that a certain portion of first day returns is the result of deliberate underpricing, but much of the underperformance is due to market overreaction. This is consistent with the findings of Ritter (1991) that investors are periodically overoptimistic over the earnings

6 6 Islam, Malik and Uddin potential of at least some of the newly listed companies. Finn and Higham (1988) find that the average aftermarket performance of Australian IPOs was negative during the first year, but not statistically significant. Lee, Taylor and Walter (1996) provide evidence that the underperformance of Australian IPOs is not confined to the first year aftermarket, rather continues in the ensuing three years. The IPOs lost half of their values on an average within a three-year period. The underperformance of IPOs in the first and second year is related to the degree of initial return of IPOs, but Lee, Taylor and Walter (1996) argue that long run returns are not associated with underpricing in the manner that the overreaction or fad explanation suggests. Li and Hovey (2009) analyzed the relationship between IPO performance and corporate governance in China. They document that ownership structure has an impact on the long run performance of IPOs. In China, foreign and legal person ownership has a positive influence on the long performance of IPOs. Ahmed-Zaluki, Campbell and Goodacre (2009) examined the relationship between earnings management and IPO performance in Malaysian market. They document that IPO issuers who are aggressively engaged in earnings management have worst aftermarket performance. Page and Reyneke (1997) document underperformance of South African IPOs relative to sized-matched portfolio, price-earning ratio matched portfolio and sector matched portfolio at the end of fourth year after the issue. Their analyses indicate that the small issues underperformed more than the larger issues. McGuinness (1993) find positive return of IPOs in Hong Kong during first few months, which was reversed leading to a long-term decline in returns. Aggarwal, Leal and Hernandez (1993) analyzed the long run price-performance of IPOs in Brazil, Chile and Mexico. In the Brazilian market, IPOs did not underperform in the first year, but significantly underperform in the second and third year. In the Chilean market, IPOs outperformed the market during the first two years, but underperformed in the third year, although neither the out-performance nor the underperformance was statistically significant. Unlike the other two markets, Mexican IPOs underperformed in the first year. They also document a general negative relationship between the IPO initial return and the first and third year returns in the Brazilian market. Affleck-Graves, Hegde and Miller (1996) document that IPOs providing abnormal initial return also provide abnormal return relative to matched firms of same size in the short run aftermarket. They also provide evidence that the observed conditional price trend can not be satisfactorily explained by underwriter price stabilization. The underperformance of IPOs is consistent with the fact that the operating performance of IPO firms decline substantially subsequent to the public offering. Jain and Kini (1994) document that the operating profit on assets of IPO firms decline significantly compared to their pre-ipo levels. Their analysis also indicates

7 Journal of Finace and Banking 7 that there is a positive relation between the managerial ownership retention and the post issue operative performance. Jain and Kini (1994) suggest that investors have very high expectations of future earnings growth, which are not subsequently realized. Mikkelson, Partch and Shah (1997) find that the operating performance of IPO firms exceeds the performance of matched publicly traded firms before going public, but after going public, their operating performance declines to a level below the performance of matched firms. The underperformance of IPOs in the long run aftermarket indicates that investors are overoptimistic about the long run performance of IPO firms. Rajan and Servaes (1997) document that investment analysts are overoptimistic about the earning potential and the long term growth prospects of IPO firms. They find that firms with higher projected growth underperform the market and their matched firms, while firms with lowest projected growth outperform the market and their matched firms. The significant initial return of IPOs, short run out-performance and long run underperformance contradict the efficient market paradigm. A high initial return followed by under-performance apparently indicates the existence of overvaluation, over-reaction or speculative factor in the secondary market. This is consistent with Shiller s (1990) view that investors psychology is important in the decision to buy IPOs. He documents that a vast majority of IPO investors did not make their own estimate of the value of stock. Also, investors are encouraged by the positive price change and discouraged by the negative price change. The long run underperformance of IPO stocks documented in different markets questions the efficiency of securities markets in both developed and emerging markets. Do the IPOs issued in Bangladesh behave the same way? There is dearth of study on the performance of IPOs in Bangladesh. In this backdrop, this study attempts to analyze the short and long run price performance of IPOs in Bangladesh. RESEARCH METHODS In this study, two performance measures have been taken for analysis the cumulative wealth relative to the DSE general index (CWRX) and the cumulative wealth relative to industry (CWRI). A two-month period has been considered as the short run period. The long run price performance measures have been prepared for a maximum period of five years. The cumulative wealth relative (CWRI jm ) to the industry for the period between the first trading of IPO and the end of m-th month has been estimated as follows:

8 8 Islam, Malik and Uddin 1 HPR jm CWRI jm 1 ARI km where, HPR jm is the holding period return of j-th IPO for the period until m-th month and ARI km is the average return on k-th industry matching firms for the period until the end of m-th month in the aftermarket. The cumulative wealth relative to the DSE general index CWRX jm for the period until m-th month has been estimated as: 1 HPR jm CWRX jm 1 ARM m where, ARM m stands for the change in the DSE general index during the corresponding period until the end of m-th month in the aftermarket of j-th stock. For estimating holding period returns of IPO stocks and matching firms, a return index for each stock has been prepared after adjusting cash dividends, stock dividends, right issues and splits. All the adjustments were made with reinvestment assumption. From the return index of each stock, cumulative returns from the first day of trading were estimated. These returns were taken to prepare industrial stock return indexes on equally weighted basis. For preparing the industrial return indexes, all stocks belonging to the industry at least three years before the listing of IPO stock have been included. The IPO short run and long run price performance has been analyzed by categorizing the IPO stocks by size of issue, year of issue and industrial sector. First, the mean and median wealth relatives to industry and to market were estimated for all IPOs. Second, IPOs were categorized by the size quartiles and the mean and median wealth relatives to industry and to market were estimated for each size quartile. Third, IPOs were categorized by the year of issue and the same wealth relatives were estimated for each year of issue. Finally, IPOs were categorized by industry and the mean and median wealth relatives of IPOs belonging to each industry were estimated. The sample includes 163 IPOs that were issued in Bangladesh during the period between 1992 and A few IPOs issued during this period were not included in the sample due to non-availability of data. The sample IPOs represent almost all industries including Food, Engineering, Pharmaceuticals, Ceramic, Cement, Textile, Leather, Bank, Non-Bank Financial Institutions, Insurance, Mutual Funds, Paper and Miscellaneous sectors.

9 Journal of Finace and Banking 9 The aftermarket price performance of IPOs has been followed up for a maximum period of five years. Firm specific data requirements include the size of equity issue, the number of equity shares, the issue price, cash dividends, stock dividends during the period between the year of public issue and five years after issue. In addition, the additional equity raised by rights and conversions, IPO stock prices and stock price index subsequent to the issue were collected for a period of eight years. During the period under study, few IPOs got delisted from the exchange due to bad performance and there was possibility of survival bias. A survival bias is likely to cause underestimation of the average return of IPOs in the aftermarket. To minimize the effect of survival bias, the last wealth relatives of delisted IPOs were continued till the end of study period. The data relating to the IPOs in the Bangladesh market have been collected from the Monthly Reviews and Annual Reports and other publications of the Dhaka Stock Exchange (DSE). The first day prices of IPO stocks have been taken from the electronic data provided by the DSE and from the records of the DSE and from the website of the DSE. Firm-specific data of IPO stocks have been taken from their prospectuses, publications of DSE and the website of the Securities and Exchange Commission of Bangladesh. ANALYSIS AND FINDINGS This study is based on the IPOs that were issued in Bangladesh during the period between 1992 and During this period, an average of about 11 IPOs entered the capital market every year. Among them 67 IPOs were issued during the three year period between 1994 and The highest number of 24 IPOs entered the market in In Bangladesh, there is no consistent pattern of growth in the number of IPO issues over the years. However, as observed in other countries, IPO issues had been relatively higher surrounding the period of bubbles and hot issues. Table 1 Descriptive Statistics Variable Mean Std. Dev. Size of IPO Issue (Million Taka) Initial Return (Percent) Sponsors Capital (Million Taka)

10 10 Islam, Malik and Uddin In general the average size of IPO issue has an increasing trend over time. Whereas the average size of the issue had been Taka million during the period under study, the average size had been more than Taka 123 million during the period between 2003 and The sponsors capital at the time of IPO had been Taka million with a standard deviation of Taka million. The sample IPOs provided an average initial (first trading day) return of percent on the issue price. The highest initial return had been observed in 1996 with an average of percent. This was followed by percent in 2005 and percent in The variation of initial return in different years is statistically significant. Among different industries as well, the initial return has wide variation. The initial return had been highest in the Engineering industry with an average of percent. This is followed by Leather industry with an average of per cent. The variation of initial return in different industries is statistically significant. Further analysis reveals that IPO initial return had been consistently relatively higher for the smaller issues and relatively lower for the larger issues. The average initial return of smallest, smaller, larger and largest quartiles of IPOs had been percent, percent, percent and percent respectively. The variation of initial return among IPOs in different size groups is statistically significant. Table 2 Short and Long Run Price Performance of IPO Stocks in Bangladesh Cohort Month Wealth Relative to Industry Wealth Relative to Market Mean Median Mean Median The relative performance of IPOs stocks has been analyzed by estimating the wealth relative to industry and the wealth relative to market. Analysis of post-issue price

11 Journal of Finace and Banking 11 performance indicates that IPO stocks continue to provide positive return in the short run. In Bangladesh, the short run price performance reached its peak in the first month of trading in the secondary market. This is a bit shorter than the IPO price rally in other markets. In many markets IPO price performance reaches its peak in the second month of secondary market trading. The mean wealth relative to industry reached its peak in the first month with a mean of and median of In the same period, the median wealth relative to market reached its peak. The mean wealth relative to the market however continued to rise in the long run. The relative performance to industry in the long run has been shown in Figure 1. The median wealth relative to industry continued to slide down till the fourth year. The median wealth relative had been less than unity throughout the period under study. This indicates that majority of IPO stocks underperformed the comparable firms within the same industry. The median wealth relative reached its worst in month 44 with a value of This means that the IPO investors lost 38.4 percent compared to their industry comparables. There had been slight recovery of wealth relative in the fifth year. But within few months, the recovery lost its momentum and started to slide down again during the last few months of fifth year. The mean wealth relative to industry also showed a similar pattern of performance. The relative performance over the market had not as sliding as it had been over the industry. Figure 2 shows the mean and median wealth relative of IPO stocks to market. The median wealth relative to market continued to be less than unity till the fourth year. It reached its bottom in month 49 when median wealth relative had been The median value hovered around 0.90 from the second year till the fourth year. This means that the IPO investors had been 10 percent worse than investors who invested in other securities. A temporary recovery of median wealth relative for few months in the fifth year lost its momentum and started to fall again. The mean wealth relative to market however remained greater than unity throughout the period under study. The mean value hovered around 1.10 from second year till the fourth year. A higher mean coupled with low median indicates that few IPOs had outstanding performance compared to the market although the majority IPOs underperformed the market. This provides indications why investors possibly buy IPO stocks that underperform in the aftermarket. It is likely that investors are overwhelmed by the abnormal performance of few IPO stocks and consistently overestimate the value of other IPO stocks.

12 12 Islam, Malik and Uddin 1.10 Figure 1 Wealth Relative of IPOs to Industry Mean Median Cohort Month Figure Wealth Relative of IPOs to Market Mean Median Cohort Month

13 Journal of Finace and Banking 13 Performance by Size of IPO: Empirical evidences in other countries indicate that the performance of IPO stocks in the aftermarket vary significantly by size of issue. Smaller issues underperform their industry counterpart and the market much more than larger issues. Do IPO stocks behave the same way in the aftermarket in Bangladesh? This has been examined by categorizing IPO stocks of each year into size quartiles. This categorization has been done on yearly basis because the size of IPOs increased over time significantly with the growth of the market and the largest IPO of an earlier year is likely to appear as a smaller IPO in the later year. The performance of IPO stocks by size quartiles has been presented in Table 3. Analysis of wealth relative to industry by size quartile indicates that majority IPOs belonging to the fourth (largest) quartile outperformed the market marginally in the first month in the secondary market. In the long run IPOs of all quartiles underperformed their industry counterpart during the entire period under study. However, IPOs belonging to the first three quartiles underperformed the industry more than those belonging to the fourth quartile. Performance of IPOs belonging to the first two quartiles continued to slide down till the fourth year, while that of third quartile continued to slide down till the end of fifth year. The median wealth relatives to industry came down to and at the end of fourth year respectively for the first and second quartile. This means that IPO investors are worse off by percent and percent respectively compared to their industry counterpart. The median wealth relative for the third quartile went down to at the end of fifth year. The performance of IPOs belonging to the fourth (largest) quartile continued to decline during the first two years and started to recover thereafter. The median wealth relative to industry of this quartile came down to at the end of second year, but recovered to by the end of fifth year. Three findings emerge from the analysis of wealth relatives to industry by size quartile. First, IPOs belonging to all size quartiles underperformed the industry in the long run. Both median and mean wealth relatives of all quartiles remained less than unity in the long run aftermarket. Second, the variation of IPO performance among different size groups in Bangladesh is not as high as it had been documented in other countries. Only IPOs belonging to fourth (largest) quartile perform better than those belonging to other three quartiles. This is different from the evidences in other markets where only the smallest IPOs perform the worst in the aftermarket. Third, the median wealth relatives to industry for the first three quartiles reached their worst in the third or fourth year, but the same for the fourth quartile reached its worst in the second year.

14 14 Islam, Malik and Uddin Table 3 Performance of IPO Stocks by Size Quartiles Month Quartile 1 (Smallest) Wealth Relative to Industry by Size Quartile Quartile 2 Quartile 3 Quartile 4 (Largest) Median Mean Median Mean Median Mean Median Mean Month Quartile 1 (Smallest) Wealth Relative to Market by Size Quartile Quartile 2 Quartile 3 Quartile 4 (Largest) Median Mean Median Mean Median Mean Median Mean Analysis of wealth relative to market indicates that in the short run, the majority IPOs of smallest quartile underperformed the market marginally. Their median wealth relative to market declined to at the end of second month. This means that investors lost about 5 percent compared to the market. Similarly, the median wealth relative to market for IPOs belonging to quartile two and quartile three declined to and respectively at the end of second month. Interestingly IPOs belonging to fourth quartile outperformed the market marginally by the same

15 Journal of Finace and Banking 15 time. This contrary to the evidences in other markets where smallest IPOs outperformed the market and largest IPOs underperformed the market in the short run. In the long run, there is no consistent pattern of IPO price performance when they are classified by size. During the first two years, IPOs belonging to the smallest quartile outperformed the market by around 5 percent. In the third and fourth years, these IPOs underperformed the market significantly. Their median wealth relative to market declined to by the end of fourth year. In other words, investors of these IPOs lost about 30 percent compared to the market. In the fifth year, however, they recovered fully. It is notable that the mean wealth relative to market continued to be well above unity throughout the period under study. This implies that the outstanding performance of few IPOs in this quartile contributed to their high average performance. Majority IPOs belonging to the second quartile underperformed the market significantly till the end of fourth year. The median wealth relative of these IPOs to market reached its worst to at the end of third year. These IPOs however started to recover in the fourth year. By the end of fifth year they outperformed the market by about 10 percent. Majority IPOs belonging to the third quartile underperformed the market in the first year, but outperformed the market in the second year. Their median wealth relative to market reached values of and at the end of first year and second year respectively. Till then, their performance started to decline again and reached a median value of at the end of fifth year. However, their mean wealth relative to market remained well above 1 throughout the period under study. By the end of fifth year, their mean wealth relative to market reached This essentially indicates that the outstanding performance of few IPOs in this quartile contributed to a high mean wealth relative to market. The fourth quartile IPOs outperformed the market excepting in the second and third year. These large IPOs had a median wealth relative of at the end of second year, but started to recover in the third year. Two findings appear from the analysis of wealth relative to market by size of IPOs. First, there is no consistent pattern of IPO price performance relative to the market by the size of issue. Second, the majority IPOs in all but third quartile recovered in the long run and outperformed the market by the end of fifth year. Performance by Year of Issue: In many countries, it has been documented that IPOs issued in a hot issue market performed the worst in the long run while IPOs issued in a cold issue market outperformed the market. In Bangladesh, the majority IPOs in all years of issue underperformed their industry counterpart till the fifth year aftermarket. The degree

16 16 Islam, Malik and Uddin Table 4 Long Run Price Performance of IPO Stocks by Year of Issue Wealth Relative to Industry by Year of Issue Month Median Mean Wealth Relative to Market by Year of Issue Median Mean Note: The performance of IPOs issued before 1994 and after 2001 had not been reported in this table.

17 Journal of Finace and Banking 17 of long run underperformance relative to the industry had been worst for IPOs issued in 1994, 1996, 2000 and In Bangladesh there had been a hot issue market in The performance of IPOs issued in 1996 continued to slide down till the fifth year and by then, the median wealth relative declined to This means that investors of these IPOs are around 47 percent worse off the comparable firms in the industry. IPOs issued in 2000 and 2001 experienced even worse performance compared to other firms within the same industry. The median wealth relative of IPOs issued in 2000 to industry reached a low of in the fourth year while the same for IPOs issued in 2001 reached their low of at the end of third year. After the market crash in 1996, the stock market in Bangladesh remained depressed for a few years. IPOs issued performed relatively better for two years in the aftermarket. Their median wealth relative to market at the end of second year reached , and respectively for the IPOs issued in 1997, 1998 and However, they could not sustain their performance and their median wealth relative to industry continued to decline for the next few years. The mean wealth relative to industry also had similar values. When compared to the market, the IPOs issued in the hot issue market underperformed in the long run. The median wealth relative to market of IPOs issued in 1996 continued to decline till the end of fifth year and reached Similar decline had been observed for IPOs issued in 2000 and The median wealth relative of IPOs issued in 2000 reached at the end of fifth year while the same for IPOs issued in 2001 reached as low as at the end of fourth year. On the other hand, IPOs issued in the cold issue period did not underperform the market by the same extent. In fact IPOs issued in 1998 outperformed the market significantly throughout the period under study. Their median wealth relative to market reached as high as at the end of third year, but declined to at the end of fifth year. IPOs issued in 1999 outperformed the market till the fourth year, while IPOs issued in 1996 outperformed the market till the second year. Similarly IPOs issued in 1994 and 1995 underperformed the market till the third and second year respectively, but recovered significantly during the later years. However their mean wealth relative to market remained more than unity throughout the period under study. Performance of IPO Stocks by Industry: Empirical studies document that few industries had been subject exogenous shocks and IPOs in these industries provided high return in the short run, but eventually underperformed in the long run. For this reason, the performance of IPO stocks in the aftermarket has been analyzed by industry. Analysis of wealth relative to industry by industrial sector reveals startling findings.

18 18 Islam, Malik and Uddin Table 5 Short and Long Run Price Performance of IPOs by Industry Month Leather MF Insurance NBFI Textile Food Pharma Cement Paper Misc Engin. Ceramic Bank Median Wealth Relative to Industry Mean Wealth Relative to Industry Median Wealth Relative to Market Mean Wealth Relative to Market

19 Journal of Finance and Banking Volume 9, Number 2 December 2011 There have been strong evidences of IPO underperformance that can be attributed to specific industries. IPOs belonging to five industries are underperformed severely in the aftermarket. These are Leather, Engineering, Paper, Ceramic and Food. Wealth relative of IPOs in the Leather sector to industry continued to decline gradually till the end of fifth year. The median wealth relative to industry came down to as low as in the fifth year. That means, investors in these stocks were worse off their industry counterpart by about 90 percent. Similarly, the median wealth relatives of IPOs in the Engineering, Paper and Ceramic sector declined to , , and respectively at the end of fifth year. In other words, IPO investors in these sectors were worse off their industry counterpart by more than 50 percent. IPOs of textile, mutual fund and cement industries declined gradually till the third year but recovered slightly in the next two years. Their median wealth relatives came down to , and respectively at the end of third year but increased to , and respectively at the end of fifth year. Similarly IPOs of insurance industry declined till the second year but recovered significantly by the end of fifth year. Their median wealth relative to industry came down to in the second year but recovered gradually to at the end fifth year. On the other hand, in few sectors IPOs outperformed the industry in most of the years. The median wealth relative to industry of IPOs in the banking sector remained above unity in all years excepting for second and third years. Their median wealth relative to industry went up to at the end of fourth year. Similarly, in the nonbank financial institution sector the median wealth relative remained above unity in all years excepting for third year. Their median wealth relative also reached its peak at in the fourth year. Analysis of wealth relatives of IPOs to market by industrial sector also reveals similar industry specific performance, but few industries distinctly outperformed the market. IPOs of engineering, leather, food, paper and textile underperformed the market throughout the period under study. Their median wealth relatives to market declined to , , , and respectively at the end of fifth year. On the other hand, IPOs of bank, non-bank financial institutions and cement industry outperformed the market significantly throughout the period under study. Their median wealth relative to market reached as high as , and respectively at the end of fifth year. In addition to these, IPOs of insurance and pharmaceutical industries outperformed the market mostly during the period under study. Analysis of wealth relatives to market reveals two major findings. First, IPO price performance is attributable mostly to industry. IPOs of few industries were extreme

20 20 Islam, Malik and Uddin out-performers while IPOs of few other industries were severe under-performers in the long run. Second, IPOs of most of the industrial sectors, particularly the long run out-performers, underperformed in the short run. These findings indicate that IPO investors were overwhelmed by the exogenous shocks in particular industries which led to frustrations in the long run. CONCLUSIONS AND POLICY IMPLICATIONS This study had been conducted to analyze the short and long run price performance of IPO stocks in the aftermarket in Bangladesh. In the short run, IPOs provide abnormal return that peaks in the first month after the issue. This is a bit shorter than other markets where short run price rally peaks in the second month. In the long run IPOs severely underperform their respective industry till the fourth year. Majority IPOs also underperform the market, but the degree of underperformance is not as severe as it is to market. The long run underperformance of IPOs is attributable to the hot issue markets and few industries. In Bangladesh, IPOs in Leather, Engineering, Paper, Ceramic and Food industries severely underperformed while IPOs in banks and non-bank financial institutions industries outperformed in the long run. The findings are consistent with the evidences in other markets. The behavior of IPO price performance has significant policy implications for investors, firms, regulators and policymakers. It appears that IPO investors are overwhelmed by the exogenous shocks in few industries that dry up in the long run. There is scope for research why investors make the same mistake and the inefficiency continues to exist in the market. It is possible that firms resorts to earnings management immediately before the IPO issue, which creates over-optimism among the investors about the long run performances of IPOs. In that case, the regulators and policymakers have to frame the rules to minimize the possibility of earnings management by issuing firms. REFERENCES: Affleck-Graves, S. Hegde, and R.E. Miller (1996), Conditional Price Trends in the Aftermarket for Initial Public Offerings, Financial Management, 25(4), Aggarwal, R., R. Leal, and L. Hernandez, (1993), The Aftermarket Performance of Initial Public Offerings in Latin America, Financial Management, 22(1), Ahmed-Zaluki, Nurati A., Kevin Campbell and Alan Goodacre (2009), Earnings Management in Malaysian IPOs: the East Asian Crisis, Ownership Control and Post-IPO Performance, Working Paper, Universiti Utara Malaysia. Allen, F. and G.R. Faulhaber (1989), Signaling by Underpricing in the IPO Market, Journal of Financial Economics, 23(2),