SPECIAL ARTICLE Do Stock Markets Allocate Resources Efficiently? An Examination of Initial Public Offerings
Vineet Kohli
This paper examines the pricing of Initial Public Offerings in relation to their future operating performance and risk. IPO firms have lower profitability but receive higher valuation than their industry peers on the expectation that their earnings will grow in the future. The expectation of superior growth is not realised in the post-issue period. It thus appears that low profitability firms conduct IPOS when investors are excessively optimistic about their growth potential. The paper concludes that stock markets in India have suffered from excessive optimism and poor evaluation.
I am grateful to Prabhat Patnaik and C P Chandrasekhar for very helpful comments and discussions. The usual caveat applies.
Vineet Kohli (vineetjnu@gmail.com) is with the Tata Institute of Social Sciences, Mumbai.
Speculators may do no harm as bubbles on a steady stream of enter
prise. But the position is serious when enterprise becomes the bubble
on a whirlpool of speculation. When the capital development of a
country becomes a by-product of the activities of a casino, the job is
likely to be ill-done.
–Keynes (1936, p 159)
T
The research on pricing and operating performance of IPOs also has policy relevance. It is sometimes argued that stock markets behave efficiently by providing correct market valuation of companies. For two companies in the same size and risk class, the one with higher expected earnings will receive a higher stock price and will find it more profitable to issue equity. The more profitable firm will thus grow bigger than the less profitable firm and its relative importance in the economy will increase. This view of efficient allocation of resources through stock markets is often contrasted with inefficient allocation of resources in a government dominated bank-based system. Shah (1999), for example, argues that public sector banks face incentive problems and bank officials are prone to external pressure. On the other hand, individuals trading in the stock market have full incentive to evaluate the firm correctly since they stand to lose their wealth if they their assessment proves wrong. Also, stock market investors trade anonymously and are not prone to external pressure. The argument that bank-based systems allocate resources to unprofitable projects became extremely popular after the east Asian c risis. It was argued that a select group of favoured large corporations could corner a large share of credit from banking systems in these countries without any regard for the underlying profitability of their investments. Arguing that the east Asian crisis was an outcome of the structural weaknesses of their bank-based s ystems, Pomerleano (1998) said:
This relationship suggests lax credit allocation processes (possibly supporting projects of politically connected individuals and organisations), without reference to project viability.
Since development of stock markets is logically premised on a superior allocation of resources when funds are routed through them, it is a legitimate exercise to evaluate how stock markets a llocate financial resources. The role of resource allocation by
Economic & Political Weekly
EPW
stock markets is often done even before market valuation is a rrived. This happens when firms raise money through IPOs. The extent to which capital markets encourage efficient allocation of resources thus depends to some extent on how IPOs are priced. If investors fail to evaluate the true profitability and/or risk of firms at the IPO stage, serious doubts can be raised about the efficiency of resource allocation through stock markets. This paper will, however, not only examine the rationality of pricing at the IPO stage but will also delve at some depth into evolution of prices in the post-IPO period when shares of the company become available for trading in the stock market.
The paper is organised as follows. Section 1 discusses two different views on the functioning of stock markets. These two theoretically contrasting views will form the basis for interpretation of empirical results. In Section 2, we discuss the objective of this study and the choice of sample, variables and statistical methodo logy used. In Section 3, the results are stated. Section 4 clearly states the central question arising from the data exercise. S ections 5 and 6 discuss the results. In Section 7, the impact of primary market regulation is investigated, while Section 8 concludes the paper.
1 Two Views on the Functioning of Stock Markets
The current state of understanding about how real world stock markets function, may broadly be classified into two schools. The first, the asymmetric information school, emphasises the role of asymmetric information in producing sub-optimal outcomes but retains the assumption of rational investor behaviour. The s econd, the behavioural school, accords centrality to the irrational behaviour of investors.1
Since the main element of difference between these two views rests in the assumption of investor rationality, a brief explanation of the concept of investor rationality seems warranted. An investor is said to be rational if he/she values every asset at its fundamental value. The fundamental value of a share is equal to the present discounted value of rationally expected earnings per share. The assumption of rationality ensures that investors quickly process available information to obtain the probability distribution of outcomes (in this case regarding earnings of the firm) that will actually prevail. This has two implications. First, rationality implies that outcomes cannot deviate from expectations. Second, since available information is readily incorporated in the formation of prices, future changes in prices cannot be predicted on the basis of available information. This basically means that prices will follow a random walk.
In recent years, a large literature taking clues from the Chapter 12 of Keynes’ General Theory has highlighted how markets may deviate from this benchmark of rationality in various ways.2 As opposed to the view of rationality in which investors know the outcomes in advance, Keynesian view emphasises the role of i gnorance in forming expectations. According to the Keynesian view, investments produce returns in a distant future of which investors only have a vague idea. The ignorance of investors (and the necessity of acting in spite of this ignorance) forces them to rely on conventions in forming expectations. One such convention, the role of which Keynes himself emphasised, was the use of the recent past as a guide to the future. According to Keynes, investors often
64 extrapolate the current into the future since they know little about the actual character of future. All in all, once we enter the Keynesian world of ignorance, the deviation of share prices from fundamental values becomes possible because investors have no idea of what the fundamental value of the asset is. In the Keynesian world, it is perfectly possible for outcomes to (systematically) deviate from expectations and for share prices to be predicted on available information (even when it gets stale) since ignorant i nvestors do not know in what direction and to what extent is the new information to be incorporated.
Let us now discuss the asymmetric information view in detail.3 According to this view, while managers know the fundamental value of assets of the firm and its investment opportunities, outside investors can only observe the probability distribution of these fundamental values. Therefore, the market price of each firm is equal to the mean of fundamental values of all firms.4 This implies that higher than average fundamental value firms will be underpriced by the market, whereas lower than average fundamental value firms will be overpriced.5 It is further posited that managers maximise the wealth of existing shareholders and both new and existing shareholders know this to be the objective function of the firm. If managers always work in the interest of existing shareholders, an overpriced firm will always issue equity even if the best use it can make of these funds is to park them in a zero NPV investment. Suppose a company has 10 outstanding shares, each carrying a fundamental value of Rs 10. If due to asymmetric information, each share is valued above its fundamental value (say) at Rs 20, issuing equity and parking the proceeds in a zero NPV investment will cause the wealth of existing shareholders to increase. If only one share is issued and the proceeds are invested in a zero NPV investment, the fundamental value of 11 shares becomes Rs 120 and thus the fundamental value of each share increases above Rs 10. On the other hand, an underpriced firm will issue equity only if the loss due to issuing equity is compensated by the NPV of investment. In the previous example, if each share had a market price of Rs 5, the firm would issue 2 shares to buy an investment project of Rs 10 only if this investment project had NPV of at least Rs 10, for otherwise, the post-issue fundamental value of 12 shares will fall below Rs 120 and existing shareholders will have to suffer a loss.
Having laid down the structure of the problem in this manner, it is not difficult to see that the decision to issue equity conveys a negative signal about the value of the firm. Since investors know that all non-issuers must be underpriced firms, those firms that issue equity must have a higher probability of being overpriced. This implies that rational investors will reduce the market price of any firm that seeks equity finance in the presence of asymmetric information. Moreover, since managers know that their firm will be devalued on the announcement of an equity issue, they avoid issuing equity. This view thus predicts that firms should sparingly rely on equity as a mode of financing investment. These models have found some validation in the empirical literature. Asquith and Mullins (1986) have shown that the stock price of a firm falls on the announcement of an equity issue.
august 15, 2009 vol xliv no 33
It must be noted here that costs of issuing equity due to asymmetric information arise because investors behave rationally. It is rational that if the firm is issuing equity, it is likely to be overpriced, which causes investors to devalue the firm on the announcement of the equity issue. An equity issue acts as a mechanism that allows investors to screen lower value firms from higher value firms and then price firms according to their fundamental values.
We can now move to the behavioural view of the functioning of the stock market. However, to maintain continuity with the argument so far, we begin with some empirical studies on the long-term price performance of firms that raise equity finance.6 These studies show that return on equity issuers (both IPOs and seasoned equity offerings, SEOs) remain below that of matched non-issuing firms from almost six months after the date of issue to five years after it.7 These results have been interpreted as the correction in pricing of issuers at the time of issue. The argument goes that issuers are overpriced at the time of issue. However, as investors realise with experience that the true value of the firm is lower than what they had thought at the time of issue, the share price of the firm falls and its return remains below the return on matched non-issuers. This shows that investors do not behave in a fully rational manner at the time of issue. If investors behaved in a fully rational manner, the information content of the equity issue decision should have been readily incorporated. There should have been a one-shot reduction in the value of the firm after the equity issue. After that the return on issuer should remain the same as the return on the matched non-issuer.
The overpricing of equity issuers at the time of issue and their subsequently lower returns warrants an explanation. A particularly persuasive explanation is in terms of the fall in operating performance of equity issuing firms from their pre-issue levels. Jain and Kini (1994) find that operating performance of IPO firms dips after the issue. They document that in the post-IPO period, both the operating return to assets and operating cash flow to asset ratios are significantly lower when compared to the year just preceding the IPO. Moreover they find that these ratios are higher for IPO firms when compared to non-IPO firms in the year preceding the IPO but become comparable in the post-issue period. They also find that both the market to book and price to earnings ratios calculated at offer price register a significant dip after the year of IPO. To unite these findings, the Keynesian understanding of how expectations are formed comes in handy. Before the IPO, firms have a good earnings performance. Investors extrapolate good performance into the future and overvalue the firm. After the IPO, earnings performance dips and investors revalue the firm downwards. A similar decline in operating performance in the post-issue period was documented for SEOs by Loughran and Ritter (1997). An interesting argument to explain the dip in postissue operating performance is made by Teoh et al (1998). They argue that firms actively manage their earnings before conducting an IPO. To the extent that earnings are artificially inflated before the IPO, they are likely to fall in the post-IPO period. All in all, these studies imply that investors do not behave in a rational manner. They fail to foresee that earnings performance will dip after the IPO while rewarding these firms with high share prices.
Economic & Political Weekly
EPW
We can now summarise the two views on how financial resources are allocated through stock markets. In the asymmetric information view, stock markets play only a limited role in providing finance because investors perceive equity issues as attempted rip-off by low quality firms and respond by imposing costs on issuing firms. In the behavioural view, firms are able to sell overpriced shares to optimistic investors. In this view, investors do not fully comprehend the information content of an equity issue and overreact to past earnings performance of IPO firms by rewarding them with high prices. However, actual earnings performance in the post-IPO period turns out to be worse than expected and investors revalue the firm downwards. Thus, in the behavioural view, stock markets do play their role of providing equity finance to firms but they play this role inefficiently by allowing low-value firms to foist their overpriced shares on investors.
2 Empirical Investigation
The objective of the following empirical investigation is to ascertain whether IPOs in India are priced efficiently. Specifically, the objective is to check whether prices are related to earnings and risk profile of firms. If pricing is unrelated to earnings and risk and reflects some kind of optimism on behalf of the investors, resources in the stock market-based system may not necessarily flow to their most profitable use. Some eminent financial commentators have drawn attention to overpricing of new equity issues in India.8 The purpose of this empirical investigation is to ascertain whether IPOs are overpriced in India. If they are indeed overpriced, what are the reasons behind this overpricing? This paper argues that there is something structural in the way expectations are formed that causes IPOs to be overpriced.
2.1 Choice of Sample
To begin with, a small clarification is necessary on the use of the term IPO in this paper. The term IPO in the following is taken to mean public issues of equity, including issues even by firms that are already listed. Although, the second public issue of equity by firms in the PROWESS database have been excluded, some firms may have conducted a public issue even before entering the PROW-ESS database. The number of such firms has not been ascertained.
IPOs in India exhibit a high degree of industrial concentration. The industry-wise distribution of IPOs in the Indian economy u sing the National Industrial Classification (NIC) at the two digit level shows a high degree of concentration within a few industries. Almost one-fifth of all the 3,480 issues between 1990 and 2006 have taken place in financial and leasing services (Figure 1, p 66). Outside the financial and leasing services, largest number of issues has taken place in chemicals (largely due to issues by drugs and pharmaceuticals companies) and second largest number of issues has taken place in the computer industry. However, b etween 1994 and 2006 (the period of this study), the share of financial and leasing services goes up to 23.9%; the share of chemicals falls slightly to 12.9%, whereas that of computer software goes up to 8.6%. However, these industries remain in the top three in terms of number of issues in our sample period. As is the usual practice in studies on IPOs, the financial services industry has been excluded. For the purpose of this study, only IPOs in the chemicals and computer software industries have been included. Together, 499 IPOs in these two industries account for more than one-fourth of IPOs in the non-financial sector in the sample period, namely, April 1994 to March 2006.
Figure 1: Industrial Composition of IPOs in India (% of Total, April 1990 to March 2006)
Chemicals (13.19)
Others (40.66)
Computer
Software (7.30)
Textiles (6.84)
Financial and Leasing Services (19.20)
Food, Beverages
Trade (6.21)
Source: Information generated from the CMIE PROWESS database.
2.2 Choice of Variables
The objective of this study is to examine whether IPOs are priced efficiently, that is, to examine whether the initial price of these IPOs is justified by their future operating performance. The stock price variable used for the purpose of this study is the priceearnings (PE) ratio. An IPO may command a higher PE ratio when compared to its industry peers if it is expected to grow at a higher rate compared to these peers and/or it carries lower risk. Earnings of IPO firms may be expected to grow either due to expectation of improvements in profitability or due to superior growth of assets. An important point that needs to be underscored here is that higher assets growth will translate into higher earnings growth from the point of view of original investors only if this growth is financed internally or through debt (although debt financing is considered risky). If growth is financed through further issues of equity, higher assets growth of the firm will translate into higher earnings growth from the point of view of original investors (who purchased shares in the IPO) only if investment is made in a more profitable project, for otherwise the growth in earnings will be offset by dilution of shareholding of the original investors.
To check whether IPO firms are overpriced or not, information on the following variables has been generated from the CMIE PROWESS database: Stock Price Variable
1 PE ratio. Operating Performance Variables
2 Profit after tax (PAT) as a percentage of total assets.
3 Profit before tax (PBT) as a percentage of total assets.
4 Capital expenditure (CAPEX), defined as a percentage change in the sum of fixed assets and inventories over their previous year v alues.
5 Earnings growth (EG), defined as percentage change in PAT over its previous year values.9 Risk Variable
6 Beta from the capital asset pricing model.
66
For each IPO firm, information for all five operating variables has been generated from the year prior to the IPO to four years after the IPO. Since the PE ratio of IPO firms is not available before the offer date, only five years of information on the PE ratio is used. The PE ratio at the offer date is calculated by dividing the offer price by earnings per share (EPS) in the previous year. EPS is calculated as PAT of previous year divided by shares outstanding after the IPO. Since PE ratios cannot be negative, they could only be calculated for firms with positive earnings. Information has been generated on PE ratios from the first to the fourth year after the issue.
The measure of risk used by this study is beta. Beta is a measure of risk since it tells us how the return on IPO stocks co-varies with the return on the market portfolio. If the return on IPO stocks is not correlated with the return on the market portfolio, the stock will have a low beta and it will carry lower risk since adding such a stock to portfolio will reduce the variance of the portfolio. Since beta is calculated by regressing returns on the stock against the returns on the market portfolio, one should a llow sufficient time to elapse after the IPO (so that there are sufficient data points to carry this regression). Beta figures are thus taken for two years after the issue. Beta is, however, a controversial measure of risk. Beta tells us how variance on the overall portfolio (in this case the CMIE index) changes when we add a particular stock to the portfolio, but most investors do not hold a diversified portfolio. For example, Indian Household Investors’ Survey-2004 showed that most retail investors (who are the domi nant participants in the new issue market) hold a portfolio of 3 to 10 stocks. Investors’ own assessment of risk from holding a stock may thus deviate from its beta.
2.3 Benchmark Non-issuers and Choice of Sample Period
Profitability, growth, risk and PE ratios vary across industries. For example, typically a computer software firm will trade at much higher price earnings multiple when compared to, say, a firm in the iron and steel industry. The reason is that the computer software industry has a higher expected growth when compared to the iron and steel industry. Thus, even though a software firm may have lower current earnings, its earnings may end up growing at such a rapid rate that its higher PE ratio today is rationally justified. Thus, in this study there is a need to benchmark variables of IPO firms with corresponding variables of non-issuing firms in the same industry.
The sample of non-issuing firms in each of the issue years is defined as those firms that have not carried out a public issue of equity in the year of the issue and any of the four years preceding the issue. Since comparison between IPO and non-IPO firms is made up to four years after the issue, IPO firms have not been a llowed to enter the sample of non-issuers for these four years. Also, literature has shown that stock prices of IPO/SEO firms u nder-perform their benchmarks for three years after the issue. Thus, including firms that have conducted IPOs in this horizon in the sample of non-issuers will contaminate PE ratio comparisons between issuers and non-issuers.
This is also the reason for starting the sample in 1994 when data are available in the PROWESS database for earlier years as
august 15, 2009 vol xliv no 33
well. Most of the IPOs in the Indian economy took place in the period between 1990 and 1995. Now, given the condition that IPO firms can enter the sample of non-issuers only four years after the issue, the size of the non-issuers’ sample progressively decreases as we move backwards in the sample period. In the year 1994, the non-issuers’ sample for the computer software industry contains only 10 listed firms. Since, the industry effect can be controlled with a reasonable degree of confidence only when there are sufficient firms in the sample of non-issuers, 1994 has been decided as a cut-off year.
2.4 Statistics
For each of the variables in the sample, median values have been calculated. The reason for focusing on median rather than mean is that mean is often affected by extreme values. Also, PE ratio has a minimum value of zero and is thus skewed to the left.10 M edian is thus a better indicator of central tendency of such a variable. Both raw and industry-adjusted median values have been reported. To calculate industry-adjusted median values, the corresponding medians of the
Table 1: PE Ratio of IPO Firms
whether there is a significant change in the post-issue values of various variables when compared to their pre-IPO levels. Throughout this paper, the convention of using one-sided sign test and two-sided signed rank test has been followed.
3 Results
The results will be discussed in this section. The results of PE ratio, profitability, growth rates and risk will be discussed in that order.
3.1 PE Ratio
Table 1 reveals that IPO firms have a median PE ratio of 28.57 at offer price. The median PE ratio however falls after an IPO and reaches a low of 6.49 two years after the IPO. There is a slight r ecovery in the third and fourth year after the IPO. However, even in the fourth year, the median PE ratio remains at less than half of its value on the offer date. The Z-statistic, obtained through Wilcoxon signed rank test between PE ratios in period t=0 and PE ratios in subsequent years, is significant at 1% level for all four years. This suggests that PE ratios of IPO firms fall below their offer date levels. However, the
non-issuers’ sample have been fall in the PE ratio of IPO firms
Stats t=0 t=1 t=2 t=3 t=4
subtracted from values (of all may simply be an outcome of a
N 210 275 212 178 203
firms) for each of the variables. fall in the PE ratios of indus-
Median (raw values) 28.575 8.03 6.49 9.52 13.16 Thus, from the PAT ratio of IPO Z 8.999*** 6.04*** 6.399*** 3.971*** tries of which these IPOs are a
firms in the year of issue, Median (industry adjusted) 13.256*** 0.51 -1.205** -1.1175 0.94 part. Therefore, the industry
m edian PAT ratio of non-issuers Ż 5.334*** 3.132*** 4.098*** 2.979*** adjusted medians of PE ratios
(1) Triple, double and single stars refer to 10%, 5% and 1% levels of significance respectively.
in the same industry in that have also been looked at.
(2) t refers to years since IPO. For example, t=0 refers to the year of IPO.
year have been subtracted. Adjusted PAT ratios in the year of issue for each of the years between 1994 and 2004 are then placed adjacent to each other in the same column. The median value of this column is the adjusted PAT r atio of issuers in the year of issue. The same exercise is then repeated for the year prior to the issue as well as each of the four years a fter the issue.
Since quarterly reporting in India started after 1998, data on operating variables is only available annually between 1994 and 1998. Thus operating variables of issuers are adjusted by median values of non-issuers in the same year. However, PE ratios are available at a higher frequency (since prices change even if earnings remain constant at their previous year levels) therefore PE ratio comparisons are made on a quarterly basis. Thus, PE ratios of IPO firms at offer date are compared with PE ratios of non-issuers in the same quarter in which IPO took place. Subsequent comparisons of PE ratios are made in the last quarter of the financial year. To take an example, for an IPO that took place in the first quarter of the financial year 1994-95, comparisons are made with the median PE ratio of non-issuers (appropriately defined for the year 1994-95) in the first quarter of 1994-95. Subsequent comparisons are made in last quarters of 1996, 1997, 1998 and 1999.
Two statistical tests have been used in this study. The first is the sign test which tells us whether industry adjusted median v alues are significantly different from zero or, what is the same, whether the median value of IPO firms is significantly different from that of non-issuers in the same industry. The second important test is the signed rank test which shows how different variables evolve from their IPO and pre-IPO levels. This test will reveal
A look at industry-adjusted v alues reveals that IPO firms have a much higher median PE ratio at the time of issue. The difference between the PE ratio of IPO firms and non-issuers is 13.25. The sign test reveals that this value is significantly different from 0 at 1% level. However, the industryadjusted medians also fall over time. The fall in PE ratios is quite drastic and, by the second year after the IPO, the PE ratio of IPO firms falls significantly below the PE ratio of industry peers. The Ż-statistic obtained through Wilcoxon signed rank test is also s ignificant at 1% level. The fall in the PE ratio of IPO firms thus cannot entirely be attributed to a fall in the PE ratio of benchmark non-issuers. Table 1 also shows that the PE ratio of non-issuers also falls in the post-IPO period. If we treat the difference between entries in the third and fifth rows as median values of non-issuers, we find that even these values register a decline in the post-IPO period. The median PE ratio of non-issuers thus falls from 15.32 at the offer date to around 7.7 two years after the IPO. Table 1 thus reveals three important facts about IPOs in India:
3.2 Profitability
Two variables of profitability, namely, profit before tax as a percentage of total assets and profits after tax as a percentage of t otal assets, have been used in this study. Cash flows have not been e mployed, as is usually the practice in firm-level performance
Economic & Political Weekly
EPW
studies, because cash flow statements are not available for most firms in the early years of our sample period.
The PAT ratio of IPO firms is the highest in the year prior to an IPO (Table 2). After that, there is quite a free fall and the PAT ratio falls to 0.5% by the end of the fourth year after the IPO. Here again, the decline is significant at 1% level.
Table 2: PAT Ratio of IPO Firms
Stats t=-1 t=0 t=1 t=2 t=3 t=4
N 323 402 401 391 404 385
Median (raw values) 3.87 2.735 1.68 0.34 0.34 0.5
Z 5.039*** 6.355*** 9.253*** 9.827*** 8.792***
Median (industry adjusted) -1.705*** -2.522*** -1.77*** -2.46*** -2.1725***-1.72***
Ż 3.269** 2.265** 5.09*** 6.426*** 5.27***
The industry-adjusted median PAT ratio is negative, suggesting that IPO firms have significantly lower profitability when c ompared to benchmark non-issuers. The reason may be that the companies with higher profita-
Figure 2: Box Plots of Industry-Adjusted PAT Ratios of IPO Firms in t=-1 and t=4
bility generate sufficient re-comparison is made between
100
PAT (t=-1)
sources to finance their growth PAT (t=4) t=-1 and t=0. So, while we may internally. However, lower
0
profitability firms do not gener
the profitability of IPO firms ate enough resources internally -100
d eclines after an issue, we may and since the amount of debt
be less confident that they rethat a firm can assume is a -200
main permanently depressed function of the profits it can
after the IPO.
-300
generate internally, these firms
The main conclusion of this must also find it difficult to fi
sub-section is that IPO firms
-400
nance their growth through bank loans (Kalecki 1937). The only way such firms can grow is through an equity issue. This may be the reason behind the lower PAT r atio of IPO firms.
The decline in profitability in the post-IPO period can be understood in two ways. First, there could be a decline in profitability due to agency problems caused by greater diffusion in share ownership after the IPO (Jensen and Meckling 1976). In a manager-owned firm, managers bear the full cost of any perquisites they consume on the job; whereas when the shareholding of the firm becomes diffuse, managers derive the full benefit of these perquisites but the cost is shared by a wider population due to dispersed shareholding.11 Thus, managers may have a greater tendency to indulge in wasteful consumption of perquisites in the post-IPO period and this may cause the decline in profitability in the post-IPO period as well. Second, profitability may decline b ecause firms may have indulged in earnings management before the IPO. To the extent that earnings were managed before the IPO, they are likely to fall after it. This was, for example, found by Teoh et al (1998) who showed that firms use accruals to boost earnings before an IPO.
The industry-adjusted PAT ratio of IPO firms also registers a decline in the post-issue period. The fall is significant at 1% level if we look at Ż-statistic. An interesting feature is that Ż-statistics remain significant even when medians remain more or less s imilar. For example, the industry-adjusted median of PAT ratios is -1.705 in t=-1 and -1.72 in t=4. The reason for decline in p rofitability, even when medians remain more or less unchanged is that values above and below the median may have fallen s ignificantly over time. This emerges clearly from the comparison of box plots of PAT ratios in the year prior to an IPO and four years after the IPO. While, the median remains the same in both box plots, the lower quartile is smaller for PAT ratios at t=4. Also, both positive and negative outliers have much smaller values for t=4.
Moreover, as in the case of PE ratios, there is some decline in the PAT ratio of non-issuers as well. In t=-1, the PAT ratio of non-issuers is 5.575. By the fourth year, the PAT ratio of non-issuers is 2.22. Thus, there is a decline in the PAT ratio of non-issuers as well.
The PBT ratios also register a decline in the post-IPO period relative to pre-IPO values (see Table 3). Moreover, the PBT ratio of IPO firms is significantly smaller than industry benchmarks, confirming the earlier result that IPO firms have a significantly lower profitability when compared to their industry benchmarks. The industry-adjusted PBT ratios also register a decline but this
d ecline is significant only when
have a lower profitability than non-issuing firms. Profitability is much lower even before IPOs are conducted. After the IPO, the profitability of IPO firms declines. There is a certain decline in raw values. Industry-adjusted values fall significantly in the case of the PAT ratio. The decline in adjusted values is, however, not significant in the case of the PBT ratio. Our central point from the perspective of this study is that IPO firms have a significantly lower profitability and thus difference in profitability (which either falls or at best remains stagnant in the post-IPO period) cannot explain why IPOs have higher PE ratios than non-IPOs at the time of issue.
Table 3: PBT Ratio of IPO Firms
Stats t=-1 t=0 t=1 t=2 t=3 t=4
N 323 402 401 391 404 385
Median (raw values) 10.6 6.21 7.39 5.635 5.175 6.63
Z 6.226*** 4.478*** 5.917*** 5.951*** 5.083***
Median (industry adjusted) -5.26*** -8.24*** -4.8*** -6.095*** -5.16*** -4.66***
Ż 3.183*** 0.026 1.3 1.62 -0.12
3.3 Growth
The two indicators of growth we have used have an interesting story to tell. A look at Table 4 (p 69) shows that IPO firms are growing at a breakneck speed. In the year of IPO, (raw) median capital expenditure of these firms is 97%. This means that IPO firms almost double their size in the year of IPO. However, even before the year of IPO these firms have an impressive growth of
august 15, 2009 vol xliv no 33
around 54%. In Tables 4 and 5, besides the Z-statistic, we have also calculated the Z1-statistic. The Z1-statistic compares growth rates in year 0 with growth rates in subsequent years. The growth figures for many IPO firms in the sample are not available in the year prior to the IPO. Hence comparisons are also made between year 0 and subsequent years. The Z-statistic reveals that the growth rate does not fall in the year of IPO, suggesting that firms use funds raised in the IPO to acquire real assets. However, the capital expenditure of these firms tapers off subsequently and both Z and Z1 values are significant. Similar results obtain when we look at industry-adjusted capital expenditures. IPO firms have much higher capital expenditures compared to benchmark nonissuers in t=-1 and t=0. However, overtime, the gap between these two sets of firms closes down and by the third year, the capital expenditure of IPO firms becomes indistinguishable from that of the non-issuing firms. An interesting feature is that capital expenditures of non-issuing firms also decline from their IPO and pre-IPO levels. The capex of non-issuing firms in t=-1 is 9.165%. By t=4 this figure has fallen to 4.613%. However the decline is not comparable to the decline in capital expenditure of IPOs, which falls from 53.75% in t=-1 to 5.095% in t=4.
Table 4: Capital Expenditure of IPO Firms
Stats t=-1 t=0 t=1 t=2 t=3 t=4
N 101 301 370 365 364 354
Median (raw values) 53.75 97.26 34.4 11.245 6.14 5.095
Z -0.352 4.334*** 5.823*** 5.874*** 6.803***
Z10.033*** 11.647*** 11.796*** 11.719***
1
Median (industryadjusted) 44.585*** 84.19*** 21.5925*** 2.9*** 0.23 0.4825
Ż -0.079 4.626*** 5.42*** 5.147*** 6.197***
Ż1 9.576*** 10.98*** 11.157*** 11.059***
Table 5: Earnings Growth of IPO Firms
Stats t=-1 t=0 t=1 t=2 t=3 t=4
N 125 233 247 196 179 188
Median (raw values) 68.15 63.96 14.71 -7.14 9.62 27.505
Z 4.06*** 5.781*** 6.03*** 4.837*** 3.459***
Z1 5.856*** 7.853*** 5.077*** 3.275***
Median (industryadjusted) 44.25*** 32.45*** -16.21** -36.342*** -23.12*** -20.52**
Ż 4.408*** 6.446*** 5.935*** 5.143*** 4.012***
Ż1 8.309*** 7.678*** 5.147*** 4.07***
The earnings growth figures in Table 5 reveal a picture more or less similar to the one in Table 4. However, industry-adjusted earnings growth registers a far steeper decline when compared to capital expenditure. While IPO firms start with earnings growth figures that are significantly above benchmark non-issuers, they end up having much lower earnings growth after one year of an IPO. For example, the industry-adjusted median of earnings growth is 44% in t=-1 and falls by 80% to -36% in t=2.
This sub-section shows that a significant reversal in the growth performance of IPO firms takes place in the post-IPO periods. In the year before issue, IPO firms have higher earnings as well as assets growth. In the post-issue period, assets growth is similar whereas earnings growth is lower for issuing firms.
Economic & Political Weekly
EPW
3.4 Beta
Since we are only interested in comparing the risk of investing in equity issuers compared to the risk of investing in non-issuers (and not how the risk itself evolves), betas have been compared for two years after an IPO. The raw median value of beta of IPO firms is 1.14 (Table 6). A value of beta greater than 1 is usually associated with a risky stock, but high beta values of IPO firms may be an outcome of high industry betas. Stocks in some industries such as computer software may fluctuate much more than the market index, causing such firms to have high betas. Indeed, the last row shows that much of the risk of investing in IPO firms is explained by their industry membership. Moreover, industry-adjusted values are not signifi-Table 6: Beta of IPO Firms
| cantly different from 0, sug- | Stats | t=2 |
|---|---|---|
| gesting that high valuations | N | 233 |
| of IPO firms cannot be ex- | Median (raw values) | 1.14 |
Median (industry adjusted) -0.015
plained by their lower risk.
4 Central Question
The above data exercise shows that IPO firms receive higher PE ratios. Higher PE ratios however do not seem rationally justified since IPO firms have lower profitability when compared to nonissuers to begin with and if at all this gap between profitability of IPO and non-IPO firms widens after IPO. Moreover, asset growth of IPO firms is similar and earnings growth is in fact much smaller. Higher valuation of IPO firms in spite of their lower quality contradicts the assumption of investor rationality.
The results do not support the adverse selection hypothesis. Lower profitability and efficiency of IPO firms coupled with similar risk and subsequently similar growth rates suggests that the basic intuition of the adverse selection theories that lower quality firms will try to sell overpriced equity is right. However, the second aspect of these theories that investor rationality will prevent low quality firms from selling overpriced equity is not borne by data. The finding that equity issuers charge much higher prices than their established counterparts at the offer date suggests that investors do not fully understand the information content of e quity issues. Moreover, the evidence of investor rationality is further called into question when it is acknowledged that the price of IPO firms typically jumps on the day of listing.12 If IPOs are already overvalued at their offer price, the (market-adjusted) positive return earned by them on the day of listing cannot be explained rationally.13
This evidence is thus more compatible with some form of investor irrationality. An examination of the source of investor irrationality needs to be carried out. But before doing so, let us reiterate the central question that arises from the data exercise. The central paradox arising from the data exercise is the higher PE ratio of IPO firms at the offer date but their poorer subsequent earnings growth (and similar risk) when compared to nonissuers. Why do investors subscribe to overpriced IPOs? Over longer periods of time after the IPO, investors seem to realise their mistake and revalue IPOs downwards causing PE ratios to fall. But why they make these mistakes in valuation in the first instance is intriguing.
69
5 Towards a Tentative Explanation
We have already argued that a firm with low current earnings will receive a higher share price if its earnings are expected to grow at a superior rate than the firm with higher current earnings. Given the fact that IPO firms have lower current earnings, the reason behind their higher valuations must rest in expectation of higher growth of these firms. This expectation is of course never met since the earnings growth of IPO firms turns out to be lower than that of non-issuers. The basis for this expectation may rest in the past growth performance of the IPO firms which is far superior to that of the non-IPO firms. While there are limited observations for growth figures in the year prior to IPO (101 in the case of capital expenditure and 125 in the case of earnings growth), strong results in Tables 4 and 5 tempt us to put forth this explanation of IPO overpricing in the Indian context. Given the superior earnings and assets growth performance of IPO firms before going public, investors may feel that this superior growth performance is permanent and thus price these IPO firms above their listed counterparts. However after the IPO, investors realise that superior growth performance before the IPO was in fact transitory and thus the PE ratio of IPO firms falls. To put the same argument differently, investors extrapolate the past growth p erformance of IPO firms into a distant future and price them above industry counterparts. However after an IPO, growth figures tend to decline and as actual growth figures fall below the levels expected at the time of the IPO, investors revise their growth expectations downwards and this causes the PE ratio of IPO firms to fall. Our explanation is based on the Keynesian a rgument that uncertainty forces investors to rely on certain rules of thumb such as extrapolating past data to arrive at v aluations of assets. In his summary of the General Theory, K eynes (1937) argued:
We assume that the present is a much more serviceable guide to the future than a candid examination of past experience would show it to have been hitherto.
The reliance on past growth rates to value shares is common in the financial industry. For example, a commonly used method of valuing equity shares is the PEG ratio. The PEG ratio is obtained by deflating the PE ratio by projected earnings growth. The usual practice is to treat firms with a lower PEG ratio as cheaper. Since the projected growth rate is a function of past growth, IPO firms often appear cheaper (even at higher PE ratios) compared to their industry peers due to their better past growth record. The fall in growth rates of these firms in the post-IPO period causes their PEG ratios (calculated at past valuations) to increase above their industry counterparts. Investors respond by dumping IPO firms and buying their low PEG industry counterparts. Thus, the dip in growth performance of IPO firms causes its price to fall and that of its counterparts to rise. This is reflected in the lower industry adjusted PE ratios of IPO firms in the post-issue period.
6 Similarities with Existing Literature
These results bear close resemblance to the arguments of Baker and Wurgler (2007). According to Baker and Wurgler (2007), small, low profitability and higher growth firms are most prone to be affected by waves of investor sentiment. This is because of
INDIGENOUS SYSTEMS OF MEDICINE
April 18, 2009
| Medicine, State and Society | –V Sujatha, Leena Abraham |
|---|---|
| ‘Commercialising Traditional Medicine’: Ayurvedic Manufacturing in Kerala | –M S Harilal |
| Can Maternity Services Open Up to the Indigenous Traditions of Midwifery? | –Mira Sadgopal |
| Recovering from Psychosocial Traumas: The Place of Dargahs in Maharashtra | –Bhargavi V Davar, Madhura Lohokare |
| Medicine as Culture: Indigenous Medicine in Cosmopolitan Mumbai | –Leena Abraham |
| The Patient as a Knower: Principle and Practice in Siddha Medicine | –V Sujatha |
| For copies write to: Circulation Manager, Economic and Political Weekly, 320-321, A to Z Industrial Estate, Ganpatrao Kadam Marg, Lower Parel, Mumbai 400 013. email: circulation@epw.in |
august 15, 2009 vol xliv no 33
the inherent subjectivity involved in valuing such firms. E laborating on this, they argued:
We believe that the crucial characteristic is the difficulty and subjectivity of determining their true values. For instance, in the case of a young, currently unprofitable but potentially extremely profitable growth firm, the combination of no earnings history and a highly uncertain future allows investors to defend valuations ranging from much too low to much too high, as befits their prevailing sentiment. During a bubble, when the propensity to speculate is high, investment bankers can join the chorus arguing for high valuations. By contrast, the value of a firm with a long earnings history, tangible assets, and stable dividends is much less subjective, and thus its stock is likely to be less sensitive to sentiment.
According to this view, in periods of high optimism investors disregard current earnings and dividends of firms and mainly focus on their growth performance. This allows low profitability, high growth firms to time their equity issue in such periods. If these issues are timed at the peak of such bubbles of optimism, subsequent waning of this optimism and concomitant shift in i nvestment strategy away from low profitability to high profitability firms will cause valuation of (low profitability) IPO firms to fall relative to that of higher profitability firms.
issues thus turn out to be a money losing proposition for investors who purchase them.
These results are also similar to those of Rajakumar (2005) who showed that equity financed firms are also liquidity constrained, suggesting that capital market imperfections are more acute for equity financed firms than for debt financed firms. He also showed that after the initial spurt in the early years of economic reforms, the investment tempo of equity financed firms suffered much more than that of debt financed firms. He thus concluded that the overall shift towards equity financing in the first half of the 1990s has acted as a drag on investment.
7 Regulation
It may be the case that these results are an outcome of lack of regulation in the primary market. It is often argued that the excesses of the early 1990s may never be repeated due to stronger regulation of the stock markets. Let us just summarise the argument briefly. Till 1996, there was virtually no restriction on firms raising equity through the public issue route. After 1996, only firms with a history of dividend payments were allowed to enter the primary market. In 1999, this stipulation was diluted to ability to pay dividends rather than actual payment of divi-
Table 7: Pricing and Operating Performance Variables of IPO Firms in the Post-Regulation Period
dends. Thus, after 1996 only firms with a minimum prof
(industry-adjusted values)
itability could issue equity. In the data used in this study,
t=-1 t=0 t=1 t=2 t=3 t=4 Z
| PE | 5.7350* | -3.055 | .045 | -.195 | .38 | profitability is likely to be higher for firms that have con- | ||
|---|---|---|---|---|---|---|---|---|
| PAT | .93 | -.5425 | .4575 | -.795** | -.925* | -1.12 | 2.36** | ducted an IPO after 1996. Thus, even if firms conducting |
| PBT | -2.54 | -4.1** | .7 | -3.72*** | -4.6** | -3.92** | 0.603 | IPOs in the post-regulation period were as aggressive in |
| CAPEX | 63.28*** | 77.125*** | 15.15*** | 1.235 | .9275 | .765 | 3.595*** | pricing IPOs as firms in the pre-regulation period, the |
| PATGROWTH | -4.96 | 32.365 | -17.165 | -10.797 | -33.66*** | -13.075 | 0.202 | higher profitability of post-regulation IPOs may make |
Evidence of overpricing of IPOs is also consistent with the ideas developed by Edward Miller. According to Miller (1977), uncertainty regarding returns from the asset produces divergence of opinion regarding expected returns from holding such an asset. Moreover, greater is the uncertainty, greater is the likelihood of expected returns across investors being divergent.14 Since most of the IPO firms are new and do not have any established track record of performance, there is tremendous uncertainty and thus also great divergence of opinion across investors regarding prospects of these firms. It thus follows that, at the time of the IPO, the shares of the firm can be sold at high prices to a small number of optimists. Overtime, as information about the IPO firm increases, divergence of opinion falls and optimists have lesser reason to remain optimistic about the firm and consequently they revise their valuation of the IPO firm towards the average opinion.
These findings are also similar to those of Purnanandam and Swaminathan (2001) who showed that IPOs typically receive higher valuations than their matched industry counterparts. They also showed that IPOs had a higher price even when a nalysts’ growth forecast is taken into account, suggesting that investors are more optimistic than analysts, who are known to be excessively optimistic themselves and are known to support high and extreme valuations during bubble years. The decline in the PE ratio of IPO firms is also consistent with evidence presented in the beginning of this paper that return on IPOs and SEOs is dismal when compared to matched non-issuers.15 New
Economic & Political Weekly
EPW
their higher offer date prices look justified. To compare pre- and post-regulation periods, the data of 106 IPOs (in the sample) that took place after the financial year 1996 have been separately analysed. The relevant figures are provided in Table 7.
It emerges from Table 7 that firms conducting IPOs have lower profitability compared to non-issuers even in the post-regulation period. Median PAT ratio is lower than that of industry counterparts in most of the years and median PBT ratio is lower in all years. However, the difference between issuers and non-issuers in the post-regulation period seems to have fallen. In the postregulation period also, the PAT ratio registers a decline from its pre-IPO level. The Z value of signed rank test of PAT ratios between t=-1 and t=4 yields a figure of 2.36, which is statistically significant. The PBT ratio does not decline but remains significantly below non-issuers.
The PE ratio of issuers at the offer date seems to remain above non-issuers but the extent of overpricing seems to have fallen in the post-regulation period. This may not have anything to do with regulation since primary market regulation does not seek to influence offer price.16 This is mainly because all years in the pre-regulation period were what could be called, hot IPO periods. Investors were extremely optimistic in this period and a large number of IPOs could be floated successfully in these years. However, after 1996 there were periods of extremely low investor sentiment and hence firms were more conservative in pricing IPOs to prevent issues from remaining undersubscribed. If we only look at high sentiment years within the post-regulation period, IPOs do not seem to be any less overpriced. For example, there is data on 18 IPOs in the year 2000 was shown that investors neglect the low profitability aspect of and these 18 IPOs have an industry adjusted PE ratio of 19.29. new issues and overemphasise their growth performance. The
Finally, even after regulation, indicators of growth present a expectation of growth at the time of issue is not met subsequently. story that is not different from pre-regulation years. Investment These conclusions have some important implications for the deand earnings growth are high in the year of issue but dip after the sign of the financial system. It is often argued that a market-based IPO. The results are not significant in the case of earnings growth financial system should be promoted since a bank-based system but they are also less reliable in the case of earnings growth. For (especially the one with large presence of public sector banks) example, all but 25 companies have common data on earnings does not allocate resources to their profitable use. The large nongrowth in the year t=0 and t=4. In any case, to justify their performing assets of public sector banks are often cited as evihigher PE ratios at the time of issue, firms undertaking IPOs must dence of poor evaluation in the bank-based system. It was seen in grow at a higher rate than non-issuers. The fact that the rates of this paper that equity markets in India have mainly financed growth are lower in the post-IPO period confirms that IPOs are lower profitability firms on the expectation that their earnings overpriced even in the post-regulation period. will grow in future. However, this expectation has not been
r ealised subsequently as both profitability and growth of issuing 8 Conclusions firms have fallen in the post-issue period. This suggests that The main conclusion of this paper is that high prices of new equity stock markets in India have suffered from excessive optimism
issues are not related to their future operating performance. It and poor evaluation.
Notes
1 The view that all investors are rational and no asymmetric information exists is too inane to deserve a mention.
2 This literature appears under the broad banner of behavioural finance. See Shliefer (2000) for a presentation of some of the most important propositions of the behavioural finance school.
3 Discussion is based on papers by Myers (1984) and Myers and Majluf (1984).
4 Suppose there are only two firms: a lower value firm L and a higher value firm H. L has a fundamental value of VL and H has a fundamental value of VH. Due to asymmetric information, investors cannot differentiate between L and H, and attach a value of (VL+VH)/2 to each.
5 This is known as the adverse selection problem in equity markets.
6 For evidence on long-term stock performance of IPOs and SEOs, see Ritter (1991), Loughran and Ritter (1995) and Speiss and Affleck-Graves (1995).
7 Issuing firms are matched by non-issuing firms in the same risk class, so no difference in returns should be expected to exist between these two sets of firms.
8 For example in a recent posting – Watch Out! They Are Out to Get You – on 27 August 2008 on her web site, Sucheta Dalal a well-known financial commentator argued: “The recent cycle of IPOs that culminated in Reliance Power’s massively overvalued Rs 11,700 crore issue is nothing new” and later in the same paragraph: “A little analysis of the past would have told them that a rush of IPOs happens only when things look extremely rosy. It is then that investment bankers collude with c ompanies to stick expensive stocks on to you” (emphasis added).
9 Although, as we shall see, what matters from the point of view of investors purchasing shares in an IPO is the growth of earnings per share.
10 Although such “skewness” can be removed through appropriate transformations such as taking logs or square root of the original data. We can then use mean values of the transformed data.
11 Alternatively, we may argue that in managerowned firms, managers derive full benefits of r esources they expend on managing the firm. However, after an IPO, the benefits are distributed equally across shareholders whereas cost is borne entirely by the owner-manager.
12 Ghosh (2004) calculates the return earned by IPOs in India between offering and listing date for the period 1991 to 2001. This return was the lowest in the year 1996, when an average IPO
earned a (market-adjusted) return of 38.43%.
13 Some finance theorists have argued that IPOs are in fact undervalued at their offer price. This u nderpricing is corrected when market forces are allowed to operate after listing. It is argued that firms often try to signal their quality through such underpricing. The loss due to underpricing is made up through a further issue of equity at far more favourable terms. However, if IPOs are overvalued at their offer price, the return earned on listing cannot be explained through wilful underpricing to signal firm quality. Rather, it is more likely that the market behaves in an irrational manner by raising the price of already overvalued IPOs. For a review of both signalling and irrational explanations of the return on listing, see Ritter and Welch (2002).
14 Divergence of opinion requires uncertainty in the fundamental sense and not just risk in the sense of a well-defined probability distribution of returns. For example, consider a toss of a fair coin, so that we receive Rs 100 if the outcome is heads and pay Rs 100 otherwise. That our expected payoff in this case is 0 is hardly a controversial statement and will not be subject to d ivergence of opinion. However, in choices i nvolving true uncertainty, where there is no
o bjective basis of forming an expectation of r eturns, it is likely that investors will arrive at different expectations.
15 See Ritter (1991), Loughran and Ritter (1995) and Speiss and Affleck-Graves (1995).
16 Regulators have sought to influence IPO pricing in India through the introduction of a book building mechanism in 2000. Book building is a mechanism of discovering the demand curve of shares at the time of an IPO. However, according to the CMIE PROWESS database, there are only 50 IPOs between financial year 2001 and 2004. Moreover, not all IPOs in this period were book-built; therefore we have not been able to check whether book-built issues are any less overpriced.
References
Asquith, P and D Mullins (1986): “Equity Issues and Offering Dilution”, Journal of Financial Economics, 15:61-89.
Baker, M and J Wurgler (2007): “Investor Sentiment in the Stock Market”, Journal of Economic Perspectives, 21:129-151.
Ghosh, S (2004): “Boom and Slump Periods in the I ndian IPO Market”, RBI Occasional Papers.
Jain, B and O Kini (1994): “The Post-Issue Operating Performance of IPO Firms”, Journal of Finance, 49:1699-1726.
Jensen, M C and W Meckling (1976): “Theory of the Firm: Managerial Behaviour, Agency Costs and Ownership Structure”, Journal of Financial E conomics, 3:306-360.
Kalecki, M (1937): “The Principle of Increasing Risk”, Economica, 440-47.
Keynes, J M (1936): The General Theory of Employment, Interest and Money (London: Macmillan).
– (1937): “The General Theory of Employment”, Quarterly Journal of Economics, 209-23.
Loughran, T and J Ritter (1995): “The New Issues Puzzle”, Journal of Finance, 50:23-51.
– (1997): “The Operating Performance of Firms Conducting Seasoned Equity Offerings”, Journal of Finance, 52:1823-850.
Miller, E (1977): “Risk, Uncertainty, and Divergence of Opinion”, Journal of Finance, 32:1151-168.
Myers, S (1984): “The Capital Structure Puzzle”, Journal of Finance, 39:575-92.
Myers, S and N Majluf (1984): “Corporate Financing and Investment Decisions When Firms Have Information that Investors Do Not Have”, Journal of Financial Economics, 13:187-221.
Pomerleano, M (1998): “The East Asian Crisis and Corporate Finances: The Untold Micro Story”, World Bank Policy Research Working Paper Number 1990.
Purnanandam, A and B Swaminathan (2001): “Are IPOs Underpriced?” Working Paper, Cornell University.
Rajakumar, D (2005): “Corporate Financing and Investment Behaviour in India”, Economic & Political Weekly, September 17:4159-165.
Ritter, J (1991): “The Long-Run Performance of Initial Public Offerings”, Journal of Finance, 46:3-27
Ritter, J and I Welch (2002): “A Review of IPO Activity, Pricing, and Allocations”, Journal of Finance, 57:1795-1828.
Shah, A (1999): “Institutional Change on India’s C apital Markets”, Economic & Political Weekly, January: 183-94.
Shleifer, A (2000): Inefficient Markets: An Introduction to Behavioural Finance (New York: Oxford U niversity Press).
Society for Capital Market Research and Development (2005): Indian Household Investors Survey – 2004: The Changing Market Environment, Investors’ Preferences, Problems, Policy Issues.
Spiess, D K and J Affleck-Graves (1995): “Underperformance in Long-Run Stock Returns Following Seasoned Equity Offerings”, Journal of Financial Economics, 38:243-67.
Teoh, S, I Welch and T Wong (1998): “Earnings Management and the Long-Run Market Performance of Initial Public Offerings”, Journal of Finance, 53:1935-74.
august 15, 2009 vol xliv no 33
