Printer Friendly

The aftermarket performance of initial public offerings in Latin America.

Numerous studies examining the performance of initial public offerings (IPOs) in the United States and other countries document the existence of positive average initial returns. However, recent evidence suggests that in the U.S., what appears to be underpricing in the short-run may be overpricing when one focuses on the long-run. This raises important questions regarding the efficiency of the IPO market.

This paper empirically examines the performance of IPOs in both the short-run and long-run based on a sample of 62 Brazilian offerings in 1980-1990, 36 Chilean IPOs in 1982-1990, and 44 Mexican IPOs in 1987-1990. These three Latin American markets have recently been attracting considerable foreign interest, and all three have been among the world's best performers in recent years.(1) During 1991, the annual return (in US$) was 151.6% for Brazil, 103.0% for Mexico, and 90.1% for Chile, relative to 27.2% for the U.S., 11.6% for the UK, and 8.3% for Japan. The market correlations with the Standard & Poor's 500 Index were 0.12 for Brazil, 0.40 for Chile, and 0.54 for Mexico during the five-year period ending December 1991, using monthly returns. These markets provide substantial opportunities for portfolio diversification.

The three Latin American countries being studied are involved in ambitious privatization programs in which capital markets are playing an important role. The Chilean privatization of pension funds resulted in billions being invested in its capital market. In many Latin American countries, there is also an attempt to restructure and "deregulate" the economy and open up the stock markets to foreign investors. From January 1991 to June 1992, more than $50 billion was invested in Latin America (Business Week |3~), a significant portion of which may have been "capital flight" money returning. The renewed interest about the region is timely for this study.

Two major methods for going public are used around the world. The initial public offering can be made by either a fixed-price offer or offers for sale by tender. In the U.S. and UK, IPOs are generally issued via the fixed-price offer method, where potential investors specify the number of shares to which they wish to subscribe at a preannounced price. French companies tend to use tender offers in which applicants specify a price (at or above a minimum price) and a quantity of shares. After receiving the applications, an offering price is set which determines the cut-off below which applications are rejected. Both Brazil and Mexico use a fixed-price procedure, while the Chilean IPO market uses an auction procedure.

The study of these markets provides insight on some of the most important and recently best-performing emerging markets. It also adds to the international evidence on IPO performance. It is also useful to study these three countries because Chile uses a unique auction process to determine the prices, while IPOs in Brazil and Mexico are offered at a fixed price similar to the U.S. These differences allow for interesting comparisons to be made on the methods of going public and promotes discussion about the best method for going public.

The results provide evidence consistent with previous empirical findings that report short-term excess returns associated with IPOs. The long-run results for Brazil and Mexico are also similar to the findings of Aggarwal and Rivoli |1~ and Ritter |22~, documenting significant overpricing of U.S. initial public offerings.

The rest of the paper is organized as follows: Section I is a review of the international evidence on IPOs. Section II provides a background on the three emerging markets. The data and methodology are discussed in Section III. Section IV reports the empirical results, and the conclusions are contained in Section V.

I. The International Evidence on IPOs

The international evidence on IPOs finds strong underpricing in the short-run, but negative long-run returns are found in nine out of ten studies, as summarized in Exhibit 1. A number of studies examining the performance of IPOs in the United States document the existence of short-run excess returns in the IPO market.

It is also important to study the performance of IPOs in the long-run. Aggarwal and Rivoli |1~ note that, in the short-run, imperfections exist in the IPO market, such as restrictions on short selling and active price stabilization that is provided to new issues by the associated investment bankers. As Ritter |22~ points out, if systematic price patterns exist in the long-run, then this raises questions concerning aftermarket efficiency: from the investors' point of view, the ability to develop trading strategies is of interest; and from the issuer's point of view, important questions evolve about the costs of going public.

The empirical evidence in the U.S. seems to suggest that IPOs underperform in the long-run relative to the overall market. Ibbotson |9~ reports that IPOs underperform by an average of approximately one percent per month in the second through fourth years, although he emphasizes that he cannot reject the hypothesis that the abnormal performance is zero. Ritter |22~ finds that the matching firm-adjusted cumulative average returns, exclusive of initial returns, decline to -29.1% by the end of month 36. Similarly, Aggarwal and Rivoli |1~ report market-adjusted returns exclusive of initial returns to be -13.7% after 250 trading days for a sample of 1,598 NASDAQ-traded firm commitment IPOs during the period 1977-1987. The December 2, 1985 Issue of Forbes magazine also concluded that IPOs in the U.S. underperform the market. These empirical findings suggest another explanation for the positive average initial returns of IPOs. Instead of being systematically underpriced, IPOs may be correctly priced, but investors, on average, overvalue the new issues in the initial aftermarket trading.

The Australian and UK findings are consistent with U.S. patterns of positive initial returns followed by underperformance. Finn and Higham |7~ study the joint process of initial-issue-cum-listing for 93 Australian issues during the period 1966-1978. The average initial market-adjusted return is 29.2%. However, the one-year market-adjusted return from the closing price on day one results in significant mean returns of -6.52%. Levis |17~ reports average first day returns of 14.3% for 712 IPOs between 1980 and 1988 in the UK and long-run underperformance of -30.59% by the third year after the offer.

Exhibit 1 also presents the short-run/long-run IPO returns for Canada (Jog and Riding |11~), Germany (Uhlir |25~), Hong Kong (Dawson |4~), Japan (Dawson and Hiraki |5~), Korea (Kim and Lee |12~), Malaysia (Dawson |4~), the Netherlands (Wessels |26~), Singapore (Dawson |4~ and Koh and Walter |14~), and Switzerland (Kunz and Aggarwal |15~).

To summarize, the international evidence, as presented in Exhibit 1, consistently finds short-term excess returns. The significant long-run performance results excluding TABULAR DATA OMITTED the first day returns are always negative. The long-run results must be interpreted carefully because many markets are volatile, plus a few companies tend to dominate value-weighted indexes used in many cases to do the market adjustment.

French companies use a variety of processes to go public. Out of the 131 IPOs studied by Husson and Jacquillat |8~ between 1983 and 1986, 11% use a competitive auction procedure similar to a block trade, 36% use a sealed-bid auction procedure, 18% use tender offers, and the remaining 35% use a combination of auction and tender offers. Average initial excess returns of 4.2%, 3.9%, 11.4%, and 35.1%, respectively, are reported. Excess returns of 4.0% for the total sample are reported in Exhibit 1.

Jenkinson and Mayer |10~ further examine the extent of underpricing for 11 French issues during 1986-1987 and 20 UK issues involved in privatization programs during 1979-1987. The average discount on the first day's closing market price relative to the offer price is 25.05% for French tender offers, 32.79% for 14 UK fixed-price offers, and only 2.50% for 6 UK tender offers. An average discount of 22.20% for all UK privatizations is reported in Exhibit 1.

II. The Latin American Markets and the Process of Going Public

Significant changes have taken place in global financial markets during the last decade. The emerging markets of Latin America have begun to attract considerable interest. The market capitalization of Mexico was US$ 98.18 billion, Brazil US$ 42.76 billion, and Chile US$ 27.98 billion at the end of 1991. These markets weighed 16.10%, 6.80%, and 4.45%, respectively, in the International Finance Corporation's Composite Index during 1991. While these numbers increased from their levels of the early 1980s, combined they still represent less than five percent of the U.S. market capitalization.
Exhibit 2. Rate of Inflation and Market Capitalization,(a)
1980-1990

 Inflation (%) Market Capitalization

Year Brazil Chile Mexico Brazil Chile Mexico

1980 83 35 26 9.2 9.4 13.0
1981 106 20 28 12.6 7.1 10.1
1982 98 10 59 10.3 4.4 1.7
1983 142 27 102 15.1 2.6 3.0
1984 197 20 66 29.0 2.1 2.2
1985 227 31 58 42.8 2.0 3.8
1986 145 20 86 42.1 4.1 6.0
1987 230 20 132 16.9 5.3 8.4
1988 682 15 114 32.2 6.9 13.8
1989 1287 17 20 44.4 9.6 22.6
1990 2938 26 27 16.4 13.7 32.7
1991 441 22 23 42.8 28.0 98.2

Notes:

a Rate of inflation for the calendar year and market
capitalization at the end of each year. Market capitalization
in US$ billions.

Sources: Emerging Markets Factbook, International Finance
Corporation, 1989-1992; International Finance Statistics,
International Monetary Fund, 1989-1992.


In Brazil, there are several regional stock exchanges, but only those in Rio and Sao Paulo have significant trading volume. Most transactions are conducted on the floor of the exchange, although the stocks of large companies simultaneously trade electronically. An independent market authority, the Securities Commission (Comissao de Valores Mobiliarios -- CVM), was established in 1978. The Central Bank was responsible for monitoring the market before the CVM was formed.

In Chile there are three stock exchanges, the largest and most important being the Santiago Exchange. On the floor of the Santiago Exchange there are two posts, and all stocks are assigned to one of these two posts. There is no specialist for each stock. Instead, each member firm (42 member firms in 1992) has one representative at each post. Prices are determined using an open outcry method in which all the member firms are involved.

The Bolsa Mexicana de Valores (BMV) was formally created in 1907 in Mexico City. Operations at the Bolsa are well-automated with transactions being effected using the auction method. The markets have performed well with the recent privatization programs in Latin America, more favorable regulations for foreign investors, permission for domestic firms to issue securities overseas, and enthusiastic demand by institutional investors. Exhibit 2 presents the rate of inflation and the market capitalization for each country in the sample.

The percentage of total market capitalization accounted for by a few large companies is much greater in these countries than in the U.S. The share of market capitalization held by the ten largest stocks at the end of 1991 amounted to 31.9% in Brazil, 49.3% in Chile, and 55.0% in Mexico, compared to less than 15% in the U.S. In Brazil, the top five financial institutions in the underwriting ranking were responsible for about 65% of the total volume issued in 1988. In all three markets, the number of IPOs is quite small and there is heavy oversubscription. These factors may lead to severe misevaluation of the new issues in the aftermarket.

In Brazil, a firm may go public by issuing common or preferred stocks. Preferred stocks are analogous to class B common shares in the U.S., which have differential voting rights. Most common shares, which are similar to the U.S. class A shares with superior voting rights, are not liquid because they are held by controlling shareholders that do not trade them. Most Brazilian IPOs analyzed in this paper are preferred stock. These preferred stocks have some of the same characteristics as the common stock in other countries, and their dividends are variable.(2) In Chile and Mexico, all issues in our sample are those of common stocks.

The Brazilian firms in the sample had been in business for 25 years, on average, before going public. Their asset size averages US$ 25 million. Many Chilean IPOs are a result of privatizations. The privatized firms included well-established and fairly large utilities and mining companies and even the national airline carrier. This suggests that IPOs in these countries are of older firms than typical U.S. IPOs of small growth companies.

The process of IPOs is unique in Chile since these shares are not sold through underwriters; instead, brokers sell them directly to the public on a commission basis similar to a block trade. The issue is offered through special auction sessions or during normal trading hours on the exchange floor. The special sessions are popular for issues involved in privatizations. The issuing firm can auction the issue in several blocks. Since there are no underwriters, there are no commitments.

III. Data and Methodology

The data on Brazilian new issues were collected from the Securities Commission's registration filings. The 1980-1990 period was chosen because it corresponds to the new regulatory environment introduced with the formation of the new Securities Commission. Before this, the new issues were regulated through a specific department of the Central Bank; the regulation was not very demanding in terms of disclosure requirements. Aftermarket prices were collected from the Sao Paulo Daily Stock Exchange Bulletins. Prices for one, two and three years after the offer were collected. If a market price on an anniversary was unavailable, the market price on the nearest market day within a ten-day trading window around the first-, second- and third-year anniversary of the offering, respectively, was used.(3)

A total of 122 new stock issues were registered during this period. No price data were available for 41 issues because they were acquired by funds and banks involved in regional or national development programs, resulting in 81 publicly traded issues. The final sample of 62 IPOs includes only those firms that started trading within 60 days of the Securities Commission's approval.(4) All issue prices were adjusted for stock splits. The market index for Brazil was adjusted five times during the sample period due to high inflation. These changes are appropriately incorporated in the analysis.

Preferred stock alone was issued in 51 of the 62 cases. Preferred share returns are reported for an additional nine issues that included both common and preferred stock because of lack of liquidity in the aftermarket for common stock. Only two issues were exclusively common stock, and common share prices are used for these. The average issue size was approximately US$ 4.4 million for the sample. The highest annual average size was US$ 5.4 million in 1986.(5) The anti-inflation stabilization plan, announced in February 1986, caused a boom in the stock market, generating what could be called a "hot issues" market. From the total amount issued during the 11-year sample period, 72% occurred after 1985, and 39% occurred in 1986 alone.

The Chilean data were collected from the Bolsa de Comercio de Santiago for the period 1982 to 1990. Out of a total of 36 Chilean IPOs in the sample, 21 involved privatizations. As discussed by Perotti and Guney |20~, in Chile the first wave of privatization took place during the period 1974 to 1981 and a second wave from 1986 onwards. An auction process is used to go public in Chile. The issuer does not report the auction price to the Chilean Securities Commission or the Santiago Stock Exchange. The offer prices were obtained in 19 cases from the individual firms.

There were 44 new offerings in Mexico that started trading on the Bolsa Mexicana de Valores during the period 1987-1990. The offer price, closing price on day one, the month-end prices, and closing price one year after the offer were available for new issues in Mexico. Month-end prices were available only from the month an IPO started trading to the end of the calendar year. Thus, if an issue came out in January, then month-end prices were available from January to December, but if the issue came out in December, only the closing price for the month of December was available. All price data on these issues were collected from the exchange and the International Finance Corporation.

The total return for stock i in period t is calculated as in Equation (1) where |P.sub.it~ is the price of stock i at time t and |P.sub.i0~ is the offer price:

|R.sub.it~ = |P.sub.it~ / |P.sub.i0~ - 1. (1)

The return on the market index during the same time period is calculated as in Equation (2). The market index value at time t is |P.sub.mt~, and the market index value on the offer day is |P.sub.m0~:

|R.sub.mt~ = |P.sub.mt~ / |P.sub.m0~ - 1. (2)

The market-adjusted abnormal return for each IPO on day t is computed as:

|AR.sub.it~ = |(1 + |R.sub.it~) / (1 + |R.sub.mt~) - 1~ X 100, (3)

where |AR.sub.it~ is the market-adjusted abnormal return of stock i at time t, |R.sub.it~ is the total return on the ith security from offering to t trading days after the initial offering, and |R.sub.mt~ is the total market return for the corresponding period.(6) The Sao Paulo Stock Exchange Index (BOVESPA) is the market proxy for Brazil, the General Stock Price Index (IGPA) for Chile, and the Price and Quotation Index (IPC) for Mexico. All three of these indexes are market-capitalization weighted similar to the Standard & Poor's 500. In order to evaluate the performance of IPOs purchased in the aftermarket, market-adjusted returns are calculated as in Equation (3), but now |R.sub.it~ is the total return on the ith security from day 1 to t trading days after the initial offering, and |R.sub.mt~ is the total market return for the corresponding period.

Wealth relatives are also calculated using the procedure employed by Ritter |22~ and Levis |17~, as seen in Equation (4). WR is the wealth relative; |R.sub.it~ is the return of stock i on day t from the offer day; |R.sub.mt~ is the market return during the same time period. The total number of IPOs in the sample is represented by N. A wealth relative above one implies that IPOs outperformed the market in that period. A wealth relative below one indicates underperformance. Similarly, wealth relatives from day 1 are also calculated.

|Mathematical Expression Omitted~.

IV. Empirical Results

A. The Short-Run and Long-Run Performance of IPOs

Exhibit 3 reports the aftermarket performance for 62 Brazilian new issues. Consistent with previous studies on IPOs in Asian emerging markets reported in Exhibit 1, large positive mean and median excess returns, of 78.5% and 36.5%, respectively, are found for Brazil as seen in Panel A of Exhibit 3. The t-statistic of 6.83 is significant at the five percent level.(7)

Only eight of the 62 individual IPOs showed negative market-adjusted performance on the first day of trading. On average, significant gains accrue to the holders of unseasoned new issues on the first day of trading. Most of the initial price adjustment appears to take place on day 1. The median total excess returns were 44.9% by the end of the first month, and then dropped to 38.8% by the end of three months. These initial returns for Brazil are much higher than those reported in Exhibit 1 for all other countries, except Korea and Malaysia. Leal |16~ finds that Brazilian IPOs with high initial returns tend to be associated with firms that are small in size, have high leverage, and poor past performance.

Inflation may also influence the high returns. There are usually a few weeks between the day the offer price is set and the offer date. When inflation is high, the market price on the first day of trading may also reflect some compensation for inflation during the period. The Brazilian sample includes only issues that started trading within 60 days of the offer day in order to limit inflationary effects.

Since "hot" new issues tend to be allocated by underwriters and selling groups to favored clients, many investors are unable to purchase the issues at the offering price. Therefore, it is important to examine new issues from the viewpoint of the investor who purchases the stock in the aftermarket. Panel B of Exhibit 3 shows the excess returns, assuming purchase of the issues at the closing price on the first day of trading. The short-run results for months 1, 2, and 3 are not significantly different from zero at conventional significance levels as seen in Panel B of Exhibit 3. This implies that the aftermarket is quite efficient, and most prices adjust fairly quickly after trading starts. There are individual issues that continue to increase in price and TABULAR DATA OMITTED others whose price declines, but, on average, there is not any significant pricing bias.

Evidence on the U.S. new issues market suggests that IPOs underperform the market in the long-run. Aggarwal and Rivoli |1~ find the mean and median market-adjusted returns to be negative for investors who purchase new issues at the offering price and hold for one year. Ritter |22~ finds similar results over a three-year holding period assuming a purchase on the first day. In the Brazilian case, the mean excess return of 39.2% for an investor who bought at the offering price and held for one year is significantly different from zero, and for three years it is an insignificant -25.6%. Similar patterns are evident in the wealth relatives, which drop from 1.79 on day 1 to 0.60 at the end of three years.

As shown in Panel B of Exhibit 3, investors who purchased the issues in the aftermarket at the closing price on the first trading day and held for one year received mean and median excess returns of -9.0% and -51.1%, respectively. The mean excess returns become -34.9% and -47.0% for holding periods of two and three years, respectively.(8) By the end of year 3, the wealth relative drops to 0.67. The results must be interpreted cautiously due to the small samples and extremely volatile and inflationary economies. In Brazil, the annual rate of inflation during the sample period reached its peak at 2938% in 1990, while the minimum during the sample period was 83% in 1980.

The returns in year 3 are not impacted by survivorship bias since the 57 firms with one-year excess returns continued to be listed in year 3. All 57 firms could not be utilized in the year 2 and year 3 analysis either because they did not trade during the 10-day trading window (due to low liquidity) or because their second- and/or third-year anniversary was outside the sample period.(9)

TABULAR DATA OMITTED

The results for Chilean IPOs are presented in Exhibit 4. Offer prices were available only for a limited number of firms. If the IPOs were purchased at the offer price and held for one day, then mean market-adjusted excess returns of 16.3% would be earned, although this number is not different from zero at conventional levels of significance due to small sample size. Although the magnitude of Chilean mean initial returns is above those reported for the United States and other developed markets, the median is only 0.5%. This may be a result of the auction process used to TABULAR DATA OMITTED go public. It is consistent with previous empirical evidence, such as from France, indicating that auctionlike processes result in lower initial returns. The mean returns continue to be positive but not significant at the five percent level for months 1, 2, and 3, as seen in Panel A of Exhibit 4. The wealth relatives also continue to be above one until the end of the third month. For the firms involved in privatization, the short-run mean excess returns are also not significant, as seen in Panel B of Exhibit 4.

The mean market-adjusted short-run excess returns for purchasers who buy at the closing price on day 1 and hold for one, two and three months are positive but not significant, as shown in Panel C of Exhibit 4. These results are similar to those reported for Brazil. The privatization group also has positive mean excess returns, but none of them are significant at conventional levels as seen in Panel D of Exhibit 4. The mean day 1 return for the privatization sample is 7.6%, while the same mean return for the whole sample is 16.3%.

In the long-run, after one, two and three years, the mean market-adjusted returns are -9.8%, 33.9% and 0.8%, respectively, assuming purchase at the offer price. The corresponding medians of -23.0%, -12.2% and -19.0% are all negative. The mean excess returns for privatization firms of -29.9% after year 1 are significant. Because the sample size is very limited, the results for years 2 and 3 are not reported.

If the IPOs are bought in the aftermarket on day 1 and held for one, two or three years, then none of the results are significantly different from zero even though mean excess returns are 1.1%, -2.0%, and -23.7% for years 1, 2, and 3. The corresponding mean values are -5.5%, -5.4%, and -13.7% for the privatization group. The group of privatization firms has negative mean and median excess returns for all three years. At the end of year 3, the full sample and the privatization group have wealth relatives below one. The Chilean results for the long-run analysis are similar in direction to those reported by Aggarwal and Rivoli |1~ and Ritter |22~ for the U.S. But, these results are not significant using standard t-tests for mean returns and nonparametric tests for the medians (not reported). The small sample size does limit the conclusions that can be drawn from the results.

In the case of Mexico, the mean positive excess return of 2.8% for day 1 is not statistically significant at conventional levels, as reported in Panel A of Exhibit 5. The median market-adjusted return is only 0.7%. The mean continues to be positive for the next three months. Since a large number of IPOs took place during the third quarter of 1987, the sharp drop in excess returns during month 3 is a result of the October 1987 crash. As a result of the crash, the Mexican market dropped by nearly two-thirds, and recent IPOs dropped by an even larger amount. This decline continued for months. If the IPOs were purchased TABULAR DATA OMITTED in the aftermarket at the closing price on day 1 and held for one, two or three months, then the mean returns are positive but not significant, as seen in Panel B of Exhibit 5.

As pointed out earlier, closing prices were available for 38 Mexican IPOs one year after the offering.(10) After a holding period of one year, the market-adjusted returns, relative to the offer price, become negative with mean and median values of -17.7% and -35.3%, respectively. Investors who purchase in the aftermarket at the first day's closing price and hold for one year have significant underperformance with excess returns being -19.6%, as seen in Panel B of Exhibit 5. None of the firms in our Mexican sample was involved in privatization. Even though the privatization program was actively being pursued during these years, most privatized firms were first placed in private hands, and after a time lag, these firms were expected to issue new stock to the public. This contrasts with other markets where privatizations have been accomplished concurrently with their IPOs (see Perotti and Guney |20~).

Finally, Exhibit 6 shows the long-run behavior of the IPOs relative to initial returns. There seems to be a negative relationship between the first day excess returns from the offer price and the first- and third-year returns from day 1 prices in Brazil, as shown in Panel A of Exhibit 6. This suggests that investors tend to buy the new shares at an overvalued price on day 1, and later realize losses after one and three years. This result is in line with the evidence in other markets that IPOs with higher initial returns underperform in the long-run. For Chile and Mexico, this pattern is not clear.

B. Explanations for IPO Price Behavior

Tinic |24~ has explained the short-run underpricing as protection against legal liability. But the implications of legal liability are quite different in Latin America relative to the U.S. The claims for compensation due to lack of due diligence are much more difficult to carry out. Explanations other than the risk of legal liabilities are more appropriate to explain underpricing in Latin American. Additionally, results by Drake and Vetsuypens |6~ question Tinic's hypothesis for the U.S. markets.

The information asymmetry problem is critical in emerging markets. In a thin market, underwriters may have the market power to enforce a large discount on a high quality firm to provide a firm commitment arrangement and to compensate for the performance of firms that are subsequently revealed to be of low quality. As there are not many underwriters and the prestigious ones form syndicates, there are not many alternatives left to the issuer. Logue |18~ and Reilly |21~ have advanced the hypothesis regarding the monopsonistic power possessed by investment bankers. Considering the small size of the markets studied here, it is also probable that severe asymmetries exist between investors, as discussed by Rock |23~.

Another interesting implication regards Baron's |2~ model. It states that the initial positive returns exist because underwriters have superior information about the future market price of the IPO and profit on it. In the U.S., empirical tests by Muscarella and Vetsuypens |19~ reject the model by finding initial positive returns when the underwriters place their own IPO. In the case of Chilean IPOs, a market where there are no underwriters and which uses a competitive auction system to place the new shares, the positive initial returns also do not support the hypothesis.

V. Conclusions

Initial one-day returns are found to be 78.5%, 16.7%, and 2.8% for Brazil, Chile, and Mexico. Previous international evidence on new issues consistently finds excess returns in the short-run. The auction process used to go public in Chile may result in the excess returns not being significant. There appears to be no major differences between the Chilean privatization IPOs and the rest of the sample.

The long-run mean market-adjusted return is -47.0% in Brazil after three years. The three-year mean excess return is -23.7% for Chile and the one-year mean excess return is -19.6% for Mexico. They indicate long-run underperformance. For Brazil, there seems to be a negative relationship between the initial returns and the long-run returns, suggesting the overpricing of IPOs on the first trading day. These findings for the Latin American markets are similar to the U.S. and UK pattern of long-run underperformance. Based on the international evidence, it appears that these long-run patterns are not just sample- or country-specific. The evidence from Latin America must be interpreted cautiously because of the small sample sizes and the fact that most IPOs are concentrated during a few years. These phenomenon, in fact, exist in nearly all markets except the U.S. and UK.

The authors thank Kendall Hill, Roger Kunz, Alan Roshwalb, Pietra Rivoli, four anonymous referees, and especially Jay Ritter (the editor) for helpful comments. Reena Aggarwal acknowledges grants from the School of Business Administration, Georgetown University, Center for Business-Government Relations and from the Fulbright Commissions of Brazil and Chile. Ricardo Leal acknowledges grants from the Catholic University of Rio de Janeiro and from the Brazilian Council for Research (CNPq).

1 See Business Week |3~, for example.

2 During 1991, nearly 98% of total trading volume involved preferred stock. Common stock is not liquid and these shares barely trade on the exchange floor. The little activity in common stock is mostly related to foreign controlled companies that are not allowed to issue preferred stock.

3 This window was used for collecting prices in order to avoid having too many missing observations.

4 There were 35 issues that started trading within 30 days of approval. This criteria was used because long periods of time before any trading could cause distorted returns in the presence of high inflation rates.

5 The exchange rate used is the official selling rate on the IPO approval date.

6 Excess returns estimated in this manner are more appropriate since Brazil had very high inflation during part of the sample period, resulting in high nominal returns. Excess returns calculated as |R.sub.it~ - |R.sub.mt~ would result in distortions due to the high nominal returns.

7 The two-tailed t-statistic is calculated as the mean/standard error where the standard error is the standard deviation divided by the square root of the number of observations. The t-statistics assume normality and independence and therefore are biased and must be interpreted cautiously.

8 It should be pointed out that initially the sample started out with 62 issues, but by the end of year it was reduced to 57. The number of firms dropped because four firms were delisted or stopped trading and the fifth did not trade. Unlike CAR computations, the mean holding period market-adjusted returns do not assume reinvestment of the proceeds from delistings.

9 If a stock did not trade during the 10-day trading period, we also tried using the next available price, but it does not change the results.

10 As discussed in the data and methodology section, monthly prices were available only for the rest of the year. This means that if an IPO started trading in the market on November 1, 1990, then prices were available for the end of November (month 1), end of December (month 2), and on November 1, 1991 (year 1). But this IPO could not be included in the month 3 analysis.

References

1. R. Aggarwal and P. Rivoli, "Fads in the Initial Public Offering Market?" Financial Management (Winter 1990), pp. 45-57.

2. D.P. Baron, "A Model of the Demand for Investment Banking Advising and Distribution Services for New Issues," Journal of Finance (September 1982), pp. 955-976.

3. Business Week, June 15, 1992 and June 22, 1992.

4. S.M. Dawson, "Secondary Stock Market Performance of Initial Public Offers, Hong Kong, Singapore and Malaysia: 1978-1984," Journal of Business Finance and Accounting (Spring 1987), pp. 65-76.

5. S.M. Dawson and T. Hiraki, "Selling Unseasoned New Shares in Hong Kong and Japan: A Test of Primary Market Efficiency and Underpricing," Hong Kong Journal of Business Management (1985), pp. 125-134.

6. P.D. Drake and M.R. Vetsuypens, "IPO Underpricing and Insurance Against Legal Liability," Financial Management (Spring 1993), pp. 64-73.

7. F.J. Finn and R. Higham, "The Performance of Unseasoned New Equity Issues: Common Stocks Exchange Listing in Australia," Journal of Banking and Finance (September 1988), pp. 333-351.

8. B. Husson and B. Jacquillat, "Sous-evaluation des titres et methodes d'introduction au second marche (1983-1986)," Finance (December 1990), pp. 123-134.

9. R. Ibbotson, "Price Performance of Common Stock New Issues," Journal of Financial Economics (September 1975), pp. 235-272.

10. T. Jenkinson and C. Mayer, "The Privatization Process in France and UK," European Economic Review (March 1988), pp. 482-490.

11. V.M. Jog and A.L. Riding, "Underpricing in Canadian IPOs," Financial Analysts Journal (November-December 1987), pp. 48-55.

12. E.H. Kim and Y.K. Lee, "Issuing Stocks in Korea," in Pacific Basin Capital Markets Research, S.G. Rhee and R.P. Chang (eds.), Amsterdam, North Holland, 1990, pp. 243-253.

13. I. Krinsky, J.B. Kim, and J. Lee, "Motives for Going Public and Underpricing," Working Paper, McMaster University, 1992.

14. F. Koh and T. Walter, "A Direct Test of Rock's Model of the Pricing of Unseasoned Issues," Journal of Financial Economics (August 1989), pp. 251-272.

15. R.M. Kunz and R. Aggarwal, "Explaining the Underpricing of Initial Public Offerings: Evidence from Switzerland," Journal of Banking and Finance, forthcoming.

16. R. Leal, "Less Competitive Stock Markets and the Validity of IPOs' Ex Ante Uncertainty Measures," Working Paper, Georgetown University, 1992.

17. M. Levis, "The Long-Run Performance of Initial Public Offerings: The UK Experience 1980-1988," Financial Management (Spring 1993), pp. 28-41.

18. D.E. Logue, "On the Pricing of Unseasoned Equity Offerings: 1965-69," Journal of Financial and Quantitative Analysis (January 1973), pp. 91-103.

19. C.J. Muscarella and M.R. Vetsuypens, "A Simple Test of Baron's Model of IPO Underpricing," Journal of Financial Economics (September 1989), pp. 125-135.

20. E.C. Perotti and S.E. Guney, "The Structure of Privatization Plans," Financial Management (Spring 1993), pp. 84-98.

21. F.K. Reilly, "New Issues Revisited," Financial Management (Winter 1977), pp. 28-42.

22. J. Ritter, "The Long-Run Performance of Initial Public Offerings," Journal of Finance (June 1991), pp. 3-27.

23. K. Rock, "Why New Issues Are Underpriced," Journal of Financial Economics (March 1986), pp. 187-212.

24. S.M. Tinic, "Anatomy of Initial Public Offerings of Common Stock," Journal of Finance (September 1988), pp. 789-822.

25. H. Uhlir, "Going Public in the F.R.G.," in A Reappraisal of the Efficiency of Financial Markets, R. Guimaraes, et al (eds.), Berlin, Springer-Verlag, 1989.

26. R.E. Wessels, "The Market for initial Public Offerings: An Analysis of the Amsterdam Stock Exchange," in A Reappraisal of the Efficiency of Financial Markets, R. Guimaraes, et al (eds.), Berlin, Springer-Verlag, 1989.

Reena Aggarwal is an Associate Professor of Finance and Ricardo Leal is a Visiting Instructor, both at the School of Business Administration, Georgetown University, Washington, D.C., and Leonardo Hernandez is a Professor of Finance at the Catholic University of Chile, Santiago, Chile.
COPYRIGHT 1993 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:New Issues Markets Special Issue
Author:Aggarwal, Reena; Leal, Ricardo; Hernandez, Leonardo
Publication:Financial Management
Date:Mar 22, 1993
Words:6385
Previous Article:The long-run performance of initial public offerings: the UK experience 1980-1988.
Next Article:The opening price performance of initial public offerings of common stock.
Topics:


Related Articles
The long-run performance of initial public offerings: the UK experience 1980-1988.
IPO underpricing and insurance against legal liability.
The effect of the trading system on the underpricing of initial public offerings.
Conditional price trends in the aftermarket for initial public offerings.
Don't leave IPOs to the pros.
Assets in place, growth opportunities, and IPO returns.
The impact of the quiet period on IPO volatility.
Underpricing and ex post value uncertainty.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters