DO SEASONED EQUITY OFFERINGS REALLY UNDERPERFORM IN THE LONG RUN? EVIDENCE FROM NEW ZEALAND

DO SEASONED EQUITY OFFERINGS REALLY UNDERPERFORM IN THE LONG RUN? EVIDENCE FROM NEW ZEALAND By Marcus Traill and Ed Vos* University of Waikato Department of Finance Private Bag 3105 Hamilton, New Email: evos@waikato.ac.nz Phone: 0064 7 838 4466 ext 8110 Fax: 0064 7 838 4145 *address all correspondence to Ed Vos ABSTRACT This study investigates the long run performance of seasoned equity offerings on the New Stock Exchange. The sample consists of 39 seasoned equity offerings between 1984 and 1994. The research investigates the relationships between the long run performance and issue size, issuer size, the proportion of issue size to firm size, and survivorship. The results of the research indicate that the sample of SEO firms significantly underperform the market for the five years after the issue. However, the reason for the underperformance is not an issuing effect. The results suggest that the long-run underperformance of SEOs is a more general manifestation of firms delisting from the New Stock Exchange, suggesting that the performance of SEOs are not a special class of share that lose money. Firms who do not survive five years after issuing seasoned stock are responsible for the long-run underperformance of seasoned equity offerings in New. Moreover, surviving seasoned equity offerings do not significantly underperform the market. A share market crash explanation is convincing, and consistent with the survival explanation. In addition, there is evidence that is consistent with firms making use of windows of opportunity that existed in the New market to sell overvalued stock. There is also support for agency costs playing a role in the performance of seasoned equity offerings.

2 CONTENTS PAGE DO SEASONED EQUITY OFFERINGS REALLY UNDERPERFORM IN THE LONG RUN? EVIDENCE FROM NEW ZEALAND... 1 ABSTRACT... 1 SECTION 1 INTRODUCTION...3 SECTION 2 LITERATURE REVIEW...3 2.1. LONG RUN PERFORMANCE OF SEOS...3 2.2. LONG RUN PERFORMANCE OF IPOS...4 2.3. WINDOWS OF OPPORTUNITY...5 2.4. AGENCY THEORY...5 SECTION 3 DATA...6 METHODOLOGY...7 3.1. CUMULATIVE AVERAGE ADJUSTED RETURNS...7 3.2. HOLDING PERIOD RETURNS...7 SECTION 4 RESULTS...8 SECTION 5 CROSS-SECTIONAL AND TIME-SERIES PATTERNS IN THE POST- OFFERING PERFORMANCE OF SEASONED EQUITY OFFERING FIRMS...11 5.1. POST-OFFERING PERFORMANCE BY YEAR OF OFFERING...11 5.2. POST-OFFERING PERFORMANCE CATEGORISED BY FIRM SIZE...12 5.3. POST-OFFERING PERFORMANCE CATEGORISED BY ISSUE SIZE...13 5.4. POST-OFFERING PERFORMANCE CATEGORISED BY INDUSTRY...14 5.5. POST OFFERING PERFORMANCE CATEGORISED BY SURVIVORSHIP...15 5.6. REGRESSION...16 SECTION 6 SUMMARY AND CONCLUSION... 16 FURTHER RESEARCH... 17 REFERENCES... 17 2

3 DO SEASONED EQUITY OFFERINGS REALLY UNDERPERFORM IN THE LONG RUN? EVIDENCE FROM NEW ZEALAND SECTION 1 INTRODUCTION The long run performance of seasoned equity offerings (SEOs) has become the subject of more recent attention. The finding that SEOs are poor long-run performers is more recent (1990s). In some of the older academic literature from the 1960s, e.g., Stigler (1964) and Friend and Longstreet (1967), there is evidence using small samples that issuing firms do poorly in the long run, but this literature has been largely forgotten. 1 It is generally accepted that SEOs underperform in the long run, with different stock markets experiencing virtually identical patterns to the US (UK - Levis (1995), Japan - Cai (1996), and Kang, Kim, and Stulz (1996)). 2 However, it seems unlikely that all equity markets will experience these very similar patterns. Therefore, further investigation of the effect of issuing seasoned stock in a different equity market may provide additional insight into the long-run performance of this corporate financing choice, particularly in small market like New. This study investigates the long run performance of SEOs in the New market. The research will investigate the relationship between the long run performance and potential explanatory variables. This paper examines whether there are any identifiable characteristics of New issuing firms that explain their long-run performance. The hypothesis of the study is that the seasoned equity offerings of New firms on the New Stock Exchange between 1985 and 1994 underperform the market for the five years following the offer. To investigate the hypothesis, the methodology developed in Ritter (1991) is applied to the New sample. 3 In the next section, this paper reviews the current financial literature on the long-run performance of seasoned equity offers. The report then describes the data and research methods in Section 3. Results on the long-run stock price performance following seasoned equity offerings are presented in Section 4 and the several cross-sectional and time-series patterns of the long-run performance are documented in Section 5. The implications of the results are discussed in Section 6 and the paper closes with a brief summary. SECTION 2 LITERATURE REVIEW 2.1. Long Run Performance of SEOs Spiess and Affleck-Graves (1995), using a sample of 1,247 U.S. firms making primary seasoned equity offerings during the 1975-1989 period, find that there are negative long run abnormal returns for seasoned equity offerings in the United States. The median return in the five-year holding-period period following a seasoned equity offering was 10.0%, compared with a median five-year holding-period return of 42.3% for similar size, 1 Loughran and Ritter (1995) at p 47. 2 Loughran and Ritter (1997) at p. 1849. 3 This methodology is also used in Loughran and Ritter (1995) and Spiess and Affleck-Graves (1995). 3

4 non-issuing firms in similar industries. The underperformance persists even after controlling for trading system, offer size, and the issuing firm s age and book-to-market ratio. They conclude that the results of their study are most consistent with managers taking advantage of firm-specific information to issue equity when the firm s stock is overvalued. While it was suggested that the long-run underperformance following both IPOs and seasoned equity offerings could reflect mismeasurement of the relative risk of the sample and matched firms, the magnitude and robustness of the underperformance suggest otherwise. Spiess and Affleck-Graves (1995) also report that the adjusted returns are positive and statistically significant in the first month after the offering. The short-term positive impact following the offering is consistent with the short-term post-issue positive abnormal returns found in several studies such as Loderer, Sheehan, and Kadlec (1991), Tripathy and Rao (1992), and Barclay and Litzenberger (1988). Loughran and Ritter (1995), using a sample of 3702 seasoned equity offerings from 1970 until 1990, report that the average raw return for issuing firms is only 7 percent per year during the five year after the offering, compared to 15 percent per year for non-issuing firms of the same market capitalisation. The authors conclude that firms take advantage of windows of opportunity by issuing equity when they are substantially overvalued. Healy and Palepu (1990), Hansen and Crutchley (1990) and Loughran and Ritter (1997) investigate the operating performance of SEOs. Loughran and Ritter (1997) find that the median issuer substantial improvement in operating measures in the year of issue. The stock price run up in the year prior to the issue is 72 per cent and may be justified if the firm holds the improved operating performance. However, the improvement in operating performance deteriorates to below pre-issue levels. Loughran and Ritter (1997) also find that there is an independent new issues effect. Levis (1993) reports that firms conducting SEOs in the United Kingdom subsequently underperform. Marsh (1979) reports that firms conducting SEOs in the United Kingdom during the 1962 to 1972 outperform the market during the following year then underperform in the second year after the offering. Cai (1996) and Kang, Kim, and Stulz (1996) report the patterns of underperformance of SEOs, similar to those in the US, in Japan. 2.2. Long Run Performance of IPOs The long-run performance of IPOs will be briefly discussed as their performance is similar to SEOs. Ritter (1991) first notes the long-run underperformance of IPOs and the underperformance has been confirmed in several studies, including Aggarwal and Rivoli (1990), Loughran and Ritter (1995), and Loughran, Ritter and Rydqvist (1994). Strang (1995) reports that the long-run underperformance of IPOs is present in New as a result of the overreaction hypothesis and the hot issue market. Spiess and Affleck-Graves (1995) report that the performance of IPOs is similar to the performance of SEOs. They find that the underperformance documented by Ritter (1991) is not an IPO phenomenon, but a more general public offering effect. Brav and Gompers (1997) report that size and book-to-market equity contribute more to underperformance of IPOs than the issuance of new equity. 4

5 2.3. Windows of Opportunity One potential explanation for the long run underperformance of SEOs is the windows of opportunity framework. Loughran and Ritter (1995) report that the long-run underperformance of new issues is consistent with firms taking advantage of transitory windows of opportunity by issuing equity when their stock is substantially overvalued. Bayless and Chaplinsky (1996) use the volume to define what they refer to as hot and cold markets with hot market experiencing high volumes of equity issue. They report that the price reaction to equity issue in a hot market is lower than for a cold market, with this figure being economically significant to firms who will try and raise equity capital at the most favourable terms. 2.4. Agency Theory Another explanation for the long-run underperfomance of SEOs is agency theory. Jensen and Meckling (1976) argue that managers will not always act best interests of the shareholders. The agency costs of an equity issue should be considered. As emphasized by Jung, Kim, and Stulz (1996) and McLaughlin, Safieddine, and Vasudevan (1996), the cash inflow and reduced managerial percentage ownership associated with an equity issue may intensify agency problems and result in lower operating margins. If these increased agency problems are not fully anticipated by the market, they will be manifested in low post-issue stock returns. 5

6 SECTION 3 DATA The sample consists of seasoned equity offerings of operating companies during 1984 through 1994 on the New Stock Exchange. 4 In common with other studies, financial institutions have been excluded. The issue data was obtained from the New Stock Exchange Annual Report (1984 to 1989), the New Sharemarket Review (1990 to 1994), which become the New Stock Exchange Fact Book (1995 to 1998). Share price data was collected from Datex and the National Business Review. Several data restrictions are present in this study. The offer must be a cash offer of ordinary shares. Because multiyear performance of issuers are examined, Healey and Palepu s (1990) procedure, excluding SEOs by the same firm during the five years after an SEO that is in our sample, is followed. These offerings are excluded in order to reduce dependence for the statistical tests. Thus, once a firm has a seasoned equity offering, that firm cannot re-enter the SEO sample until five years from the issue date have passed. Firms that either form mergers or are taken over are excluded from the sample due to the possible influence in measuring an issuing effect. Firms are included who conduct a SEO and subsequently go into liquidation or receivership or are delisted from the New Stock Exchange for breach of requirements or listing rules. The inclusion of such firms should reduce survivorship bias. Table 1 reports the number of SEOs by calendar year. This study contains 39 offerings and is smaller than Loughran and Ritter s (1997) sample of 1338. Healy and Palepu (1990) had a sample size of 93 and Hansen and Crutchley (1990) used 109 issuing firms for their studies. Table 1 Distribution of Sample by Year Year of Issue Total No. Total Gross Total No. of Percentage in Gross Proceeds of SEOs Proceeds SEOs in Sample Sample $ $ 1984 45 330,690,000 5 11% 29,365,707 1985 53 516,310,000 6 11% 148,260,073 1986 51 1,528,200,000 9 18% 171,281,871 1987 32 850,050,000 9 28% 43,582,752 1988 3 372,290,000 0 0% - 1989 6 821,380,000 1 17% 547,204 1990 8 255,250,000 2 25% 44,688,185 1991 6 609,850,000 1 17% 140,000,000 1992 5 535,550,000 0 0% - 1993 14 933,650,000 4 29% 83,713,572 1994 5 460,180,000 2 40% 52,051,485 Total 228 7,213,400,000 39 17.74% 713,490,849 4 The New Stock Exchange only came into existence in 1983. The sample originally contained four SEOs from 1983 but lack of data for these stocks prevented their inclusion in the final sample. 6

7 METHODOLOGY To measure the long-run performance of these seasoned equity offerings, aftermarket returns from purchasing the shares at the closing price on the day of the offering are calculated. The aftermarket consists of the following 60 months. One performance measure was used: Five-year holding-periods for both the SEO and the market index. Cumulative average returns were also calculated. 5 To simplify data collection the event period is defined as the closing price of the stock on the last day of the month following the offering (or the closest day possible). 3.1. Cumulative Average Adjusted Returns The monthly market adjusted returns are calculated as the monthly raw return on a stock minus the monthly market return for the corresponding period. The market adjusted return for stock i in event month t is defined as: ar it = (r seo, it r mkt, it ) The average market adjusted return on a portfolio of n stocks for event month t is the equally-weighted arithmetic average of the market-adjusted returns: AR t = (1/n) ar it i = 1 The cumulative market adjusted aftermarket return performance from event month q to event month s is the summation of the average market adjusted returns: n CAR q,s = (1/n) AR t t=q s 3.2. Holding Period Returns Holding period returns are computed and is defined as: b HPR (i;a,b) = [ Π (1+r it )] 1 i = a Holding-period returns are calculated for the market index also. A wealth relative is calculated to provide an easily understood measure of long-run performance. Similar to Ritter (1991) the wealth relative is the average of the holding period returns of the SEOs divided by the total returns of the market index. WR = _ (1 + average holding-period return for SEO) (1 + average of holding-period return for market index) Thus, a wealth relative less than one is evidence that the SEO underperformed the market. If the offering firm is delisted prior to the anniversary of the holding period of 5 Cumulative average returns were calculated on an equal-weighted basis resulting in a 500% return. The big percentage swings of the small valued firms effected the accuracy of the results on an equal-weighted portfolio. However, due to difficulty in securing data a value-weighted portfolio could not be constructed for either the CARs or the HPRs. 7

8 its seasoned equity sale, the holding period returns is still compared to the market index for the full holding-period. SECTION 4 RESULTS Table 2 shows the average market-adjusted returns (AR) and the cumulative average market-adjusted returns (CAR) for all 60 event months of the 39 SEOs. 8

9 Figure 1 Cumulative Average Returns Cumulative Average Returns Cumulative Average Return 80.00% 60.00% 40.00% 20.00% 0.00% -20.00% -40.00% -60.00% -80.00% 1 6 11 16 21 26 31 36 41 46 51 56 Months Relative to Date of SEO - Event Cumulative Average Market Adjusted Returns Cumulative Average Raw Returns Cumulative Average Market Returns Figure 1 shows the cumulative average raw SEO returns, the cumulative average market returns and the cumulative average market-adjusted returns. The negative value of the cumulative average market returns reflects the time period chosen for the study. Approximately 72 (29 of 39) per cent of the sample SEOs were before 1987, so the majority of the sample and the corresponding market index went through the October 1987 stock market crash and subsequent recession. These facts account for the negative average returns for the period under study. The other factor that contributes to the poor performance is the inclusion of firms in the sample that were delisted from the New Stock Exchange either through going into liquidation or breaching exchange requirements. The inclusion of these firms helps reduce survivorship bias and contributes substantially to the poor performance of the SEOs. The short-term-positive abnormal return found in several studies (Loderer, Sheehan, and Kadlec (1991), Tripathy and Rao (1992), and Barclay and Litzenberger (1988) was not found to exist in the sample of SEOs. The market-adjusted returns of the first month were negative and insignificant (at the 0.05 level). 9

10 Table 2 Distribution of Five-Year Holding Period Returns following Seasoned Equity Offerings Five-Year Holding Period Returns Rank Seasoned Equity Offerings Market Index Difference Lowest -99.20% -59.84% -39.36% Lower Quartile -92.50% -44.83% -47.67% Median -52.90% -5.80% -47.10% Upper Quartile 52.40% 47.64% 4.76% Highest 437.50% 115.54% 321.96% Mean -6.95% 5.25% -12.20% Standard Deviation 112.30% 54.46% Table 2 shows the distribution of the holding-period returns for SEOs between 1984 and 1994. The average five-year holding period return for the SEOs was 6.95% compared to an average five-year holding-period return for the market index of +5.25%. This can be described as a dollar invested in SEOs at the end of the first trading day to the fiveyear anniversary from 1984 to 1994 would produce a value of $0.9305. Alternatively investing in the market index at the same time would return $1.0525. These results compare to Spiess and Affleck-Graves (1995) and Loughran and Ritter (1995) who report a mean five-year holding-period return for SEOs of +55.72% and +33.4% respectively. The five-year holding-period returns for the SEOs have a far larger distribution than the market five-year returns. The SEOs have a standard deviation of 112.30% compared to 54.46% for the market. The sample contains two particularly large positive returns that influence some results. The large range of the market also reflects the turbulent nature of the share market over the sample period. Initially, the SEO sample did not significantly underperform the market-index, in a statistical sense (t-stat -0.71 at the 0.05 level). Because the distribution of holding-period returns are clearly negatively skewed for the SEO firms additional tests are carried out to check for outliers and their impact. Conducting a simple sign test it is found that the fraction of firms that underperform the market (66.67%) is significantly different than 50% at the 0.05 level. When cleaned of two outliers with large influence, the SEO sample significantly underperformed the market index (t-stat 2.95 at the 0.05 level). Removing these two outliers does not alter the significance of the tests that are conducted later in this paper. Because Conrad and Kaul (1993) have documented a potential upward downward bias induced by calculating short-term abnormal returns over long periods, the focus for the remainder of the paper is on holding-period returns. 10

11 SECTION 5 CROSS-SECTIONAL AND TIME-SERIES PATTERNS IN THE POST-OFFERING PERFORMANCE OF SEASONED EQUITY OFFERING FIRMS To investigate the possible relationships and explanations of the long run underperformance phenomenon various cross-sectional patterns are examined. The long run performance of the SEO firms is segmented by year of issuance, issuer size, issue size, industry, and survivorship. 5.1. Post-Offering Performance by Year of Offering Table 3 Distribution of Five-Year Holding-Period Returns Average Five-Year Holding Period Return Year No of SEOs Issue Size SEOs Market Wealth Relative 1984 5 29,365,707 30.05% 65.51% 78.57% 1985 6 148,260,073 18.54% 7.90% 109.86% 1986 9 171,281,871-85.54% -46.63% 27.09% 1987 9 43,582,752-78.39% -43.13% 37.99% 1989 1 547,204 437.50% 25.03% 429.90% 1990 2 44,688,185 11.24% 66.07% 66.98% 1991 1 140,000,000 199.19% 115.54% 138.80% 1993 4 83,713,572-1.99% 71.34% 57.20% 1994 2 52,051,485 145.83% 39.73% 175.93% 1984 to 1987 29 392,490,403-41.86% -14.92% 68.34% 1989 to 1994 10 321,000,446 94.29% 63.76% 118.64% 1984 to 1994 39 713,490,849-6.95% 5.25% 88.41% In Table 3 the SEOs are categorised by year of issuance. The sample period can be divided into two separate time periods. The first period is 1984 until 1987 and the second period is 1989 to 1994. Notably, 74% of the sample firms issued equity in the first period. In this first period the five-year holding-period returns are significantly lower than the market return for the same period (t-stat -2.1 at the 0.05 level). The second period the returns are higher than the market but this result is not significant (tstat +1.1). In comparing the two periods against each other it was found that the 1984 to 1987 period underperformed the 1989 to 1994 period (t-stat 2.19 at the 0.05 level). In Table 1 (p. 6) a summary of all equity raising is listed. The notable features of this table are the significant use of equity raising in 1986 and 1987. The largest number of seasoned equity offerings exist in these years with 1986 being a highlight in terms of both the number of seasoned equity offerings and the volume of equity raised. The results from Table 3 are supportive of Ritter s (1991) window of opportunity framework where firms take advantage of transitory windows of opportunity by issuing equity when, on average, they are substantially overvalued. 6 It is not surprising that the nine firms from the sample that did not survive the five-year period came from the years 1986 and 1987, reflected with these two years possessing the lowest wealth relatives. The desire to of firms to raise equity becomes greater in hot issue markets (periods of high equity issue volume). 6 Overvaluation is considered both in a relative sense (a firm is overvalued relative to comparable firms at a point in time) and in an absolute sense (a firm is overvalued relative to its past or future market value). 11

12 5.2. Post-Offering Performance categorised by Firm Size Table 4 Post-Offering Performance categorised by Firm Size Market Capitalisation Average Five-Year Holding Period Return Quintiles SEO Market Wealth Relative Significant Results t-stat No. of SEOs Less than 10,000,000-50.88% 7.28% 45.78% -3.1 (5%) 7 10,000,000 to 18,000,000 3.13% 4.27% 98.90% 8 18,000,000 to 58,800,000-64.00% -19.65% 44.81% -2.03 (5%) 8 58,800,000 to 110,000,000 56.91% 8.14% 145.10% 8 Greater than 110,000,000 14.60% 26.46% 90.62% 8 Table 4 shows the performance of firms segmented by firm size. The smallest quintile sized firms show significant underperformance with a wealth relative of 45.78%. There is also significant underperformance in the third quintile. In an unreported result, one of extreme outliers was removed from the second quintile resulting in significant market underperformance. The poorer aftermarket performance of the smallest SEOs suggests there may be a size effect in the issuance of seasoned equity. Spiess and Affleck-Graves (1995) report that underperformance is more severe for the smallest SEOs. The insignificance of the results for the larger firms shows that the large firms do not significantly underperform the market. Brav and Gompers (1997) suggest in the IPO market that size and book-to-market account for the new issue effect. While there is a small amount of evidence suggesting that a size effect may be present here it is more sensible to conclude that it has an influence on how underperformance is distributed rather than a cause of the underperformance. 39 12

13 5.3. Post-Offering Performance categorised by Issue Size Table 5 Post-Offering Performance categorised by Issue Size Gross Proceeds Average Five-Year Holding Period Return Quintiles SEO Market Wealth Relative Significant Results t-stat No. of SEOs Less than 2050000 49.61% -14.43% 174.85% 7 2050000 to 3500000-64.93% -10.59% 39.22% -6.0 (5%) 8 3500000 to 6100000-41.61% 32.52% 44.06% -3.2 (5%) 8 6100000 to 14000000 3.25% -18.49% 126.67% 8 Greater than 14000000 26.01% 34.78% 93.49% 8 Table 6 Post-Offering Performance of Issue Size to Firm Size % Issue Size to Firm Size Average Five-Year Holding Period Return Quintiles SEO Market Wealth Relative Significant Results t-stat 0% to 8% 32.59% -12.12% 150.87% 8 8% to 15.5% 39.12% 40.62% 98.93% 8 15.5% to 23% -65.30% -22.91% 45.01% -4.09 (5%) 8 23% to 32% -46.21% -8.10% 58.53% -2.15 (5%) 8 Greater than 32% -14.71% 32.12% 64.55% -1.7 (10%) 7 Table 5 shows the performance of firms segmented by issue size. The second and third quintiles show significant market underperformance with wealth relatives of 39.22% and 44.06% respectively. The insignificance of the other quintiles suggests that issue size alone does little to help us understand SEO underperformance. Table 6 shows the performance of firms segmented by the size of the issue as a percentage of the size of the firm issuing the stock (before the issue). The results show that firms in the highest three quintiles (with a higher percentage of issue size to firm size) underperform the market significantly showing wealth relatives of 45.01%, 58.53% and 64.55% respectively. This may be viewed to be consistent with the agency theory of Jensen and Meckling (1976). Seasoned equity offerings dilute managerial shareholder ownership and provide cash for management, two motivations for managers to be less interested in shareholder wealth maximisation. 39 No. of SEOs 39 13

14 5.4. Post-Offering Performance categorised by Industry Table 7 Post-Offering Performance categorised by Industry Average Five-Year Holding Period Return Industry Issue Size SEO Market Wealth No. of SEOs Relative $ 4,620,800-60.43% -42.60% 68.93% 2 Agriculture and Associated Services Chemical and Fertiliser 55,385,513-8.28% 14.45% 80.14% 3 Construction 2,985,458-78.41% -5.80% 22.92% 1 Energy and Fuel 58,711,892 121.18% 47.64% 149.82% 1 Engineering 19,456,790-6.78% 11.15% 83.86% 1 Food 1,560,757 30.32% -50.40% 262.75% 1 Finance and banks 46,836,806-65.94% 17.01% 29.11% 2 Forestry and forest products 117,603,954 183.34% 49.23% 189.87% 3 Investment 192,144,739-72.09% -21.39% 35.50% 9 Liquor 12,201,701 90.00% 38.35% 137.33% 1 Media and Communications 10,837,200-46.76% -30.41% 76.51% 2 Medical Supplies 3,981,710-92.50% -43.43% 13.26% 1 Miscellaneous Services 9,031,831 382.63% 114.39% 225.12% 2 Retail 323,481-80.62% -44.77% 35.09% 1 Transport and Tourism 169,272,947 15.66% 53.75% 75.22% 5 Mining 8,535,270-73.00% -12.02% 30.69% 4 39 Table 7 shows the performance of SEO s segmented by industry. The sample represents a well-diversified cross-section of the New market. The investment, transport and tourism, and mining industries were the most active industries over the sample period. Deregulation of the investment and transport sectors were likely causes of the strong representation in the sample. Investment firms performed poorly as they were able to raise money easily in the euphoria of the share market before the crash in 1987 at, as it turned out, grossly overvalued prices. It is difficult to draw any industry generalisations from this data due to there only being, in many cases, one or two firms from the industry represented in the sample. 14

15 5.5. Post Offering Performance categorised by Survivorship Table 8 Post Offering Performance categorised by Survivorship Firm Class No of SEOs SEOs Market Wealth Relative Difference of Means t-stat Surviving Firms 30 15.99% 18.51% 97.87% -0.10 Non-surviving Firms 9-91.15% -45.60% 16.27% -6.51 39 Table 9 Number of Companies that Delist per Year Percentage of Failing Firms on NZSE Year No. of Deletions No. of Companies % Failing Firms 1984 11 237 4.64% 1985 16 254 6.30% 1986 32 273 11.72% 1987 49 288 17.01% 1988 56 234 23.93% 1989 41 195 21.03% 1990 57 139 41.01% 1991 32 112 28.57% 1992 9 112 8.04% 1993 2 127 1.57% 1994 4 139 2.88% Mean all 28 192 15.15% Table 8 shows the performance of firms segmented by whether they survived the duration of the 60-month period without being delisted from the New Stock Exchange. The results are conclusive. The nine firms that do not survive are the reason that the general sample population underperform the market. The t-tests show that the non-surviving firms significantly underperform the market (t-stat 6.51 at the 0.05 level). What is just as important is the fact that the surviving firms do not significantly underperform the market (t-stat 0.1). The two outliers were taken out of the sample with no difference to the results (t-stat for surviving firms 1.2). The question then arises: are the chances of picking a SEO that is going to fail any bigger from picking a failing stock at random? In other words is underperformance merely a function of market survivorship as opposed to an issue effect? Table 9 shows the number of firms on the New Stock Exchange that are delisted each year. Over the period of the sample it is found that, on average, 15.5% of the firms that are listed on the New Stock Exchange are delisted each year. The sample contained nine stocks out of 39 (23%) that were delisted over the ten-year period. If you were to randomly select 39 stocks and measured their returns over the same period, it would be very feasible to select a portfolio that would have the same number or a greater number of stocks that were delisted. It is concluded that SEOs are just as likely to delist as any other stock in the market. Therefore, it suggested that the underperformance of SEOs in New from 1984 to 1994 is merely a function of market survivorship. This means that SEOs should not be considered a special class of investment or a way to lose money, more a result of the greater workings of the survival 15

16 in the market. Essentially, there is no issue effect present. This is contrary to Loughran and Ritter (1997) who report that a new issues effect exists. The survival effect results are linked with a share market crash driven explanation for the underperformance of SEOs. The share market euphoria and turmoil surrounding the crash in 1987 resulted in a large number of SEOs and also a large number of firms failing. The performance of firms that issued in the hot markets of 1986 and 1987 underperformed significantly with a number of them failing after raising capital. It is likely that the firms failed as a result of the share market crash (6 of the 9 failures were 1987 issues), suggesting that the real reason for the underperformance was ultimately the crash of 1987. 5.6. Regression A simple O.L.S. regression was carried out to test for any statistically significant crosssectional relationship between the long-run five-year holding-period returns and several potential explanatory variables. Table 11 summarises the results. Table 10 Regression Results with Five-Year Holding Period Returns as the Dependent Variable Issue Size (t-stat) Firm Size (t-stat) Issue Size to Firm Size (t-stat) Intercept -.2104(-0.97) -0.0924(-0.42) -0.089(-0.42) X variable 7.5956E-09(1.37) 2.5710E-10(0.28) -0.2133(-0.5) R-Squared 4.80% 0.10% 0.70% None of the regressions show a statistically significant relationship with the five-year holding-period returns. SECTION 6 SUMMARY AND CONCLUSION This study has investigated the long run performance of SEOs on the New Stock Exchange from 1984 to 1994. The sample was conducted over a period where New was deregulating a number of industries and the country experienced recession. Questions may remain regarding the generality of the results beyond the sample period and to the general SEO population. It was found that the average fiveyear holding period return for SEOs was 6.95% compared to an average holding period return for the market index of +5.25%. The performance of SEOs are significantly lower than the market. Several explanations for the long-run underperformance were suggested. A size effect was investigated and while there was evidence that it may be present in the results it was concluded that it was an influence on how the underperformance is distributed rather than a cause of the underperformance. The clustering of equity issues around the earlier pre-share-market-crash period supports the widely held belief that firms make use of windows of opportunity and issue stock when their shares are overvalued. The underperformance of firms that issue a large proportion of stock compared to their market value provides some evidence that perhaps agency costs may play a role in 16

17 explaining the underperformance of seasoned equity offerings in New. If increased agency problems were not fully anticipated by the market, the manifestation in low post-issue stock returns suggests agency costs may influence long-run underperformance in the New market. When the sample is broken down into two sub-groups, it is found that firms that survive the five-year post-issue period do not significantly underperform the market. The performance of seasoned equity offerings is a direct result of the failing firms who significantly underperform the market over the period of the study. When the whole New stock market was analysed it was found that the sample firms had just as much chance of failing as firms in the general market. This leads to the conclusion that the performance of SEOs in the period is merely a manifestation of firms failing in the stock market as opposed to an issue effect. Therefore, firms that conduct seasoned equity offerings are no more likely to lose money as any other share selected randomly from the market. The number of failing firms is also consistent with a share market crash explanation for the underperformance of SEOs. It was these failing firms that were the reason for the underperformance, and as the crash was the underlying cause of a number of the failures, the share-market crash theory is another potential explanation. A number of the findings in this paper are contrary to other studies who find that SEOs underperform the market in the long run. The investigation of the long-run performance highlights how different markets can provide different insights into a problem that existed consistently in other markets of the world. The different results suggest that the New market may be unique in its make-up, yet it may also help to provide an explanation to the long-run underperformance anomaly in other world equity markets. FURTHER RESEARCH There are a number of areas that warrant further investigation for the performance of SEOs in the New market. In particular, the further investigation into the reasons for SEO performance needs further investigation. Similar to Spiess and Affleck- Graves (1995) firm characteristics such as size, age and book-to-market equity may be examined to explain performance with more intensive statistical analysis of these crosssectional variables. Further, the pre-and-post-issue operating performance should be examined (Loughran and Ritter (1997)) as a fundamental building block for further investigation. The further investigation of a survival effect in other world markets would be interesting. REFERENCES Barclay, M. J., Litzenberger, R. H., 1988, Announcement Effects of New Equity Issues and the Use of Intraday Data, Journal of Financial Economics 21, 71-99. Bayless, M., Chaplinksy, S., 1996, Is There a Window of Opportunity for Seasoned Equity Issuance?, Journal of Finance 51, 253-278. Brav, A, Gompers, P. A., 1997, Myth or Reality? The Long Run Performance of Initial Public Offerings: Evidence from Venture and Nonventure Capital-Backed Companies, Journal of Finance, 52. 17

18 Cai, J., 1996, The Investment and Operating Performance of Japanese SEO firms, Working Paper, Hong Kong University of Science and Technology. Choe, H., Masulis, and Nanda, V., (1993), Common Stock Offerings Across the Business Cycle: Theory and Evidence, Journal of Empirical Finance 1, 1-33. Conrad, J., Kaul, G., 1993, Long-term Market Overreaction or biases in Computed Returns?, Journal of Finance 48, 39-63. Hansen, R. S., Crutchley, C., 1990, Corporate Earnings and Financing: An Empirical Analysis, Journal of Business 63, 347-371. Healy, P. M., Palepu, K. G., 1990, Earnings and Risk Changes Surrounding Primary Stock Offers, Journal of Accounting Research 28, 25-48. Jensen, M., Meckling, W., (1976), Theory of the Firm: Managerial Behaviour, Agency Costs and Ownership Structure, Journal of Financial Economics 3, 305-360. Loderer, C. F., Sheehan, D. P., Kadlec, B. 1991, The Pricing of Equity Offerings, Journal of Financial Economics, 29, 35-57. Marsh, P., 1979, Equity Rights Issues and the Efficiency of the U.K. Stock Market, Journal of Finance 34, 839-862. Myers, S., Maljuf, N., 1984, Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have, Journal of Financial Economics 13, 187-222. Ritter, J.R., 1991, The Long Run Performance Of Initial Public Offerings, Journal of Finance 46, 3-27. Ritter, J.R., Loughran, T.,1995, The New Issues Puzzle, Journal of Finance 50, 23-51. Ritter, J.R., Loughran, T.,1997, The Operating Performance of Firms Conducting Seasoned Equity Offerings, Journal of Finance 52, 1823-1850. Spiess, K., Affleck-Graves, J., 1995, Underperformance In Long-Run Stock Returns Following Seasoned Equity Offerings Journal of Financial Economics 38, 243-267. Strang, K., 1996, The Long-Run Performance of New Initial Public Offerings, unpublished paper, University of Waikato. Tripathy, N., Rao, R. P., 1992, Adverse Selection, Spread Behaviour, and Over-The- Counter Seasoned Equity Offerings, Journal of Financial Research 15, 39-56. Vos, E.A., Cheung, J., 1993, New IPO Underpricing: A Reputation Based Model, Small Enterprise Research 1, 13-22. 18