Excerpt
TABLE OF CONTENTS
CHAPTER 1 - Abstract
CHAPTER 2 - Introduction
A Growing Trend
Every Action is Part of the Larger System
Preview of the following Sections
CHAPTER 3 - Literature Review
The Winner´s curse Hypothesis
The Lawsuit avoidance Modell
Signaling Theory of Underpricing
Other Models for the Underpricing of IPOs
Long-Term Underperformance after IPO Underpricing
Diversion of Opinion Hypothesis
CHAPTER 4 - Methodology and Results
Data Analysis
Results
Limitations
CHAPTER 5 - Conclusion and Directions for Future Research
References
Abstract
Current scholarship offers a large body of research on the underpricing of IPOs, especially in the U.S. market. In a parallel manner, there is a significant body of research on the fact that a large percentage of stock underperform what should in all likelihood be their market value for years. However, what is in large manner missing from current research is an examination of how these two phenomena are related. Providing further data and insight into this relationship is the primary goal of this thesis. That goal has been met in some measure; however, even more importantly vis-a-vis this goal this thesis demonstrates that this relationship is a much more complicated one than expected and must include a wider range of variables than has been attempted. The second primary goal of this research is to begin to establish the reasons for the above phenomena: Why are some IPOs so underpriced and why is the trend toward underpricing accelerating? Why do some stocks that seem to be very sound nonetheless continue to underperform for at least five years after their IPOs? What, if anything, are the reasons behind the relationship between these two? This thesis puts forth some preliminary answers that provide beginnings of the answers to these questions. All of the research in this thesis is based on data is based on a large sample of companies that have floated their IPOs since 2000 on the New York Stock Exchange.
Introduction
The 19th century had its gold strikes and silver strikes, economic anomalies that rattled the financial worlds of their times, helping to demonstrate what was normal by defining what was not. Economic history is in fact replete with such anomalous events occur around whatever product is valuable within the economy at a given time, whether it be enduring glitter of gold or the far more idiosyncratic allure of tulip bulbs. It is in fact arguable that markets exist to be disrupted, although this is not the most conventional view of how markets work. Markets are generally assumed to seek homeostasis and predictability. This emphasis on the connection between rational behaviour by the different actors in a marketplace and a well-functioning economic sphere has existed for centuries.
Initial public offerings occupy the same niche in today’s market: They can create a whirlwind of intense speculation and giddiness just as a gold strike could. More importantly for the purposes of this thesis, they were seen as aberrant, distorting the normal functioning of the market. The large body of recent research on IPOs (of which there is indeed a very large body) has examined why the dynamics of IPOs have drawn so much attention and the reasons that researchers (and market experts) have defined the behaviour of IPOs as anomalous. In particular, this study will focus on two key aspects of IPOs that have drawn particular critical scrutiny because they are seen as incongruous: Underpricing and long-term underperformance and the connections between these two.
IPOs, which are the focus of this thesis, have drawn significant interest as a subject of research for a number of years by dozens of scholars. One of the reasons for this is that there is a substantial amount of precise data on the behaviour of IPOs, and such large databases are inherently attractive since they present to researchers the possibility of providing precise insights into market dynamics. IPOs, with their precisely documented behaviour, are also tempting as subjects of research because of their (seeming) anomalousness, but even more for the fact that the IPO is inherently important to the lifespan of a company.
Studying the initial behaviour and subsequent behaviour of IPOs may provide much broader insights into the overall health of companies in a predictive fashion. Data of IPOs can be linked to some fascinating long-term results on the behaviour of firms, specifically that they remained underpriced. But how, if at all, are these different sets of economic behaviour connected to each other? Is there something about the nature of certain IPOs themselves in the current market that results in such dramatic poor long-term performance? Do such conditions exist only in certain sectors of the market? If not, are there other defining characteristics of the structure of these IPOs or of the companies behind them that be linked to their long-term underperformance?
A Growing Trend
The questions enumerated above have taken on more urgency recently because the degree to which IPOs are underpriced, suggesting that this discrepancy in offering price to end of first day price will continue to grow and expand to other sectors of the market that have not yet been affected. Given that, as demonstrated by this thesis, there is a clear connection between under- pricing and long-term undervaluation, if underpricing as a phenomenon is becoming more common and more severe, one must consider the probability that there will be growing consequences for the market as a whole because of growing underperformance of more and more stocks.
Over the past fourty years, the average first-day return has measured about 20% (Logue, 1973; McDonald and Fisher, 1991; Ritter, 1984; Ibbotson, 1975; Ibbotson, Ritter, 1991; Sindelar & Ritter, 1988; Ritter and Loughran, 2002; Ritter and Welch, 2002; among innumerous others have examined and documented the general topic of underpricingof IPOs). Data on the underperformance of stocks that were initially underpriced demonstrate that these stocks tend to underperform for at least five years: This trend is documented and analyzed in the later sections of this thesis. This underpricing and underperformance represents an immense amount of profits lost to those companies compared to the behaviour of other publicly traded firms that at least appear to have analogous characteristics (Loughran and Ritter, 1995; Ritter, 1991).
As noted above, models of marketplace dynamics tend to favour explanations that are built on rational behaviour. It is hardly surprising that this model of the functional marketplace should have precedence over other models, for businesses and investors thrive when the market is predictable. When it is impossible to know what tomorrow will bring to either manager or investor, the market is more likely than anything else to collapse into chaos while the economy spins itself out. Given that such market disruptions have significant negative consequences for nearly all stakeholders, being able to understand the mechanisms that contribute to the connections between underpricing and underperformance becomes vital.
However, as this paper examines, the idea of a market that runs without significant and recurring anomalies falls more into the category of the desired-for rather than the truly expected. The data collected and analysed in this thesis demonstrate that anomalies occur at a regular rate, calling into question how anomalies should be defined within market dynamics and thus calling into question larger assumptions about the ways in which markets work.
This paper examines how these models of the marketplace (the one in which anomalies are truly anomalous and not a regular characteristic of the marketplace and the one in which they are in fact a regular characteristic of the marketplace and therefore not truly anomalous) in terms of their relationship to the behaviour of IPOs. As noted, the thesis to be examined here in the context of different models of the marketplace is that stocks are underpriced as IPOs continue to be undervalued for years. This pricing strategy of the IPO should be defined as anomalous in the sense that it does not maximize the possible profits that the company could have expected to bring in, and the failure to maximize profits is by definition anomalous. Thus the degree of deviance (in a purely mathematical sense) of the underpricing of IPOs becomes increasingly deviant as the stock continues to underperform.
This brings us to a pair of interesting paradoxes. First, IPOs that are underpriced continue to underperform for years despite the fact that this runs counter to the most basic economic principle, which is that businesses exist to make money. Second, the marketplace, rather than minimizing the occurrence of anomalous events, has on a tacit level been designed to acknowledge them. To test how the performance of IPOs (both as initial offerings and in the longer term) intersects with larger market dynamics as well as with ongoing conceptions of how markets should work, this thesis examines a large body of data on IPOs offered since the year 2000.
This database allows an examination of the relationship between the underpricing of IPOs and the underperformance of those stocks to two different ends. The first of these is that relationship between IPO underpricing and long-term underperformance has not as yet received sufficient attention: The current research attempts to remedy this under-representation in the current research as regards to the long-term behaviour of IPOs. Indeed, this was the initial question that prompted and guided this research. However, in seeking to understand the precise relationship of underpriced IPOs to longer term undervaluation, a second question came to the fore. This broader question of how the anomalous behaviour that occurs around IPOs and the later behaviour of these stocks fits into a larger question of what constitutes anomalous behaviour in the stock market.
Every Action is Part of the Larger System
While this thesis is focused on the question of the relationship between underpriced IPOs and long-term undervaluation, the theoretical approach used to assess and analyze this relationship takes place within the larger questions posed above. Thus, this thesis takes a somewhat unusual approach in that while focusing on micro economical question, it does so while acknowledging that all economic processes take place within a larger context of general and ongoing market forces.
Among the theoretical models that will be considered here is the efficient market hypothesis, which is based on the specific idea that all stocks are priced accurately for current market conditions. The consequence of this assumption is that an identification of what appear to be abnormal returns cannot be discovered through the process of identifying stocks that are mispriced. In other words, stock prices contain an element of randomness, which makes the concept of anomalous behaviour problematic; however, given that there exists this aspect of randomness in the market, it cannot be ascertained through common methods. This makes good common sense: If the methods used to analyze the marketplace arise from a model of the marketplace as a rational instrument, then random or anomalous behaviour will almost necessarily be inadequate to the task of predicting or analyzing them. This is one of the basic challenges of economic analysis, how to account for both what is predictable and regular and what is not.
Thus it must be noted that there are some detectable market patterns that predict abnormal returns - creating the fascinating but paradoxical case that there are market events that can be predicted through indirect means. One of these is the connection between underpriced IPOs and their continued undervaluation in the market.
While both of the anomalous aspects of IPO companies have been well documented and analysed (even if this analysis has not lead to consensus about the underlying reasons for these two forms of economic behaviour) what has so far been missing from almost all of these examinations of these two IPO issues is any assessment of how the two phenomena are related to each other. Such a lack of investigation as to the interdependence of underpricing and underperformance is puzzling given the amount of data that exist on both phenomena. It is certainly possible that researchers in these areas have decided that there is no cause-and-effect relationship between underpricing and underperformance rather only a correlative or coincidental one.
A number of different explanatory models have been proposed to explain the phenomenon of underpriced IPOs, although not necessarily for the reason that there has been an increasing discrepancy in the behaviour of IPOs in terms of offer price and end-of-first-trading- day price. These models (which can also be used to explain other economic models) including the signalling-based model and the winner’s curse theory. Another explanation (which is not a model per se) is that this phenomenon exists as an idiopathic attribute of IPO performance.
That is, IPOs act in the way that they do because IPOs act the way that they do. This may sound flippant, but it most certainly not intended to be.
Rather, this idiopathic explanation is intended to support an idea explored above, which is that IPOs behave in a way that must be considered to be anomalous and adhering only to the case of IPO behaviour. While this last is less satisfying in many ways than being able to fit the behaviour of IPOs into a regular pattern that can be correlated with the behaviour of other sectors in the market, it is imperative for good research to acknowledge that any dynamic or phenomenon may be subject to a unique explanation.
While this may seemingly violate the requirement that explanations be as parsimonious as possible, in fact it is in accord with this requirement, for parsimony demands that we seek for the simplest possible explanation that is consistent with all of the facts. Because none of the established models is able to provide a sufficiently accurate explanation, other explanations must be put forward and tested to see if they are more accurate until there is an explanation or model that answers at least some of the fundamentally important questions that remain about the behaviour of IPO underpricing (Aggarwal and Rivoli, 1990 and Ritter, 1991).
To create a model or explanation this thesis incorporates a large body of recent data. These data are plugged into current methods of explanation to demonstrate the inadequacy of such models, although the models are also compared as to their efficacy. That is, while the current models are insufficient to answer certain key questions about the underpricing of IPOs, some models are more accurate and efficacious than are others. Bearing this in mind, examining the models that are more effective should prove to be an effective strategy of creating a new model. There is no reason to recreate the wheel in this case. While current models are not perfect, they offer significant insights into the workings of markets and so should be examined.
The most important aspect of current models to be considered in this thesis is whether any of the current models provide an explanation (in whole or part) of the ways in which underpricing and long-term under-performance are related to each other.
Preview of Following Sections
Before beginning a review of the literature on these topics, it is important to reiterate once again the importance of this topic in terms of the behaviour of the markets in real time. If it is indeed the case that there is a causal link between underpricing of IPOs and long-term underperformance, this fact could have highly significant implications for the ways in which IPOs are structured and the extent to which companies are willing to tolerate underpriced IPOs. IPOs are an essential aspect of the current marketplace in the United States and there is no current indication that this shall change in any way in the near future. Given that underpricing has been increasing in terms of rate and amplitude recently, it seems likely in fact that the overall market will be affected to an ever-greater degree by this phenomenon.
Without the capital that firms can raise through IPOs, these companies could not in most cases survive. Without these companies, in other words, there would develop a marketplace that is more monopolistic because it allows for fewer chances for new firms to enter into established economic sectors. This would have consequences so large in amplitude that it is almost impossible to imagine what the American marketplace would look like without a substantial reliance on IPOs, especially given the fact that the marketplace as it is currently structured developed around the presence of IPOs in an organic way and any change to the ways in which IPOs are floated would necessarily have ripple effects.
As noted above, no legitimate analysis of the ways in which IPO underpricing and long- term underperformance can be made in the absence of an examination of the data. Thus, the first task of this thesis has been to create a database that is sufficiently large and appropriately targeted to allow for the types of analysis needed. This seemingly simple task is in fact rather more complicated than it would appear to be given that while databases that would be appropriate to use for this thesis in terms of content and reliability are proprietary, and thus prohibitively expensive.
After having gathered the data independently by the process of merging several different publicly available databases, these data were analysed using a model, more explicitly using (market corrected) returns. Such a relatively simple methodological strategy proved to be appropriate since the calculations themselves are not the most difficult task at hand: These in many ways are less onerous and complicated than the gathering of the data or the development of the thesis in terms of its theoretical orientation. The result of the analysis on the data collected and examined in some ways supports previous models along with a general (although untested and to some extent merely conventional belief) that underpriced IPOs are linked to long-term (up to five years) underperformance of the stock for these companies. However, and this will of course be detailed in the results section, this model only holds true for companies in certain sectors of the marketplace.
This is one of the most important findings of this thesis: There is certainly a relationship between underpricing and underperformance but at least as far as this group of stocks over this time period is concerned obtains only for some stocks and not for others.
The above touches on both of the key goals of this thesis. The first of these is to document whether or not such a relationship (between underpricing and underperformance) exists; that is, to critically review the conventional wisdom that it does and, moreover, that it is not a dynamic of past markets but continues to affect the dynamics of the recent and current marketplace.
The second, which is nearly always the more interesting aspect of any piece of research, is to propose explanations as to why this should be the case. Implicit in this second question is the assumption that the relationship between the two is causal rather than coincidental, an assumption supported by the fact that the relationship exists not in a randomly distributed way but in terms of market sectors, suggesting that there is an underlying reason.
CHAPTER 2: LITERATURE REVIEW
Given how important IPOs are within the current market structure of the United States and given how fond economists are of creating their own models that are consonant with their own philosophies of how markets do and should work, it should hardly be surprising that there are numerous different models of the reasons why IPOs are so often underpriced. The first section of this literature review summarizes and assesses some of the most important (and most commonly used) theories and models used to explain the ways in which IPOs come to be underpriced as well as the reasons that stocks may suffer from long-term underperformance, often to a significant degree.
Winner’s-Curse Hypothesis
One of the most respected of these models is Rock’s (1986) 'winner-curse hypothesis'. This model is used to describe the dynamics of auctions in which there is incomplete information on one side of the auction participant. (This is also more generally referred to as an asymmetric situation.) Such asymmetry is common, of course; indeed it is arguable that there is never a case in which the information is precisely symmetrical. The ‘curse’ referred to in this model is that the winner of such auctions is in many cases in the position of having overpaid for the stock. They may have made the mistake of placing a bid that turns out to be higher than the value of the stock or, alternatively, she may be cursed because the asset proves to be lower in value than the winner thought that it would be. (If the information is not evenly distributed between the stakeholders, it should be easy to understand how such a ‘curse’ could be laid upon a winning bidder.)
While this thesis focuses on the United States, the phenomenon of the underpriced IPO exists worldwide, as the following list documents:
Country: Size, Period, Negative Return
Australia: 266 1976-89 11.9%
Austria: 67 1964-96 6.5%
Belgium: 28 1984-90 10.1 %
Brazil: 62 1979-90 78.5%
Canada: 258 1971-92 5.4%
Chile: 19 1982-90 16.3%
China: 226 1990-96 388.0%
Denmark: 32 1989-97 7.7%
Finland: 85 1984-92 9.6%
France: 187 1983-92 4.2%
Germany: 170 1978-92 10.9%
Greece: 79 1987-91 48.5%
Hong Kong: 334 1980-96 15.9%
India: 98 1992-93 35.3%
Israel: 28 1993-94 4.5%
Italy: 75 1985-91 27.1%
Japan: 975 1970-96 24.0%
Korea: 347 1980-90 78.1%
Malaysia: 132 1980-91 80.3%
Mexico: 37 1987-90 33.0%
Netherlands: 72 1982-91 7.2%
New Zealand: 149 1979-91 28.8%
Norway: 68 1984-96 12.5%
Portugal: 62 1986-87 54.4%
Singapore: 128 1973-92 31.4%
Spain: 71 1985-90 35.0%
Sweden: 251 1980-94 34.1%
Switzerland: 42 1983-89 35.8%
Taiwan: 168 1971-90 45.0%
Thailand: 32 1988-89 58.1%
Turkey: 138 1990-95 13.6%
United Kingdom: 2,133 1959-90 12.0%
United States: 13,308 1960-96 15.8%
(Ritter, 1998, p. 5)
The fact that underpricing occurs across so many different markets in very different economies suggests that there is something intrinsic to the IPO process that exists independent of other market dynamics and local cultural, economic, and political conditions.
Importance of Asymmetry
Michael Spence, the Nobel Prize winner for the Prize in Economics in 2000, was one of the first scholars to investigate the ways in which uneven distribution of information has the most profound consequences within market systems. His very low-key summation of his understanding of the consequences of unequal information follows, demonstrating how widely such a dynamic exists in the market:
“We noticed that there are many markets with informational gaps. These include most consumer durables, virtually all job markets, many financial markets, markets for various types of food and pharmaceuticals, and many more.
These informational gaps were widely acknowledged and those of us who taught applied microeconomic theory, freely admitted that these gaps might change some of the performance characteristics, not to mention the institutional structure, of markets.” (Spence, 2002).
It can be argued that the winner’s curse only occurs if the winner of the bid is unaware of this potential dynamic; knowledge of this market error can be avoided through foresight about the dynamic.
Such a case of forewarned-is-forearmed is especially likely to occur during the IPO floating process since many, indeed most (indeed sometimes nearly all) of the investors are not skilled either in market dynamics in general or in the particular strengths and vulnerabilities of the stock in question. Investors in IPOs may well not be inclined to invest under other conditions: Because IPOs are seen as an attractive form of investment they will attract people who are novices to the process.
Asymmetry of information results in uninformed buyers being far more uninformed than buyers in other circumstances and so much more subject to the winner’s curse.
[...]