Memoire Online - Investor sentiment and short run IPO anomaly: a behavioral explanation of underpricing

Abstract ..............................................................................................6

INTRODUCTION .....................................................................................7

Section 1- Short run IPO anomaly and traditional explanations ......................10

Introduction .............................................................................................10

I- Underpricing anomaly: a persistent phenomenon that characterizes IPO market .....11

I-1\ Underpricing definition ......................................................................11

I-2\ A persistent anomaly in time ..................................................................12

I-3\ A persistent anomaly in all countries .........................................................13

I-4\ A persistent anomaly in all industries .........................................................15

II- Theoretical explanations of short run underpricing: A literature review .............15

II-1\ Asymmetric information ..................................................................16

II-1-1\ The issuer is more informed than the investors .......................................16

II-1-2\ The investors are the most informed .........................................................19

* Information Revelation Theories ..................................................................19

* Winner's Curse .....................................................................................22

* Agency conflicts ....................................................................................23

II-2\ Symmetric information ...........................................................................25

II-2-1\ Risk premium ....................................................................................25

II-2-2\ Characteristics of the Initial Public offering ................................................26

* Risk ......................................................................................................26

* Issue size .............................................................................................27

* Bargaining power ....................................................................................28

II-2-3\ Lawsuit avoidance: legal liability .........................................................29

II-2-4\ Underpricing as a substitute of marketing expenditures ..............................30

II-2-5\ Internet Bubble ...........................................................................31

II-2-6\ Price stabilization and partial adjustment ................................................32

Section 2- Behavioral explanations .........................................................34

Introduction .............................................................................................34

I- Definitions .............................................................................................35

I-1\The sentiment's notion ...........................................................................36

I-2\ Hot IPO market's phenomenon ..................................................................37

I-3\ Investors typology ...........................................................................40

II- Literature review of behavioral explanations ................................................42

II-1\ Informational cascades ...........................................................................43

II-2\ The prospect theory ...........................................................................43

II-3\ Investor sentiment by Ljungqvist, Nanda and Singh (2004) ..............................45

II-4\ The use of Grey Market Data ..................................................................47

II-5\ The use of market conditions to value investor's sentiment ..............................49

II-6\ Discount on closed-end funds as proxy for investor sentiment ......................51

II-7\ Other proxies and empirical results .........................................................52

Section 3- The model and empirical implications ................................................55

Introduction .............................................................................................55

I- The model and explanatory variables .........................................................57

I-1\ The model ....................................................................................57

I-2\ The explanatory variables ..................................................................58

I-2-1\ Informational Asymmetry Theory .........................................................58

I-2-2\ Theory asserting informational symmetry and IPO market efficiency .............60

I-2-3\ Investor sentiment and Behavioral approach ................................................61

II- Data Description ....................................................................................62

III- Empirical implications and analysis .........................................................70

CONCLUSION ....................................................................................77

References .............................................................................................82

List of tables

Table 1. Average first day returns of some studies ................................................14

Table 2. Descriptive Statistics of the Full Sample ................................................65

Table 3. Descriptive Statistics of Firms with Low Underpricing ..............................67

Table 4. Descriptive Statistics of Firms with High Underpricing ..............................68

Table 5. Average First-day Returns Categorized by Underwriter prestige, Share Overhang, R&D Intensity, VC Backing, Age, Assets, Sales, Firm profitability, Industry, Bargaining Power, Individual Investors' sentiment and Time ................................................69

Table 6. Ordinary Least Squares Regression Results (with Bullish proportion as investors' sentiment measure) ....................................................................................71

Table 7. Ordinary Least Squares Regression Results (with Bullish Bearish Spread as investors' sentiment measure) ..................................................................72

Table 8. Ordinary Least Squares Regression Results .......................................76

Abstract

nderpricing phenomenon has intrigued academics and practitioners over the past three decades, and has generated considerable research trying to clarify and to understand this short run puzzle: asymmetric information theories, IPO market efficiency theories and behavioral and sentiment approach. In this study, I regroup in the same model the most important explanations advanced earlier to determine which of these explanations characterizes best the data in the context of a unified framework, with a contribution in the behavioral approach. I use a direct measure of investors' sentiment obtained from the survey data of AAII and II, and I distinguish between the sentiments of the two types of investors: individual and institutional investors.

Sentiment is a primary driver of underpricing and a relevant explanation to this anomaly. Moreover, individual investors are those driving the first day closing prices and are more conducting the short run IPO puzzle than the institutional investors.

INTRODUCTION

oing public constitutes a real driver for the development of a company, enabling it to increase its equity capital and to overcome the constraint that its founders are no longer able to provide the capital needed for its expansion, and enabling it to diversify its sources of financing without the need for debt. It is also a way to provide liquidity by creating an opportunity to convert all or some founders and shareholders' wealth into cash immediately or at a future date. Going public is also an opportunity to clarify the company's business and strategy and to think about the company's future growth. Not forgetting the role of going public and being listed in enhancing the company's credibility and strengthening its image and reputation. When a company decides to go public, shares offered in the stock market should be correctly and truly valued. Issuers should also time the Initial Public Offering to coincide with a favourable period (Hot market) to succeed their first introduction in the market. But the question is, if the issuing firm's managers are shrewd enough to value the company's shares and to choose a hot market period, why underpricing is a persistent anomaly characterizing the IPO market and why so much money is leaving on the IPO table?

Stoll and Curley (1970), Logue (1973), Reilly (1973) and Ibbotson (1975), are the first who documented that when companies go public, the first day closing price is systematically higher than the issue price at which the public offering was introduced in the market. They are the first who documented the first day underpricing phenomenon. And Ibbotson (1975) is the first who offered a list of possible explanations for underpricing, many of which were formally explored by other authors in later work.

The underpricing phenomenon has inspired a large theoretical literature over decades trying to give a relevant and a convincing explanation to this first day anomaly. It has intrigued academics and practitioners over the past three decades and has generated considerable research trying to clarify and to understand this phenomenon. The first explanations that were advanced are based on the informational asymmetry between the key parties of an IPO: issuing firm, underwriter and investors. These theories have been very popular among academics and practitioners for decades and have been considered as the most relevant and convincing explanation to the short run IPO anomaly. However, other theories have been introduced asserting the informational transparency and lucidity and the IPO market efficiency, since the first theories based on asymmetric information are unlikely to give a relevant explanation to the surprisingly and severe level of underpricing of 63.5% reached in 1999 and 2000. This severe level of the internet boom years exceeded any level previously seen. Researches are continuing, and it is fair to say that this anomaly is not satisfactorily resolved. It is still a puzzle, since nor the informational asymmetry theories neither the IPO market efficiency theories are likely to give a convincing and a reliable explanation to this short run IPO phenomenon.

Many researchers come to the conclusion that IPO future researches should turn to behavioral approach and sentiment notion. Research effort should focus more on behavioral explanations to clarify and to explain the short run IPO behaviour and the underpricing anomaly: a new path trying to understand and to explain this persistent anomaly. Turning to behavioral explanations seems to be the most promising area of research to understand the short run IPO behaviour.

We can say that the underpricing anomaly may be the most controversial area of IPO research. Research effort has provided numerous analytical advances and empirical insights, but underpricing is not yet explained and the research effort is continuing.

Ø What are the most relevant and reliable explanations to the underpricing anomaly from the long list of explanations advanced?

Ø If we rely on behavioral explanations based on investors' sentiment to explain this phenomenon, what type of investors is more conducting the underpricing phenomenon and the first day returns?

To answer these two main questions, I present in a first section the phenomenon and its persistence in time, for all the countries and for all the industries. I present the two main categories of explanations: informational asymmetry and theories asserting the informational transparency and the IPO market efficiency. I summarize the most important findings and results of researchers that have considered these theories in their studies. In a second section, I present the most important studies and papers that have considered the sentiment explanation and the investors' behaviour as the most convincing and relevant driver and determinant of IPO underpricing. In this section, I give an idea about the application of the behavioral and sentiment approach to clarify and to understand the IPO patterns by numerous researchers and I shed light on the importance of the sentiment investors in the IPO market.

Finally, in the third section, I regroup the most important explanations that have been advanced in the same model to determine which of these explanations characterizes best a sample of 217 U.S IPOs for 2006 and 2007. In the context of a unified framework and model, I present the three theories: asymmetric, symmetric and behavioral approach.

For the behavioral approach, I use direct measures of sentiment obtained from the survey data of American Association of Individual Investors and Investors Intelligence. I distinguish between the sentiments of the two types of investors: individual and institutional investors, to answer the second question.

The main result is that sentiment is a primary driver of underpricing and a relevant explanation to this short run anomaly. Moreover, individual investors are those driving the first day closing prices and are more conducting the short run IPO puzzle than the institutional investors.

Section 1- Short run IPO anomaly and traditional explanations

The underpricing is a short run anomaly characterizing the IPO market. This phenomenon has inspired a large theoretical literature over decades trying to give a relevant and a convincing explanation to this first day phenomenon. Underpricing anomaly has intrigued academics and practitioners over the past three decades and has generated considerable research aimed at explaining the apparent incongruities with rational asset pricing. While this research effort has provided numerous analytical advances and empirical insights and a large list of explanations were presented, it is fair to say that this anomaly is not satisfactorily resolved. It is still a puzzle sparking much academic attention until now and requiring other explanations and much considerable research effort.

In this first section, I begin in the first paragraph by a definition of the underpricing anomaly and an illustration of its persistence over the time, all over the world and for all the industries, an order to have a complete idea about this phenomenon. I summarizing in a second paragraph the most important results and findings reached by the researchers that have been interested in this field and have been interested in explaining the short run IPO puzzle. These findings can be classified in two main categories:

Ø Explanations based on informational asymmetry between the key parties which have been considered the most convincing explanations for decades by a large number of researchers.

Ø And theories asserting the informational transparency and lucidity and asserting the IPO market efficiency.

I- Underpricing anomaly: a persistent phenomenon that characterizes IPO market

I-1\ Underpricing definition:

Before going in details, presenting the explanations advanced for the short run IPO anomaly, it is logical to begin by a definition of this notion of «underpricing» to understand this short run phenomenon:

Early writers that have been interested in IPO market, notably Stoll and Curley (1970), Logue (1973), Reilly (1973) and Ibbotson (1975), are the first who documented that when companies go public, the price of shares they sell tends to jump substantially on the first day of trading. The first day closing price is systematically higher than the issue price at which the public offering was introduced in the market. Consequently, IPOs exhibit positive first day returns on average with no exception to the industry to which the IPO belongs, to the country and to the period and date of going public. That has been the first and the most important observation in the IPO market.

From an issuer's point of view, this phenomenon is usually called underpricing, as it describes the additional amount of money which could have been raised by the issuer if the offer price had been set at an appropriate level. Indeed, if the issuer had been set a higher offer price for his offering shares, the issue price could have been easily accepted by the investors interested in purchasing the IPO shares, since a run up is observed on the first day of trading. The trading begins by the offer price which is less than the apparent maximum price achievable at the end of the first trading day. In a way it is an amount of money left on the IPO table from the issuers' point of view. Therefore, underpricing is an expression used to describe the issuer point of view, as he thinks that he has not correctly valued the IPO shares, he underpriced the real value of the shares of his company.

Underpricing is estimated as the percentage difference between the price at which the shares subsequently trade in the market (the first day closing price) and the price at which the IPO shares were sold to investors (the offer price or the issue price) and at which the offering was introduced. Underpricing is suggesting that firms which go public leave considerable amounts of money on the table. The amount of money left on the table is defined as the difference between the closing price on the first day and the offer price, multiplied by the number of shares sold. In the words of Jay Ritter, this is the first-day profit received by investors who were allocated shares at the offer price. It represents a wealth transfer from the shareholders of the issuing firm to these investors. So, underpricing is a loss, a cost for the issuers that they want to minimize and a profit for the first investors that purchase and acquire the IPO shares.

I-2\ A persistent anomaly in time:

In the United States, at the end of the first day of trading, the shares traded on average at 18.9% above the offer price at which the company sold them (1990-2007). Underpricing has averaged 21.2% in the 1960s, 9% in the 1970s, and increasing from 7.8% in the 1980s to 14.4% in the 1990s and to a surprisingly and severe underpricing of 63.5% that exceeded any level previously seen in 1999 and 2000 (reflecting the internet boom years) before falling to 14% in 2001, averaged 11.8% from 2002 to 2006 before rising to 14% in 2007. These are the average levels of underpricing observed in the United States IPO market. When we observe these different levels and percentages of underpricing, we can say that underpricing fluctuates so much and its level changes over time, but it is persistent over time. This anomaly is always observed in the IPO market, whatever the industry to which the offering belongs and whatever the period of going public. The level changes but this anomaly is persistent.

In dollar terms, IPO firms appear to leave many billions «on the table» every year in the U.S. IPO market alone. But the highest amount is in 1999 and 2000, period of internet bubble, this amount of money left on the table at IPO market has reached 66.63 billion dollars, an amount that exceeded any level previously seen. It is the period of internet bubble that attracted the attention of much research effort. Ritter documents that in 1999 and 2000 only, 803 companies went public in the United States, raising about $123 billion, and leaving about $65 billion on the table in the form of initial returns. Loughran and Ritter (2004)¹ explain low-frequency movements in underpricing (or first-day returns) that occur less often than hot and cold issue markets. On a data for IPOs over 1980-2003, they find that IPO underpricing doubled from 7% during 1980-1989 to almost 15% during 1990-1998 before reverting to 12% during the post-bubble period of 2001- 2003, there are some level differences over time but underpricing is persistent.

I-3\ A persistent anomaly in all countries:

Many researchers have been concentrated in studying the persistence of underpricing anomaly internationally. The underpricing phenomenon of Initial Public Offerings (IPOs) has been widely studied across different stock markets around the world and it is a persistent phenomenon all over the world.

Loughran, Ritter and Rydqvist (1994)² documented that the underpricing anomaly exists in all IPO markets. They collected data for 45 countries for different periods and found that underpricing is persistent for all the IPO markets with no exceptions but surely with different levels.

A comparative study by Jenkinson (1990) examines the performance of IPOs in Japan as well as IPOs in the U.S. and the U.K., and concludes that IPOs in these countries are systematically priced at a discount relative to their subsequent trading price. In the U.S. the discount is around 10% while in the U.K., it is around 7%. But when we see the next figures presented by Jay Ritter and that represent the underpricing average in 2008 for many countries, we can see that the level of underpricing has been increased in comparison to the earlier results that were presented by many researchers.

² Loughran, Ritter and Rydqvist (1994): «Initial Public Offerings: international insights».

Loughran et al. (1994) provide also a comprehensive survey of companies going public in 25 countries and find that underpricing is a persistent phenomenon. Ritter (2003) reports the extent of underpricing in 38 countries and finds the same results.

³ This table presents the results of some different studies that were advanced. It is a demonstration of the persistence of IPO underpricing for different countries regardless the period of the study, to insist on the fact that this persistence is not only due to the period of IPOs. Regardless the period, underpricing phenomenon exists for all countries.

I-4\ A persistent anomaly in all industries:

Underpricing is also persistent for all the industries: automobile, banks, chemicals, construction, financial services, food and beverages, industrial, machinery, media, pharmaceutical and health, software, technology, telecommunications, transport and logistics ... Every firm, which decides to go public, faces the phenomenon of underpricing, the price of shares the firm sells tends to jump substantially on the first day of trading regardless its field of activity. Underpricing is persistent for all the firms no exception of the activity to which the firm belongs.

Oehler, Rummer, and N. Smith (2005) in their article «IPO Pricing and the Relative Importance of Investor Sentiment, Evidence from Germany», for a sample of 410 German firms from 1997 to 2001, they classify the issuing companies by field of activity. When observing the number of companies going public and the number of companies having a negative first returns, we can say that most of companies going public have a positive first day return and then their shares are underpriced regardless the field of activity and regardless the industry they are belonging to.

II- Theoretical explanations of short run underpricing: A literature review

The research effort aimed at explaining the short run anomaly of the IPO market has provided numerous analytical advances and empirical insights, and a large list of explanations has been offered.

Ibbotson (1975) is the first who offered a list of possible explanations for underpricing, many of which were formally explored by other authors in later work.

The list of explanations that were advanced to clarify and to understand the underpricing anomaly and to resolve this short run puzzle of the IPO market is long. The considerable research effort in the field of Initial Public Offerings market has resulted in many explanations and the list can not be exhaustive. In this paragraph, I summarize the most important researches and findings and I classify the different theories and explanations advanced in two main categories:

Ø Explanations related to the asymmetric information theory that have been popular among academics and have been considered the most convincing explanations for decades by a great number of researchers, and

II-1\ Asymmetric information:

The key parties to an IPO transaction are the issuing firm, the bank underwriting and marketing the deal, and investors. Asymmetric information models assume that one of these parties knows more than the others.

II-1-1\ The issuer is more informed than the investors:

Welch (1989) and others assume that the issuer is better informed about its true value.

The high quality issuers may attempt to signal their quality and their true value, and to distinguish themselves from the pool of low quality issuers, they voluntarily sell their shares at a lower price than the market beliefs. They leave deliberately money on the IPO table to deter lower quality issuers from imitating, and to demonstrate that they are high quality by throwing money. Investors who are less informed about the issue quality and about its true value will be incited to buy the shares since the price is low and because they begin to believe on the high quality of the issuer. And the issuers with some patience can recoup the amount of money left on the table by future issuing activity. There are some issuers who have the intention to conduct future equity issues at a later date on better terms (Seasoned Equity Offerings SEO) (Welch 1989) or they look for favourable market responses to future dividend announcements (Allen and Faulhaber 1989).

Michaely and Shaw (1994) argue that some issuers voluntarily desire to leave money on the table in order to create, in the words of Ibbotson 1975 «a good taste in investors' mouths» as a signal of high quality, allowing issuers to have more successful Seasoned Equity Offerings in the future. But, surprisingly, they find that the hypothesized relation between initial returns and subsequent seasoned new issues is not present. There is no relation between underpricing and Seasoned Equity Offerings. So, we can say that the explanation of underpricing based on creating a good taste in investors' mouths in order to have the investors' confidence in the future and to conduct future equity issues on better terms and then recouping the amount of money left on the IPO table, is not relevant and it is not convincing.

Jegadeesh, Weinstein, and Welch (1993), using data on IPOs completed between 1980 and 1986, find that the likelihood of issuing seasoned equity and the size of seasoned equity issues increase in IPO underpricing, as expected. However, they note that these statistically significant relations are relatively weak economically. But, Michaely and Shaw (1994) refute completely the existence of this relation between underpricing and SEO.

Guo, Lev and Shi (2006) ⁴, using a sample of 6010 US IPOs from 1980 to 1995, find that R&D expenditures (using the ratio of R&D expenditures to sales or to expected market value for the last fiscal year before IPO as a measure) are the best proxy to informational asymmetry about the issuer quality and find a positive and statistically significant relation between the firm quality and underpricing. R&D expenditures are the intangible investment most extensively researched in economics, accounting and finance, they have to be disclosed in the corporate financial reports. R&D contributes to informational asymmetry such as R&D intensive firms are often undervalued by investors. That is why R&D intensive issuers can not set a high offer price for their IPOs. Besides, they are more willing to forgo money on the table at IPO than are no R&D issuers, because they expect to recoup money left on the table by subsequent issues of seasoned stocks when the market realizes over time the positive outcomes of their R&D (in the words of the authors), because as I presented earlier some studies find no relation between underpricing and SEO.

Guo, Lev and Shi (2006) introduce another measure of firm quality in their model, the Share Overhang Ratio (the ratio of retained shares by insiders to the number of shares issued). They find a positive and statistically significant relation between firm quality and underpricing: the percentage ownership retained by insiders serves as a signal for firm quality. Also, Grinblatt and Hwang (1989) report a positive association between the degree of underpricing and the level of insiders' ownership. For high overhang ratio and so for high quality issues, the issue price is lower a mean to demonstrate their quality and a higher level of underpricing is observed: higher quality firms underprice more than do those of lower quality.

In their article «Why Has IPO Underpricing Changed Over Time?», Loughran and Ritter (2004) use many proxies for the firm quality:

⁴ Guo, R., B. Lev, and C. Shi (2006): «Explaining the Short- and Long-Term IPO Anomalies in the US by R&D».

Share overhang which is the ratio of retained shares to the public float (the number of shares issued) and the Venture Capital a dummy variable which takes a value of one (zero otherwise) if the IPO is backed by venture capital. They find a positive and statistically significant relation between the firm quality and underpricing using the share overhang ratio, but a statistically insignificant relation using the venture capital as a proxy to firm quality. The presence of venture capitalists in the IPO firm is expected to signal issue quality, since their presence reduces the perceived uncertainty over firm value. Venture capitalists have expertise in particular industries and they are expected to make superior investments relative to other investors. In essence, venture capitalists certify the quality of an IPO and their presence signals that asymmetric information is relatively low for the issue and that the issuing firm quality is high and hence leads to a higher issue pricing and to a lower degree of underpricing (Megginson and Weiss, 1991). But in this article, this variable is found insignificant. The same result of insignificance using the venture capital backing as a signal of firm quality is found by Guo, Lev and Shi (2006). A question is then can be asked: Is Venture Capital backing really a signal of firm quality and an explanatory variable that can be introduced in the models?

Ben Dov (2003) looks at the level of institutional ownership shortly after the IPO and finds that high institutional ownership (a signal of firm quality and a proxy for information asymmetry) forecasts higher returns in hot markets.

In the same way of informational asymmetry and firm quality, we can talk about the underwriter reputation (Booth and Smith (1986), Carter and Manaster (1990), Michaely and Shaw (1994)). Some issuers use the underwriter reputation as a signal of high quality and want to hire a prestigious underwriter, since by agreeing to be associated with an offering, prestigious intermediaries «certify» the quality of the issue. When an issuer chooses a prestigious underwriter for the book-building mechanism, he sets a low offer price conducting a high underpricing on one hand, as compensation to the underwriter and on the other hand, he is sure about full subscription. He is concerned by quantity rather than price, and a lower offering price increases the probability of full subscription. It also increases the positive news coverage and the investors' interest in future equity offerings, if issuers have the intention to conduct Seasoned Equity Offerings in later date to recoup the money left on the table. ⁵

⁵ All are explanations advanced by researchers to the underpricing anomaly when issuers choose to hire prestigious underwriters. Because, logically, hiring a prestigious underwriter is a signal of high quality and issuers should require high offer prices for their high quality issues.

This intention of Seasoned Equity Offerings can explain the mystery that issuers seem to be happy and not upset when leaving money on the table, they are certain to recoup this money in later date. ⁶

This explanation of underwriter reputation can take many senses and will be discussed later in other theories.

II-1-2\ The investors are the most informed:

We are in the case of investors that are more informed than the other parties about the demand, the price they are willing to pay to acquire the IPO stocks, ...

To reduce this informational asymmetry, issuers tend to hire underwriters and to use a «Bookbuilding mechanism». This common practice of Bookbuilding consists on setting a preliminary offer price range by the issuing firm, then the underwriter will take the role of trying to collect private information from investors about the offering which is called «indications of interest» such as their demand, the price they are willing to pay to acquire the IPO shares... The bookbuilding mechanism allows underwriters to extract this private information. During pre-selling period, the underwriter tries to gauge demand and to have an idea about the price that potential investors are willing to pay, these indications are then used in setting the final offer price. If there is strong demand and the investors are willing to pay a higher price to obtain the IPO stocks, the underwriter will set a higher offer price and vice-versa.

But, Ritter and Welch (2002) add that if potential investors know that showing a willingness to pay a high price will result in a higher offer price, these investors must be offered something in return. They have to be rewarded for communicating favourable information. Otherwise, if these investors are offered nothing in return and the issuing firm will go public with a higher offer price, investors are dissuaded and they will not reveal their intentions and may even try to reveal wrong information to bias the underwriter's decision.

⁶There is a debate concerning the correlation between underpricing and SEO, many researchers use the SEO as an explanation for not requiring high issue prices for their offerings and others refute this explanation and the relation between SEO and underpricing.

So, to induce investors to truthfully reveal that they want to purchase shares at a high price, underwriters must offer them some combination of more IPO allocations and underpricing when they indicate a willingness to purchase shares at a high price : Benveniste and Spindt (1989), Benveniste and Wilhelm (1990), and Spatt and Srivastava (1991).

Lee, Taylor, and Walter (1999) and Cornelli and Goldreich (2001) show that informed investors request more, and preferentially receive more allocations.

For the issuer, underpricing is on one hand compensation to the underwriter: Baron (1982) has the same theory which is based on the fact that issuers are less informed than underwriters, they delegate the pricing decision to underwriters who possess superior information regarding the demand for the IPOs, and to induce the underwriter to requisite effort to market shares, it is optimal that issuer lets some underpricing because he can not monitor the underwriter without cost. However, we should point out the findings of Muscarella and Vetsuypens (1989) that when underwriters themselves go public and then there is no monitoring cost, their offerings are also underpriced. So underpricing is a necessary cost of going public and not a monitoring cost as advanced by Baron.

On the other hand, underpricing is compensation to investors to truthfully reveal their willingness to purchase the IPO shares and with a higher price.

But we should also mention the fact that underwriters should not underprice too much to not lose business from issuers. Nanda and Yun (1997) find that high levels of underpricing lead to a decrease in the lead underwriter's own stock market value, whereas moderate levels of underpricing are associated with an increase in stock market value.

In a similar sense, Dunbar (2000) finds that banks subsequently lose IPO market share if they either underprice or overprice too much.

As a conclusion for the information revelation theory, we can say that underpricing can be reduced by reducing the informational asymmetry between the IPO parties.

And as Ritter and Welch (2002) note, if underwriters used their discretion to bundle IPOs, problems caused by asymmetric information could be nearly eliminated.

Underwriters, are intermediaries between issuer and investors, they advise the issuer on pricing the issue, both at the time of issuing a preliminary prospectus that includes a file price range, and at the pricing meeting when the final offer price is set. In the bookbuilding period, they try to collect the investors' private information which can help in setting the price and even the volume of IPO shares. Their role is very important in reducing the information asymmetry by transmitting the investors' private information to issuers. So, the revelation information theory is based on the work of underwriters.

I presented in the previous paragraph that underpricing is positively associated with underwriter reputation. For example, consistent with evidence from the 1990s (Beatty and Welch (1996)), Ljungqvist, Nanda and Singh (2004) predict that underpricing increases in underwriter prestige, but that this relation depends on the state of the IPO market.

Also, Benveniste, Ljungqvist, Wilhelm and Yu (2003) find a positive relation between underpricing and underwriter prestige in the 1999-2000 hot market.

First, we should present the measures used of underwriter reputation: Carter and Manaster (1990) provide a ranking of underwriters based on their position in the `tombstone' advertisements in the financial press that follow the completion of an IPO. This ranking, since updated by Jay Ritter, is much used in the empirical IPO literature, it is a discrete underwriter reputation measure ranging from 0 to 9, where a 9 (0) represents the most (least) prestigious underwriter.

Megginson and Weiss (1991) measure underwriters' reputation instead by their market share, and this approach too is widely used.

The fact of using a prestigious underwriter in the bookbuilding process has an impact on signalling the high quality of the issuing company as I said earlier, but it has also a great impact on the information revelation, since prestigious underwriter is a source of confidence to investors. The prestigious underwriter has a great ability and power to make investors reveal their private information.

Many explanations were presented to the positive relation: when an issuer decides to hire a prestigious underwriter, he is sure about a full subscription, he is concerned by quantity rather than price, and by setting a low offer price, he tries to increase the probability of full subscription. There are some other issuers who are concerned by their firm quality and by hiring a prestigious underwriter, they tend to demonstrate their high quality and they set a low offer price as a signal of quality and of the real value of the company. From the point of informational revelation theory, underpricing and underwriter reputation are positively related because underpricing is used as compensation to the underwriter (Baron 1982).

In Loughran and Ritter (2004) model, top-tier underwriters are associated with more underpricing in the 1990s, and especially in the bubble period.

Loughran and Ritter (2003) argue that prestigious banks have begun to underprice IPOs strategically, in an effort to enrich themselves or their investment clients. Another explanation is that top banks have lowered their criteria for selecting IPOs to underwrite, resulting in a higher average risk profile (and so higher underpricing) for their IPOs.

But the relation between underwriter reputation and underpricing is not systematically positive and it is empirically mixed, there are some researchers who find that the correlation between underpricing and underwriter reputation is negative: when an issuer hires a prestigious underwriter, he can set a higher offer price and then lower underpricing will be observed. For example, Carter and Manaster (1990), and Carter Dark and Singh (1998) find that more prestigious underwriters are associated with lower underpricing.

There are also some researchers who have introduced the underwriter reputation variable in their models and find that it is statistically insignificant as in Guo, Lev and Shi article (2006). These authors have introduced the underwriter reputation as an explanatory variable to explain underpricing and find that it is statistically insignificant.

In practice, results are not very sensitive to the choice of underwriter reputation measure, but they are highly sensitive to the period studied.

Rock (1986) assumes that some investors are more informed than are other investors in general, the issue firm, and its underwriter. There is an imbalance of information between the potential investors themselves. These better informed investors bid only for attractively priced IPOs. In the case of unattractive offering, the better informed investors will not purchase the shares and the uninformed investors will have all shares they have bid for because they bid indiscriminately. But, these uninformed investors are unable to absorb all the shares offered, that is why underpricing is necessary to incite informed investors to bid for this offering's shares even if they think it is unattractive. The lower offer price will incite them to try to have the offering's shares. This underpricing is necessary also to not result in a negative return for this offering for the uninformed investors, whose capital is needed even for an attractive offering. They must be protected in IPO market to not lose their capital and their participation because in the case of attractive offerings, informed investors are also unwilling to absorb all the shares offered and their demand is insufficient. The uninformed investors should not fear the IPO market, positive returns are required to protect them and to ensure a continued participation in IPO market.

In the spirit of Rock (1986), Chowdry and Nanda (1996) develop a model in which price support is used as a complement to underpricing to reduce the losses supported by uninformed investors and induce them to participate in the IPO.

In conclusion for the theory of winner's curse of Rock 1986, which is an application of Akerlof's (1970) lemons problem, underpricing is necessary to not lose the capital of uninformed investors and to incite informed investors to take part of an offering allocation even if it is unattractive. Pricing too high might induce investors and issuers to fear a winner's curse.

In the same sense of inciting informed investors to participate in an unattractive offering, Ljungqvist, Nanda and Singh (2001) find and explain underpricing by the fact that noise traders cannot absorb the entire IPO because they are wealth constrained. To induce rational investors to participate in the offering, issuers must set the IPO price below the price noise traders are ready to pay, to induce underpricing and to make the offering attractive for the rational investors. These better informed investors can sell the shares to the irrational investors in the aftermarket and make a profit.

Rock's (1986) winner's curse model turns on information heterogeneity among investors. Michaely and Shaw (1994) argue that as this heterogeneity goes to zero, the winner's curse disappears and with it the reason to underprice.

Tore Leite (2007) in his article «Adverse selection, public information and underpricing in IPOs» finds that favourable public information (such as high market returns) reduce the winner's curse problem. And this induces the issuer to price the issue more conservatively in order to increase its success probability. The author finds a positive relation between public information and underpricing.

* Agency conflicts: The agency problems between the underwriter and the issuing firm.

Underpricing represents a wealth transfer from the IPO company to investors. These investors compete for allocations of underpriced stock, they can even try to collude with the underwriter by offering side-payments if they have underpriced stocks. Such side-payments could take the form of excessive trading commissions paid on unrelated transactions, or investment bankers might allocate underpriced stock to executives at companies in the hope of winning their future investment banking business, a practice known as «spinning». So the underwriters, seeking for their own interests have an incentive to underprice IPOs if they receive commission business in return for leaving money on the table, they will try to underprice the issuer's stock by not revealing the truthfully information obtained from potential investors ...

The underwriter will use a sub-optimal effort. He will use his informational advantage over issuing companies to increase his benefit and he will not do his mission perfectly. The underwriters are given discretion in share allocation. This discretion will not automatically be used in the best interests of the issuing firm, and they can also underprice more than necessary and then allocate these shares to favoured buy-side clients.

In the post-bubble period, increased regulatory scrutiny reduced spinning dramatically. This is one of several explanations why underpricing dropped back to an average of 12%.

As a conclusion, we can say that issuing firms who seek out prestigious underwriters to advice, to look after the IPO process, to bear for the risk of hot market crashing... can suffer from enormous problems if conflict of interest problems are not controlled.

Baron and Holmström (1980) and Baron (1982) construct screening models which focus on the underwriter's benefit from underpricing. In a screening model, the uninformed party (issuer) offers a menu or schedule of contracts, from which the informed party (underwriter) selects the one that is optimal. The contract schedule is designed to optimize the uninformed party's objective, which, given its informational disadvantage, will not be first-best optimal. To induce optimal use of the underwriter's superior information about investor demand, the issuer in Baron's model delegates the pricing decision to the bank (the underwriter). Given its information, the underwriter self-selects a contract from a menu of combinations of IPO prices and underwriting spreads. If likely demand is low, he selects a high spread and a low price, and vice versa if demand is high. But as we said earlier, underwriter can use this delegation to increase his own benefit and to think about his individual interest only. He can collude with the potential investors to the potential detriment of the issuer.

Ø Issuers can monitor the investment bank's selling effort and bargain hard over the price,

Ø or they can use contract design to realign the bank's incentives by making its compensation an increasing function of the offer price.

Ljungqvist and Wilhelm (2003) provide evidence consistent with monitoring and bargaining in the U.S. in the second half of the 1990s. They show that first-day returns are lower, the greater are the monitoring incentives of the issuing firms' decision-makers.

Ljungqvist (2003) studies the role of underwriter compensation in mitigating conflicts of interest between companies going public and their investment bankers (their underwriters). Making the bank's compensation more sensitive to the issuer's valuation should reduce agency conflicts and thus underpricing. Consistent with this prediction, Ljungqvist shows that contracting on higher commissions in a large sample of U.K. IPOs completed between 1991 and 2002 leads to significantly lower initial returns, after controlling for other influences on underpricing and a variety of endogeneity concerns.

So, we can say that by making underwriter's compensation an increasing function of the offer price or of the issuer's valuation, the agency conflicts between underwriter and the issuing company could be mitigated. Underpricing could even disappear if we consider the agency conflicts as the primary driver and source of underpricing.

As a conclusion for the asymmetric information theories, we can say that almost informational asymmetry theories share the prediction that underpricing is positively related to the degree of asymmetric information, and when asymmetric information uncertainty approaches zero, underpricing disappears entirely.

II-2\ Symmetric information:

There are some theories and explanations advanced by a great number of researchers that do not rely on asymmetric information, this theory that has been very popular among academics and practitioners for decades and that has been considered as the most relevant and convincing explanation to the short run IPO anomaly. There are some researchers that advanced other theories and they explain underpricing by other reasons, asserting symmetric information between key IPO parties. As hypothesis, all the key parties of an Initial Public Offering share the same information. We talk about informational transparency and lucidity and about IPO market efficiency.

II-2-1\ Risk premium:

Let's begin by the risk premium explanation. Because the hot market can end prematurely, the sentiment demand may cease and then we face a market crashing, carrying IPO stocks in inventory is risky.

Ljungqvist, Nanda and Singh (2003) in their article: «Hot market, investor sentiment and IPO pricing» argue that underpricing emerges as fair compensation to the regulars for expected inventory losses arising from the possibility that the hot market ends prematurely. If the demand is small (in comparison to the issue offer), the issuer needs the regular investor to hold inventory. So long as the hot market persists, the regular investor sells this inventory to newly arriving sentiment investors, but the problem arises if the hot market ceases and the regular investor is then left with shares priced at the fundamental value (which is less than the offer price).

The issuer underprices the stock to compensate the regular investor for bearing the risk of an uncertain sentiment demand. It is a fair payment for the regular's expected loss. It is a way of compensating the regular investor for taking on the risk of hot market crashing.

However, Ritter and Welch (2002)⁷ refute this explanation since they argue that if the underpricing is simply a compensation for bearing a systematic or liquidity risk, why do second-day investors not seem to require this premium, after all fundamental risk and liquidity constraints are unlikely to be resolved within one day.

As a conclusion for the risk premium explanation based on Ritter and Welch (2002) point of view, we can say that this explanation is refuted and can not be considered as a relevant and a convincing explanation for the underpricing anomaly.

II-2-2\ Characteristics of the Initial Public offering:

Loughran and Ritter (2004) use many measures of risk: the natural logarithm of the assets and the natural logarithm of the sales which reflect the issuing firm size and then a risk related to valuation uncertainty, internet and tech dummy variables which reflect technological uncertainty which also induces a valuation uncertainty, and the natural logarithm of one plus the age (years since the firm's founding date to the date of going public and the date of introduction in IPO market) which reflect the age of the issuing firm. For a sample including 5,990 US operating firm IPOs over 1980-2003, they find a positive relation between risk and underpricing.

If the issuing firm is risky from the investors' point of view, it can not be introduced in the market at a higher price because it will not be accepted by these investors who are dissuaded about the risky IPO shares. The offer price is set at a lower level to incite investors

⁷ Ritter and Welch (2002): «A review of IPO activity, pricing, and allocations».

The risk composition hypothesis, introduced by Ritter (1984), assumes that riskier IPOs will be underpriced by more than less-risky IPOs. Riskier firms set a low offer price to incite investors to participate in the IPO market and to buy the IPO risky shares, and then the underpricing will be higher. For example, young firms are riskier and the internet bubble period saw a high proportion of young firms going public and a high percent of underpricing, which can confirm the explanation of risk.

In the same direction of the risk, we can also talk about the uncertainty level introduced by Beatty and Ritter (1986). They relate the level of ex-ante uncertainty surrounding the intrinsic value of an IPO to the level of underpricing. The higher the uncertainty level about the intrinsic value of the IPO, the higher is the level of underpricing. It reflects the valuation uncertainty.

Besides, Bartov, Mohanram and Seethamraju (2003) report that a dummy variable for risky IPOs has no effect on the setting of the final offer price, providing evidence for the argument that risk might not be that important for the pricing of IPOs. So there is no correlation between risk and underpricing. The notion of risk can not explain the setting of a lower offer price and then the underpricing phenomenon. This important finding refutes the prior researches and results about the suitability of the risk as an explanation to the underpricing anomaly. Risk can not be considered as a convincing explanation to the short run IPO anomaly since it has no effect on the setting of a lower offer price, and then underpricing is not induced by risk.

* Issue size: The issue size or offer size is the number of shares introduced in IPO market and offered by the issuing company for sale.

Cornelli, Goldreich and Ljungqvist (2004) argue that when the issue size is large, the issue price should reflect the greater difficulty of selling the shares in the aftermarket, and then the issue price should be lower. They find a negative relation between the size and the offer price of IPOs: a discount in the offer price by ëS with S the size of IPOs. When the offer size is large, the offer price will be lower and so underpricing will be higher. They find that the issue price should be negatively correlated with the issue size, and it is a negative and statistically significant relation between size and offer price, so we can talk about a positive and significant relation between issue size and underpricing.

This relation between issue size and underpricing was studied briefly in Ljungqvist, Nanda and Singh (2003) article. Supposing V_Rthe market price of the IPO shares and V_S the value sentiment investors place on the IPO shares and so is the price these investors are willing to pay for the IPO shares, the authors set V_Sas a function of V_R and Q which represents the total number of IPO shares (the offer size), and the relation between V_Sand Q is negative. The greater is the number of shares issued, the lower is the price that sentiment investors are willing to pay, and the greater is underpricing: positive relation between issue size and underpricing.

Besides, some researchers find a negative relation between the issue size and underpricing. Guo, Lev and Shi (2006) introduce a different issue size variable in their model (the natural logarithm of issue proceeds) and find a negative relation between issue size and underpricing. They explain this finding by the fact that sizable firms are generally less risky than those making smaller issues, and they can bargain for a higher offer price conducting a lower level of underpricing.

Ljungqvist, Nanda and Singh (2003) study the impact of bargaining power on the first day return and so on underpricing. They use as a proxy for bargaining power «the ownership structure». A firm with a highly concentrated ownership, is reflecting a high incentive to bargain hard, while an increased ownership fragmentation, and an increased frequency and size of «friends and family» share allocations, make the issuing firm decision-makers less motivated to bargain for a higher offer price .

Ljungqvist, Nanda and Singh assume that the issuing firm's ownership structure is such that â of the extracted surplus from the investor sentiment is captured by the issuing firm and 1-â is captured by a combination of the regular investor and the investment bank. For a firm with highly concentrated ownership, they believe â will be close to 1 reflecting the high incentive to bargain hard over the surplus, while for a firm with dispersed ownership or other agency problems â will be significantly smaller than 1.

An issue firm with a concentrated ownership and high bargaining power requests a higher offer price and faces a lower underpricing. Ljungqvist and Wilhelm (2003) show that companies with more concentrated ownership at the time of the IPO suffer lower underpricing, consistent with the findings of Ljungqvist, Nanda and Singh (2003).

And an issue firm characterized by ownership fragmentation and so by lower bargaining power, can not bargain hard for a higher offer price. The issue price will be set at a lower level and underpricing will be higher in the first day of trading. Tim Loughran and Jay Ritter (2004)⁸also use the ownership structure as a proxy for bargaining power and find that an increased ownership fragmentation induces a decrease in the bargaining power of the issuing firm, a lower offer price is presented inducing a higher level of underpricing.

The greater the issuing firm's bargaining power relative to the underwriter, the higher is the offer price and the lower is the first-day return and vice-versa.

II-2-3\ Lawsuit avoidance: legal liability

Lawsuits are obviously costly, not only directly: damages, legal fees, diversion of management time, etc, but also in terms of the potential damage to their reputation capital. Litigation-prone investment banks may lose the confidence of their regular investors, while issuers may face a higher cost of capital in future capital issues.

The basic idea of lawsuits avoidance goes back at least to Logue (1973) and Ibbotson (1975): companies deliberately sell their stocks at a discount to reduce the likelihood of future lawsuits from shareholders disappointed with the post-IPO performance of their shares.

Tinic (1988), Hughes and Thakor (1992) and Hensler (1995) argue that issuers intentionally underprice to reduce their legal liability. They assume that the probability of litigation increases with the offer price: the more overpriced an issue, the more likely is a future lawsuit, and that the solution is underpricing to avoid future lawsuits. Ritter and Welch (2002) give a simple example for this: An offering that starts trading at 30$ that is priced at 20$ is less likely to be sued than if it had been priced at 30$, if only because it is more likely that at some point the aftermarket share price will drop below 30$ than below 20$.

Tinic identifies a sample of 70 IPOs completed between 1923 and 1930 and compares their average underpricing to that of a sample of 134 IPOs completed between 1966 and 1971. As Tinic predicted, average underpricing was lower before 1933 (year of securities enactment).

In spite of this, Drake and Vetsuypens (1993) find that underpricing did not protect issuers firms from being sued, and sued IPOs had higher and not lower underpricing.

⁸ Loughran and Ritter (2004) argue that this argument has little support as an explanation for underpricing.

They study a sample of 93 IPO firms that were sued and compare them to a sample of 93 IPOs that were not sued, matched on IPO year, offer size, and underwriter prestige. Underpriced firms are sued more often than overpriced firms. Then underpricing does not protect firms from being sued and does not protect them from lawsuits and future legal liabilities.

They also show that average initial returns in the six years after Tinic's sample period (1972-1977) were actually lower than between 1923 and 1930 which refute his findings.

Ritter and Welch (2002) think that leaving money on the table appears to be a cost-ineffective way of avoiding subsequent lawsuits. But the most convincing evidence that legal liability is not the primary determinant of underpricing is that countries in which U.S. litigative tendencies are not present have similar levels of underpricing (Keloharju (1993)):

The risk of being sued is not economically significant in Australia (Lee, Taylor, and Walter (1996)), Finland (Keloharju (1993)), Germany (Ljungqvist (1997)), Japan (Beller, Terai, and Levine (1992)), Sweden (Rydqvist (1994)), Switzerland (Kunz and Aggarwal (1994)), or the U.K. (Jenkinson (1990)), all of which experience underpricing.

After all these findings which refute the suitability and the relevance of lawsuit avoidance as an explanation to underpricing anomaly, we can say that lawsuit avoidance can not be a primary determinant and driver of underpricing, still, it is possible that lawsuit avoidance is a second-order driver of IPO underpricing.

II-2-4\ Underpricing as a substitute of marketing expenditures:

Habib and Ljungqvist (2001) argue that underpricing is a substitute for costly marketing expenditures. Using a data set of IPOs from 1991 to 1995, Habib and Ljungqvist report that an extra dollar left on the table reduces other marketing expenditures by a dollar. On the first sight, underpricing seems to be just a substitute for marketing expenditures since both have the same cost, but underpricing is much more interesting.

By going public and by underpricing and leaving money on the IPO table, issuing firms can achieve a total coverage media and good news in all media, which can be much more costly if the firm chooses to use publicity and marketing expenditures, mainly because we can not forget the possibility of recouping this money left on the IPO table if the firm has the intention to conduct Seasoned Equity Offerings in the future. So, the issuing firm achieves a large coverage media and an important publicity without spending anything, since the money left on the IPO table will be recouped later. By underpricing, the investors who bought the IPO shares have confidence in the issuing firm, they are satisfied with the gains they retired from the IPO shares and from underpricing in the first day of trading. Optimistic about the value of this firm, these investors will easily accept the price the firm set for its Seasoned Equity Offerings at a later date. Even if the price is higher than necessary, they will accept it since they have confidence in this firm, and then all the money left on the table can be recouped by conducting Seasoned Equity Offerings in the future. ⁹

II-2-5\ Internet Bubble:

One popular related explanation for the high and severe underpricing of 65% during the Internet bubble (1999-2000) for the U.S IPOs, a peak never reached before in the U.S IPO market, is that underwriters could not justify a higher offer price on Internet IPOs. Even if these firms have a high potential of profitability in the recent future and they are operating in a new but very promising field which will generate high returns later, the underwriters can not justify this and propose a higher offer price. These firms are seen as young and operating in a new field which means that their offerings are risky and they propose risky shares. Proposing a higher offer price will not be accepted by investors and will make them fear the offerings. The issuing firms have to propose a low offer price to incite investors to participate in the offerings even if they are thought to be risky. So, we can say on one hand, the newly issuing Internet firms are very important and are operating in a very promising field and then will generate high returns, but all this can not be justified by their underwriters and they do not find the convincing arguments. On the other hand, and since they are operating in a new field not very known and they are very young firms without a history of returns, they are thought to be risky. So, we are in the same explanation of risk due to valuation uncertainty which was proved to be ineffective determinant and explanation for the underpricing anomaly.

So to explain the severe level of underpricing during the dot-com period, only the inability to justify a higher offer price can be considered as a possible explanation, but the fact of young and so risky firms can not be used as a relevant explanation.

⁹ As a support for IPO as a marketing event, Chemmanur (1993) proposes that this publicity could generate additional investor interest, and Demers and Lewellen (2003) suggest that the publicity could generate additional product market revenue from greater brand awareness.

II-2-6\ Price stabilization and partial adjustment:

Recent studies have also documented the impact of public information. They find a positive link between the «market conditions» prevailing at the time of an offering which represent public information and its subsequent initial return. Favourable market conditions predict higher underpricing and vice-versa. Derrien and Womack (2003) show that the initial returns on IPOs in France in the 1992-1998 period were predictable using the market returns in the three-month period preceding the offerings. Using U.S. data, Loughran and Ritter (2002) and Lowry and Schwert (2003) obtain similar results: the initial returns in the first day of trading for IPOs are predictable using the market conditions prevailing at the time of the IPOs or at a recent past. Favourable market conditions predict higher initial returns and so higher underpricing, and critical market conditions predict lower initial returns and lower underpricing, and for some IPOs negative initial returns and overpricing.

These favourable market conditions have an impact on noise traders who will be ready to pay higher IPO prices. They are assumed to be bullish at the time of the offering since they are very influenced by market conditions: the more favourable market conditions are, the more favourable noise traders' sentiment is and the higher the price that they are willing to pay.

But the mystery and the question that arises automatically is, why underwriters do not incorporate these favourable market conditions and this favourable sentiment when pricing the IPO and propose a higher offer price which reduces the underpricing anomaly and the money threw on the table? Why market conditions and noise traders' sentiment are only partially incorporated in IPO offer prices?

The underwriter is not only concerned with the IPO price at the time of offering, but he is also concerned with the aftermarket behaviour of IPO shares in the short run as well as on the long run. He is committed to provide costly price support if the aftermarket share price falls below the IPO price (the issue price) in the months following the offering. Even if the noise traders are bullish at the time of offering, they can change their attitude in the aftermarket.

And a sharp and rational underwriter with a reasonable attitude should take this into consideration when pricing the IPO. Derrien (2003) says that the IPO price results from a trade-off: a higher IPO price increases underwriting fees, but also the expected cost of price support.

This induces the underwriter to set a conservative IPO price with respect to the short-term aftermarket price of IPO shares. The underwriter has to incorporate partially the market conditions when pricing IPO to not face a higher cost of price support if the aftermarket price falls.

When setting the offer price, we can say that the underwriter is constrained by the cost of price support if the aftermarket price of IPO shares falls below the issue price (if he overpriced the IPO shares), but he is also constrained by the lost of IPO market shares if he sets a very low issue price to reduce the risk of support's price costs inducing very high underpricing. In this sense, Booth and Smith (1986) claim that the underwriter's role is to certify that IPO shares are not overpriced. Therefore, an underwriter that underprices more than necessary the IPOs will lose market shares on the IPO market and will lose the issuing firms' confidence, this is confirmed empirically by Dunbar (2000), and an underwriter that overprices the IPO shares will pay high costs of price support. Legal costs may also refrain underwriters from blatantly overpricing new issues, even though investors' demand is very large.

Aggarwal (2000) and Ellis, Michaely and O'Hara (2000) provide evidence that underwriters intervene to stabilize the price of IPO stocks that exhibit poor aftermarket performance. The underpricing is important to stabilize the IPO prices to compensate the long run underperformance of the issues.

Section 2- Behavioral explanations

Short run IPO anomaly may be the most controversial area of IPO research. The research effort has provided numerous analytical advances and empirical insights trying to explain the first day price run up. Many explanations were introduced and studied, but all these theories are unlikely to explain the persistent pattern of high initial returns during the first trading day. It is fair to say that this anomaly is not satisfactorily resolved and is still a puzzle.

Many researchers come to the conclusion that IPO future researches should turn to behavioral approach and sentiment notion. Research effort should focus more on behavioral explanations to clarify and to explain the short run IPO behaviour, since nor the asymmetric theories neither the explanations based on the informational symmetry are likely to give a relevant and a reliable explanation to the underpricing anomaly, mainly after the surprisingly and severe level of underpricing reached in 1999 and 2000, the internet boom years. The explanations that have been advanced earlier are unable to explain the severe percentage of underpricing observed in this period, and turning to behavioral explanations seems to be a necessity to resolve this short run IPO puzzle.

Ritter and Welch (2002), for example, advance clearly that asymmetric information which has been the most convincing explanation for decades, is not the primary driver of IPO phenomena and asymmetric information models that have been popular among academics, have been overemphasized. Ritter and Welch believe that future progress in the IPO literature will come from non rational explanations.

This argument is supported by Ljungqvist (2004) who comes to the conclusion that IPO researchers should focus on behavioral approaches to explain why the extent of underpricing varies so much over time.

Future researches have to focus more on behavioral explanations and turning to the behavioral approach describing the behaviour of irrational investors and their sentiment seems to be a necessity since the explanations and the theories that were advanced earlier are unlikely to clarify and to explain the short run IPO puzzle, especially after the dot-com boom. During this period, the IPO market has reached a severe level of underpricing never seen before. So one can ask about the suitability of «traditional» explanations and theories, and turning to behavioral explanations seems to be the most promising area of research. Going further, Cook, Jarrell and Kieschnicke (2003) conclude that the role of investor sentiment is more important than previously thought.

The tendency of behavioral approach and investor sentiment is not a new field not again discovered. The investor sentiment was introduced earlier in the 1990's by Welch who presented the informational cascade theory. But this approach has attracted more attention, has intrigued more and more researchers and has taken all its impetus in this decade. Many researchers tried to introduce this behavioral approach to explain the short run IPO anomaly and the research effort is continuing.

In this second section, I present the Behavioral Approach in a first paragraph by defining important notions as investor sentiment, hot IPO market and distinguishing between individual investors and institutional investors. Then, in a second paragraph, I summarize the most important studies advanced asserting the presence of sentiment investors and sentiment as the primary and main driver of underpricing phenomenon. I present the most important findings and the main proxies used by researchers to value the investor sentiment.

I- Definitions:

The behavioral approach asserts the presence of «irrational» investors, also called «sentiment» investors or «noise traders» whose investment decisions, choices, etc, are conducted by feelings and emotions and based on sentiment which plays a major role in their decisions.

Before going more in the details, it is reasonable to begin by presenting some definitions of the most important notions in this section. Clarifying some notions is important to understand the theories and the findings of many researchers.

I-1\ The sentiment's notion:

The sentiment represents the anticipations of the investors that are not justified by the fundamental.

The notion of sentiment characterizes the presence of irrational investors who show undue interest in an investment opportunity, for example for IPOs, and who are irrationally exuberant, over optimistic and over enthusiastic about an investment. This over optimism and over enthusiasm is not justified by fundamental. Or on the contrary, these investors are over pessimistic and they are much dissuaded about the issues with any reason or relevant justification. All the decisions are conducted by sentiments and feelings.

An optimistic investor (pessimistic) expects returns that are higher (lower) to those that could be explained by the fundamental indicators. In other words, the sentiment can be defined by the fact that the investor is optimistic (pessimistic) without having good (bad) economic reasons for the being.

The sentiment felt by the investor is very complicated and very hard to surround, since there are numerous biases that can have an impact on the investor behaviour. The sentiment also differs from an investor to another, it is individual and can not be foreseen. It depends on the investor's personality, his beliefs, his thoughts...

Biases have been studied by psychologists for some time and financial economists have recently introduced them into formal models of asset pricing. For example, a large literature reports that people believe their knowledge to be more accurate than it really is (Odean (1998)), and then they overweight the information collected by themselves and underweight the information collected by the others. In the same direction, Kaustia and Knupfer (2008) on a sample of 57 Finnish IPOs from January 1995 through December 2000, study the link between past personally experienced outcomes and future IPO subscriptions. They find that personally experienced outcomes have a greater effect on behaviour and on future IPO subscriptions, than, say just reading about the same information without personal involvement. When deciding to subscript in IPOs, investors overweight their personal experience more than the information collected by the other investors, which goes in the same direction as believing in their knowledge and in their own information.

However, there are other possible reasons for systematic decision errors. In a recent review, Hirshleifer (2001) argues that many or most familiar psychological biases can be viewed as outgrowths of heuristic simplification (an imperfect decision making procedure that makes people have reasonably good decision cheaply). We can also talk about framing effects (wherein the description of a situation affects judgments and choices), money illusion (wherein nominal prices affect perceptions), and mental accounting (tracking gains and losses relative to arbitrary reference points). We also find overconfidence (a tendency to overestimate ones ability or judgment accuracy) which may be due to another bias «self attribution». Experiments have shown that people tend to attribute favourable outcomes to their abilities and unfavourable ones and failure to chance or other external factors beyond their control (Daniel, Hirshleifer, and Subrahmanyam (1998)). This bias is known as «self attribution» and may even be at the origin of «overconfidence».

These psychological biases have a great effect on the sentiment, and they also differ from an investor to another making the sentiment a very complicated notion hard to surround or to foresee.

The IPO market presents an environment which is more prone to investor sentiment and where irrational investors are more likely to exist, since IPO firms by definition have no prior share price history and tend to be young, immature, and relatively informationally opaque.

Not surprisingly, therefore, these firms are hard to value, and it seems reasonable to assume that investors will have a wide range of beliefs and feelings about their market values. Investors will be very influenced by feelings and emotions in valuing the IPO and in deciding whether to invest in.

I-2\ Hot IPO market's phenomenon:

As we said before, by their nature the Initial Public Offerings are very sensitive to the state of mind of the investors and to the investor sentiment.

Another characteristic is very important and which ensues of the first characteristic: the Initial Public Offering market is highly cyclical. The cyclical nature of this market has sparked much academic interest. This attention has produced a number of explanations for hot and cold IPO issue markets based on changing business conditions (e.g., Pastor and Veronesi, 2005), investor sentiment (e.g., Ritter, 1991), and asymmetry of information between owners and outside investors (e.g., Myers and Majluff, 1984).

But to explain the notion of «hot IPO market», I focus on the investor sentiment explanation as an important driver and determinant of the IPO market cycles, patterns and phenomena and since it is the object of this thesis.

The phenomenon of «hot IPO market» (Hot market) was for the first time noticed by Ibbotson and Jaffe (1975). These two authors, as well as Ibbotson, Sindelar and Ritter (1988), showed the existence of cycles in the IPO market concerning the monthly number of introductions and the level of initial returns of the shares introduced. Especially, a hot IPO market is a market in which recent IPOs have generated strong and positive initial returns and so high underpricing was observed in the recent past. The sentiment and irrational investors are over optimistic and over enthusiastic about the IPO market, they are exuberant and they show undue interest in newly IPOs, since they have the tendency to project the recent favourable history of the IPO market on the future. In their mind, this positive and favourable tendency will continue for sure and they will make large profits. The market appears to be irrationally optimistic towards IPOs in the short run and so these irrational and over optimistic investors are more willing to purchase the newly IPOs and to pay more to have the newly shares. These irrational investors show excessive demand, they bid up the price of IPO shares beyond true value. The price of shares tends to jump substantially on the first day of trading, the first day closing price is systematically higher than the issue price at which the public offering was introduced in the market, and then IPOs exhibit positive first day returns and so underpricing will be higher. As a result, strong demand towards IPOs and over optimism among irrational investors cause the rise of the stock price during the first trading day. Then, we see the tendency that they have projected when deciding to invest in newly IPOs is maintained. All this is due to the bullish sentiment and the investors' over optimism and their irrational excessive demand, and we talk about a phenomenon much known as a «hot IPO market».

Therefore, more firms will be incited to go public taking advantage of this mistaken belief and of the over optimism of these sentiment and irrational investors. Lee, Shleifer and Thaler (1991), Helwege and Liang (1996), Rajan and Servaes (1997) and Lowry (1999) showed that the volume of IPOs is associated to the positively state of mind of the investors and to their excessive optimism and enthusiasm. The issuing firms are encouraged to go public at this period and to increase the volume of IPO shares they are offering in the stock market.

Loughran, Ritter and Rydqvist (1994) go further in claiming that issuers time their IPOs to coincide with periods of excessive optimism, consistent with the finding in Lee, Schleifer and Thaler (1991) that more companies go public when investor sentiment is high.

Baker and Wurgler (2002) argue that firms may even be able to time their IPO to coincide with periods of excessive valuations which goes in the same direction of the investors' over optimism and over enthusiasm. ¹⁰

Helwege and Liang (1996) and Ljungqvist, Nanda and Singh (2002) models find evidence of over optimism in hot IPO markets which confirms the relevance of investor sentiment as a driver of the IPO market cycles, the explanation advanced by Ritter (1991) for the hot IPO market.

Using German data on IPO trading by 5000 retail customers of an online broker, Dorn (2003) documents that retail investors¹¹overpay for IPOs following periods of high underpricing in recent IPOs, and for IPOs that are in the news.

Consistent with all these findings showing the importance and the major role of investor sentiment and over optimism in IPO market making it cyclical, Michelle Lowry (2003)¹² advances that the instability and the movement in Initial Public Offering volume over time both in the number of IPOs and the total proceeds raised in these offerings, can be explained and attributed to the variation in investor optimism level. When investors are overoptimistic, the issuing firms should profit from this period and are incited to offer large volume of IPO and they are sure that the large volume of shares will be absorbed by sentiment investors and vice-versa. The results found by this author indicate that investor sentiment is an important determinant of IPO volume both in statistical and economical terms. The variation in IPO volume is primarily driven by changes in investor optimism, as well as the firms' demands for capital (business cycle), the adverse selection costs (investors' uncertainty about the true value of the issuing firm) which are statistically significant but their economic effect appears small.

The end of the nineties ¹³ was one of the hottest IPO markets ever. In this period, both the number of initial public offerings and the level of initial returns have reached unprecedented peaks.

Controversially, the IPO market can be in a cold period, the investors' sentiment is bearish and the irrational investors are pessimistic. They observe the recent IPO history which knows

¹⁰Pagano et al. (1998) add that firms «time» their IPOs to coincide with periods of investor optimism, not to exploit investment opportunities, but to realign their balance sheets after large investments and growth.

negative initial returns, and they project this on the future. They think that the newly IPOs are unattractive and they are dissuaded. They refuse the investment in newly IPOs and they fear the IPO shares which will be overpriced.

This period is known as «cold IPO market». Issuing firms should take into consideration this critical period and it is reasonable to delay their decision to go public until a bull market offers more favourable conditions and pricing to succeed their introduction in the market. However, they surely will fail the introduction if they do not take into consideration the state of the IPO market.

The popular press contains many examples of this viewpoint of the importance of investors' sentiment to decide whether to go public. For example, ``the [current] rule in the IPO market seems to be: Buy it at any price'' (Wall Street Journal, May 20, 1996, p. ) (a period of high sentiment and over optimism), and ``When [investors] get bearish, you can't go public. But when they go bullish, just about anyone can go public'' (Wall Street Journal, April 19, 1999, p. C1.).

Another important remark should be added, if the issuer succeeds the first sales, he is sure that later sales will be succeeded and then he succeeds his introduction in the market taking all the advantages that are related to this success. But if he fails the first sales, taking the decision to go public in a cold market, the introduction in the market will fail conducting even a bankruptcy in the recent future, going with the informational cascades theory of Welch (1992) that I will present in a following paragraph.

Ljungqvist, Nanda and Singh (2003) present another definition for a «cold market», they define it as a market in which there are no exuberant investors and so prices are set by rational investors at fundamental value, so there are nor optimistic neither pessimistic investors who can conduct the IPO pricing. The offerings are priced at their fundamental values.

I-3\ Investors typology:

Many empirical researches examining IPO allocations focus on the distinction between types of investors: institutional investors and individual or retail investors.

Institutional investors are different from retail investors, in that institutions are better informed and more important clients. The evidence to date suggests that where bookbuilding is used, institutional investors receive preferential allocations. They are favoured by underwriters in the IPO allocations compared to retail investors who are uninformed and not very important clients since they do not buy many shares. Using U.S. data, Hanley and Wilhelm (1995) and Aggarwal, Prabhala, and Puri (2002) find that institutions are favoured, as do Cornelli and Goldreich (2001) using U.K. data. Cornelli and Goldreich (2001) also find that more information-rich requests are favourably rewarded.

Ljungqvist, Nanda and Singh (2003) ¹⁴ present a model in which they distinguish between the two types of investors, and they present these definitions:

The retail investors are small, unsophisticated investors who are prone to episodes of optimistic or pessimistic «sentiment» about the stock market, especially IPOs, where sentiment denotes incorrect beliefs about the fundamental value of an asset arising from treating noise as relevant information (Black (1986)).

The second type of investor holds beliefs that correspond to an unbiased estimate of the issuing firm's future prospects. Ljungqvist, Nanda and Singh (2003) regard institutional investors as belonging to this category.

Ritter and Welch (2002) advance that institutions are naturally blockholders, potentially capable of displacing poorly performing management. That is why, this type of investors is favoured in IPO allocations, and they are able to change and to improve the management and the performance of the IPOs in which they invest.

Booth and Chua (1996), Brennan and Franks (1997), Mello and Parsons (1998), and Stoughton and Zechner (1998) all point out that underpricing creates excess demand and thus allows issuers and underwriters to decide to whom to allocate shares. Surely, they will favour the institutional investors for the advantages advanced from which they can take profit. These institutional investors will monitor the IPO firm's management, creating a positive externality.

Supporting the argument of favoured institutional investors in IPO allocations, Ljungqvist, Nanda and Singh (2004), advance another explanation. They show that value to an issuer is maximized by underwriters allocating IPO shares to their regular (institutional) investors for gradual sale to sentiment investors as they arrive in the market over time. Regulars maintain stock prices by holding IPO stocks in inventory and restricting the availability of shares.

Underpricing emerges as fair compensation to the regulars for expected inventory losses arising from the possibility that sentiment demand may cease.

But these advantages such improving the firm's performance and management may not be of primary importance for some companies for corporate control considerations.

Retaining control for some managers is the primary and the most important objective. When they are obliged to go public, they try to protect their private benefits by favouring the individual investors rather than institutions, since retail investors are not able to buy large volume of IPO shares and they end up holding small parts. The managers seek to avoid allocating large volume of shares for few important investors (institutions), in order to protect the control of the management. Greater ownership dispersion implies that the incumbent managers benefit from a reduced threat of being ousted, because the most important part of the IPO firm still retained by its managers and the other part is dispersed between many retail investors.

Going further in this point of view (control retain), Brennan and Franks (1997) argue that managers deliberately underprice the IPO shares, since underpricing generates excess demand. This gives the managers the opportunity to protect their private benefits by allocating shares strategically when taking their company public, favouring the retail investors and then a greater ownership dispersion. So retaining control can be an explanation for underpricing anomaly for some companies for which managers consider their private benefits in the first place.

II- Literature review of behavioral explanations:

As I said before, the tendency of behavioral approach and investor sentiment is not a new field not again discovered, the investor sentiment was introduced earlier in the 1990's by Welch who presented the informational cascade theory. But the Behavioral Approach has sparked the academics' attention, has intrigued more and more researchers and has taken all its impetus in this decade. The studies and investigations are more and more numerous trying to explain the short run IPO anomaly by investor sentiment and behaviour. The research effort is continuing since the behavioral approach seems to be a very promising field to resolve the short run IPO phenomenon that is still a puzzle since the other theories and explanations advanced earlier failed to resolve it.

I begin by presenting the first explanations and the first findings that were observed to explain the short run phenomenon by the investor sentiment.

II-1\ Informational cascades:

If potential investors pay attention not only to their own information about a new issue, but also to whether other investors are purchasing or not and they attempt to judge the interest of other investors, according to Welch (1992) ¹⁵, bandwagon effects or also known as information cascades may develop:

Later investors can condition their bids on the bids of earlier investors, rationally disregarding their own information. If an investor sees that no one else wants to buy, he may not buy even when he possesses favourable information about the issue: Later investors abstain because earlier investors abstain. In order to prevent this situation from happening, an issuer may have to underprice the IPO to induce the first few potential buyers, and later induce a cascade in which all subsequent investors want to buy irrespective of their own information. They will imitate the first potential investors even if they have unfavourable and negative information about the IPO. Since there are some investors interested in the IPO and bought, other investors will be also interested in the offering and will request shares even if they think that it is unattractive, their opinion will change and they will think that it is a hot offering.

So, if an issuer succeeds the first sales, he is sure that later investors are encouraged to invest in his offering whatever their own information. Conversely, if an issuer fails the first sales to the earlier investors, this will dissuades later investors from investing irrespective to their own information.

In support for this informational cascade explanation, Amihud, Hauser and Kirsh (2001) find that IPOs tend to be either undersubscribed or hugely oversubscribed, with very few offerings moderately oversubscribed.

II-2\ The prospect theory:

Prospect theory, developed by Kahneman and Tversky (1979), asserts that people focus more on changes in their wealth compared to the level of their wealth.

Loughran and Ritter (2002) apply prospect theory of Kahneman and Tversky (1979) to IPO market to argue that issuers are more tolerant of excessive underpricing and that they accept underpricing more than necessary if they simultaneously learn about an aftermarket valuation

¹⁵ Welch's cascades model remains one of the least explored explanations of IPO underpricing.

that is higher than expected. They are more concerned with an increase in their future wealth rather than instantaneous and immediate profits. Loughran and Ritter argue that the issuing firm's executives bargain less hard for a higher offer price, if they are seeing a personal wealth increase relative to what they had expected based on the file price range that they have fixed, and this is when the price is revised upwards during the bookbuilding process.

Loughran and Ritter explain more this theory by the fact that insiders of IPO companies consider not only the shares they sell in the IPO, but also those they retain which benefit from the large initial price run-up.

Combining prospect theory reference point with Thaler's (1980, 1985) notion of mental accounting, Loughran and Ritter argue that issuers fail to «get upset» about leaving millions of dollars «on the table» in the form of large first-day returns because they tend to sum the wealth loss due to underpricing with the (often larger) wealth gain on retained shares as prices jump in the after-market. Underpricing and the positive initial return is perceived as a wealth loss under the assumption that shares could have been sold at the higher first-day trading price. If the perceived gain exceeds the underpricing loss, the decision-marker (issuer) is satisfied with the IPO underwriter's performance at the IPO. He is satisfied even if he leaves large amount on the table and he accepts the underpricing even if it is higher than necessary since he can compensate this loss by a larger wealth gain. He can benefit from the first day price run-up since he will sell the shares retained at a higher price which will generate a large profit compensating the loss undergone by underpricing.

Ljungqvist and Wilhelm (2004) use the structure suggested by Loughran and Ritter's (2002) behavioral perspective to test whether CEOs deemed «satisfied» with the underwriter's performance according to Loughran and Ritter's story are more likely to hire their IPO underwriters to lead-manage later Seasoned Equity Offerings. Controlling for other known factors, IPO firms are less likely to switch underwriters for their SEO when they were deemed «satisfied» with the IPO underwriter's performance. This result confirms what has been advanced by Loughran and Ritter (2002) about issuers who are not upset about leaving money on the table and about the tolerance of underpricing believing in a future wealth increase.

Underwriters also appear to benefit from behavioral biases in the sense that they extract higher fees for subsequent transactions involving «satisfied» decision-makers.

So, for the literature review of behavioral explanations, I begin by presenting the findings of the first researchers who tried to introduce the behavioral approach and the investor sentiment, the informational cascade introduced by Welch (1992) and the application of the prospect theory developed by Kahneman and Tversky (1979) to IPO market.

These two theories were the first explanations of the short run IPO anomaly advanced based on behavioral approach and investor sentiment. The research effort is continuing and much investigations and studies are advanced later by the researchers who believe in the behavioral approach and in the importance of sentiment to explain and to clarify the underpricing anomaly and to resolve this persistent puzzle over decades in the IPO market.

The behavioral approach has sparked much academic attention and the research effort has provided numerous analytical advances and empirical insights trying to explain the first day price run up and the underpricing anomaly by the investor sentiment.

The effect of sentiment investors has been advocated particularly strongly for initial public offerings. The short run IPO puzzle could be due to overenthusiasm and over optimism among investors who may be less than perfectly rational. And to study the effect of this type of investors on IPO market and on the short run IPO anomaly, the sentiment should be measured which is a very hard to not say that is an impossible task.

Many proxies have been advanced and proposed in the literature to measure the investor sentiment, or at least to approach it since by its subjective and individual characteristics, sentiment can not be observable, and then many results have been found. We will see all these proxies that have been used to measure the investor sentiment and the findings that have been reached.

So, in the paragraphs that follow, the notion of investor sentiment will be studied by researchers in more detail and will take another direction based on the over optimism and over enthusiasm of sentiment and irrational investors.

II-3\ Investor sentiment by Ljungqvist, Nanda and Singh (2004):

We can say that the importance of investor sentiment was introduced and analyzed in the context of the underpricing phenomenon for the first time by Ljungqvist, Nanda and Singh (2004) in their article «Hot markets, Investor sentiment and IPO pricing». The work of these authors is considered the first paper to model an IPO company's optimal response to the presence of sentiment investors. They give a response to the question: what should a profit-maximizing issuer do in the presence of exuberant investor demand and short sale constraints?

They show that underpricing, long-run underperformance and hot-issue markets can be explained by the presence of sentiment investors.

The sentiment investors' behaviour and stocks demand can not be foreseen. All is conducted by feelings and individual beliefs, and the hot market period, characterized by the presence of these sentiment investors who are exuberant and presenting excessive sentiment demand, can end prematurely at any time. And because the hot market can end prematurely, carrying IPO stock in inventory is risky. The optimal mechanism for the issuing firm in the presence of sentiment investors involves underwriters allocating stock to «regulars» to hold it for gradual sale to sentiment investors and to bear the risk of inventory losses in the place of issuers¹⁶.

Regulars hold IPO stock in inventory to maintain stock prices by restricting the availability of IPO shares. And they bear the risk of expected inventory losses arising from the possibility that sentiment demand may cease and from the non-zero probability that the hot market will end before all inventory has been unloaded. Then, they are left with IPO stocks and there are no sentiment investors to buy these shares. From the point of view of these authors, underpricing is necessary, it emerges as fair compensation to the regulars for expected inventory losses arising from the possibility that hot market period may end prematurely. As regulars are running the risk of inventory losses in the place of issuers, they require the stock to be underpriced, as compensation for them. They take profit from this underpricing as a reward. But generally, a regular will invest in IPOs only if he does not expect to lose as a consequence. Then underpricing is a necessary cost for the issuing firms emerging from the fact that sentiment investors' behaviour and demand can not be foreseen, and issuers need the regulars to take on this risk in their place and to hold inventory if the demand is small.

So, the optimal selling policy, from the issuer's point of view, usually involves gradual sales. Such staggered sales can be implemented by allocating the IPO to cooperative regular investors who hold inventory for resale in the after-market. These regulars require compensation for bearing the risk of expected inventory losses if the hot market ends prematurely and this compensation is IPO underpricing.

Underpricing is then explained by the presence of irrationally exuberant and sentiment investors whose demand may cease at any time.

¹⁶ The risk premium explanation (1^st section): underpricing is compensation for underwriters and regular investors for bearing the risk of IPO market crashing and expected inventory losses.

If all the investors are rational and their demand is regular, then it is not necessary to appeal for regular investors to hold stock inventory.

It is surely undeniable that «Hot markets, Investor sentiment and IPO pricing» by Ljungqvist, Nanda and Singh, is considered as the first work that analysed the importance of investor sentiment in the context of underpricing anomaly. The work of these authors is surely the first paper to model an IPO company's optimal response to the presence of sentiment investors. However, these authors did not use a measure for the investor sentiment in their model or a valuation for this explanation of underpricing to check its relevance and its statistical and economical significance. They introduced the investor sentiment as an explanation for underpricing anomaly but did not try to give a valuation to investor sentiment to use it in an econometrical model to concretize their viewpoint and their explanation. They surely presented important ideas and explanations in their article, but did not quantify and concretize their findings and their explanations using real data.

In the following paragraphs, I present a literature review of works and papers that introduce and analyze the investor sentiment as an explanation for the short run IPO anomaly and that try to value the investor sentiment used in a model.

II-4\ The use of Grey Market Data:

Cornelli, Goldreich and Ljungqvist (2004) try to take advantage of the existence of a grey market in Europe to construct a model on European Initial Public Offerings to look at whether the presence of sentiment investors affects prices in the post-IPO market and whether investors' sentiment can explain and be considered as a driver and primary determinant of underpricing anomaly.

Before going in details and presenting the findings of these authors, I should begin by defining the grey market and presenting some assumptions in their model to understand their results.

The Grey Market is a when-issued market for shares to be issued in an IPO (the definition as given by the authors).

Most European countries have grey markets for IPOs. It is a Pre-IPO market taking place at the same time as institutional bookbuilding (before setting the issue price definitively). In this market, retail and smaller investors speculate on the post-IPO share price, so these investors are trading IPOs before they are listed on the stock market. The prices at which they speculate are considered as post-IPO share prices and that is why the grey market is considered as an opportunity to observe the post-IPO prices and the opinion of small investors and to foresee their behaviour in the after market. The grey market is an opportunity to foresee what it will be in the IPO after-market. It is the unique opportunity to observe the valuation of the investors who will be buying shares in the aftermarket and to measure the expectations of sentiment investors directly. It is also beneficial to the issuers to increase the offer price and then the IPO proceeds. In the U.S, regulations prohibit the existence of Grey Markets and the speculation of IPO shares before they are listed on the stock market.

The grey market investors may overweight the relevance of their information, so the authors allow for the possibility that grey market investors are sentiment investors but they do not impose it.

When the publicly observable grey market price is high relative to fundamental value, bookbuilding investors resell shares to these smaller investors in the aftermarket, so the offer price and the aftermarket price will be close to the grey market price. In contrast, when the grey market price is low, the offer and aftermarket prices will be close to fundamental value. Because the fundamental value is unobservable, econometricians usually take the midpoint of the initial indicative price range as a proxy for fundamental value.

The dataset consists of 486 companies which went public in twelve European countries between November 1995 and December 2002 and for which there exist grey market prices. To give a valuation to the investor sentiment, the authors use the normalized last grey market price before the issue price was set: P_GM / P _mid (the midpoint of the initial indicative price range). They find that there is a positive correlation between grey market price, issue price and aftermarket price. The grey market price is more correlated with the issue and the aftermarket prices when the grey market price is high, although there is a positive correlation even when the grey market price is low. So this positive correlation confirms the fact that the presence of sentiment investors has an impact on after market prices and so on the underpricing anomaly.

In conclusion, Cornelli, Goldreich and Ljungqvist (2004) find that high pre-IPO prices, which indicate overly optimistic investors, are a good predictor of high initial returns during the first trading day. And then, the high investor sentiment valued by high grey market prices, is a primary driver and determinant of the short run IPO anomaly.

When small investors are excessively optimistic, they are willing to pay a price above the fundamental value, resulting in a high aftermarket price (foreseen through the grey market prices), and so in underpricing. But when these sentiment investors are pessimistic about an issue, their opinion has less of an effect on the aftermarket and issue prices that will be close to the fundamental value and not to the grey market, which suggests that the underwriter and sophisticated investors can identify sentiment investors. Thus, they consider the opinion of sentiment investors biased, and they take it into account only when they can profit from it, by selling shares to them in the aftermarket.

Cornelli, Goldreich and Ljungqvist (2004), are not the first authors using the grey market in their study to value the investor sentiment trying to explain the short run IPO anomaly. There are other earlier studies that are complementary to the findings of these authors. For example, Dorn (2003) finds that the volume of grey market trading among the customers of a German retail brokerage is correlated with high initial returns and low long-run returns. This can be viewed as further evidence that participation by smaller investors in the grey market can be interpreted as sentiment. Besides Dorn (2003), two other papers study the grey market in Germany: Loffler, Panther and Theissen (2002) document that grey market prices are unbiased estimates of first-day prices. Aussenegg, Pichler, and Stomper (2003) also find that IPO offer prices are related to prices in the grey market but they show that the coefficient is smaller than one. Finally, Pichler and Stomper (2004) model the interaction between bookbuilding and the grey market when grey market investors have similar information to bookbuilding investors. They ask whether the existence of a grey market helps or hinders information aggregation in bookbuilding. In contrast, Cornelli, Goldreich and Ljungqvist (2004), introduce a class of investors who have different information from bookbuilding investors, in order to explain certain IPO phenomena and to show how these (possibly biased) investors affect prices.

II-5\ The use of market conditions to value investor's sentiment:

François Derrien (2003) explores the impact of investor sentiment on the pricing and aftermarket behaviour of IPOs, using a sample of 62 initial public offerings realized on the French stock exchange between 1999 and 2001. He tests a model in which the first day closing price of IPOs and then initial returns and underpricing are depending on the information about the intrinsic value of the company (revealed by institutional investors and it is private information) and on investor sentiment. These French offerings used a unique IPO mechanism: a modified book-building procedure in which a fraction of the IPO shares is reserved for individual investors. This mechanism was created in 1999 and is unique to the French stock exchange: the Offre à Prix Ouvert (OPO).

To value the investors' sentiment, Derrien focus on these individual investors' demand which is related to the market conditions. These individual investors are typically small and uninformed, that is why they are considered as sentiment investors and their demand is a direct measure of sentiment.

The demand for IPO shares submitted by individual investors varies considerably and is strongly and positively related to a measure of market conditions prevailing at the time of the offering. When these conditions are favourable, the investor sentiment is high and individual investors seem to be overly optimistic, presenting a large demand for IPO shares. This demand surely influences IPO prices and then the underpricing phenomenon to a large extent. It leads to high IPO after market prices and then to large initial returns: the more favourable market conditions, the larger the individual investors' demand, the higher the investor sentiment and the higher after market prices, keeping everything else constant.

Loughran and Ritter (2002) and others have provided indirect empirical evidence of this short run IPO phenomenon using various measures of market conditions as proxies for investor sentiment. For example, to measure the market conditions, the return of the industry index the IPO company belongs to in the recent period preceding the offering can be used. Loughran and Ritter use the three week period. But in this paper, the author uses a direct measure of the investor sentiment, namely the demand for IPO shares posted by individual investors in the French IPOs used in data, these investors are considered as sentiment investors and so their demand is a direct measure of sentiment.

Derrien and Womack (2003) show that the initial returns on IPOs in France in the 1992-1998 period were predictable using the market returns in the three-month period preceding the offerings. Using U.S. data, Loughran and Ritter (2002) and Lowry and Schwert (2003) obtain similar results: in that the initial returns in the first day of trading for IPOs are predictable using the market conditions prevailing at the time of the IPOs or at a recent past.

Bradley and Jordan (2002) include the 1999 `hot issue' market in their sample and find that more than 35% of initial returns can be predicted using public information available at IPO date. However, Lowry and Schwert (2003) find that the effect is economically small, the relation between underpricing and market conditions (market returns) is statistically but not economically significant.

II-6\ Discount on closed-end funds as proxy for investor sentiment:

The proxies used in empirical works investigating in short run IPO anomaly to value the investor sentiment and to measure its impact on underpricing anomaly are numerous. One of the most important proxies often used by researchers is the discount on closed-end funds. The choice by many authors of the discount on closed-end funds as a proxy of investor sentiment in their works is based on the findings of Lee et al. (1991).

Lee, Shleifer and Thaler (1991) find that both closed-end funds and small stocks are mostly held by individual investors who tend to be noise traders, suggesting that they are more likely than large stocks to be affected by investor sentiment. Lee et al. find that almost a quarter of the variation in monthly returns on the smallest decile of firms is explained by discount changes, even after controlling for general market movements. Their findings suggest that when investor sentiment is higher, individual investors pay relatively more for closed-end funds, and the discount is smaller.

Lee, Shleifer and Thaler (1991), and Lowry (2002) document that «hot issue» periods and «hot IPO market» characterized by high sentiment and over optimism among individual investors coincide with low discount on closed-end funds, their measure of noise traders' optimism and sentiment. They also show that the annual number of IPOs is negatively related to the closed-end fund discount which they argue is a measure of retail investor sentiment: the smaller the closed-end funds discount, the higher the sentiment and optimism among individual investors and the higher the number of IPOs. Over optimistic and enthusiastic investors are more accepting to invest in new issues, and companies are encouraged to take advantage of this opportunity to go public and to offer their shares in the stock market.

Based on these findings, many researchers are convinced that discount on closed-end funds is a relevant proxy that can be used to value the investor sentiment and to study its impact on IPO underpricing anomaly that still a puzzle right now. They use it in their econometrical models.

Ivanov and Lewis (2008) ¹⁷ try to identify the determinants of market-wide issue cycles for initial public offerings (IPOs) using an autoregressive conditional count model. They consider whether IPO volume is related to business conditions, investor sentiment, and time variation in adverse selection costs caused by asymmetric information between managers and investors.

¹⁷ «The determinants of market-wide issue cycles for initial public offering» (2008)

To value the investor sentiment, they use the quarterly first-difference in the index of closed-end fund discounts as proposed by Lee et al. (1991).

II-7\ Other proxies and empirical results:

The list of proxies used by researchers to value investor sentiment and to study its impact on underpricing anomaly is very long. I presented the most important proxies in the previous paragraphs: grey market prices, market conditions, demand submitted by individual investors and discounts on closed-end funds.

In the present paragraph, I try to give a brief sight on the other proxies that are used by researchers and the main results found:

Rajan and Servaes (2003) model two different types of irrational agents, feedback traders and sentiment investors. They proxy for investor sentiment using the market-to-book ratio and find that it correlates positively with first-day returns and negatively with long-run returns. So there is a positive correlation between investor sentiment valued through market-to-book ratio and underpricing phenomenon.

Other studies have looked at who owns the shares in the aftermarket: for example, Ofek and Richardson (2003) show that high initial returns occur when institutions sell IPO shares to retail investors on the first day.

Similarly, Ben Dor (2003) looks at the level of institutional ownership shortly after the IPO and finds that high institutional ownership forecasts higher returns in «hot» markets.

So, these authors use another proxy to investor sentiment: the holdings of large (institutional) investors. Since these institutional investors have the possibility to sell the IPO shares to sentiment investors in the aftermarket in the first day of trading¹⁸.

The proxies for sentiment are numerous like the holdings of large (institutional) investors, put-call ratios, trading volume, market-to-book ratio, market conditions, grey market prices, closed-end fund discounts, and the list is not exhaustive. There is an ongoing inconclusive debate about the effectiveness and explanatory power of many investor sentiment measures (Qui and Welch, 2004).

¹⁸High institutional ownership shortly after the offering is indicating the presence of high and optimistic sentiment among individual investors, and so institutional investors (rich-information investors) can easily sell the shares to these sentiment investors and making large incomes and profits.

Oehler, Rummerand Smith (2005) in their article «IPO Pricing and the Relative Importance of Investor Sentiment-Evidence from Germany» try to explain the persistent pattern of high initial returns during the first trading day. They focus on the importance of investor sentiment and of information gathered by the underwriter before the start of the bookbuilding process. From their viewpoint, there are only two different but not mutually exclusive scenarios which could lead to the observable pattern of high initial returns at the first trading day. First, it could be possible that the offer price is set too low due to ex-ante uncertainty about the true market value of the IPOs. Second, the offer price might be on average at a «fair value» but demand for new issues is overly high and therefore generating the observed high initial returns. They conclude from the estimations realized that underpricing is mainly influenced by investor sentiment and, therefore, by the uncertainty about the demand of potential investors (the 2^nd scenario), and less by ex-ante uncertainty about the underlying firm's value. Investor sentiment has the dominant influence on the fluctuations of initial returns and IPO underpricing anomaly.

To value the ex-ante uncertainty about the potential demand and the investor sentiment, they focus first on the bookbuilding price range and the subscription period which are set by the underwriter after observing the potential demand for the stock to be issued, and second on the explanatory power of pre-IPO trading, stock market performance prior to the issue and the usage of the so-called Greenshoe option.

In Europe, the length of the bookbuilding period and the width of the bookbuilding range are set after a pre-marketing period. During this time span IPO research from sell-side and buy-side analysts is produced and distributed by syndicate members to institutional clients (Jenkinson, Morrison and Wilhelm, 2005). As a result, the length of the subscription period and the width of the bookbuilding range are good indicators of how strong the underwriter expects potential demand to be. The longer the subscription period and the wider the initiative price range, the more uncertain is the underwriter about possible success and the higher is the uncertainty about potential demand inducing the importance of investor sentiment which plays a dominant role in influencing the demand. This view is supported by the argument of Jenkinson, Morrison and Wilhelm (2005) who assert that ceteris paribus less available information will lead to an increase in the bookbuilding range.

The authors run regressions with the indicative price range and the subscription period as dependent variables, and find that positive sentiment for IPOs reduces uncertainty about potential demand and then reduces the subscription period and bookbuilding price range.

To value investor sentiment, Oehler, Rummer and Smith (2005) use many proxies: buy-and-hold on the market prior to the IPO which covers a period of 30 days (market movement), Grey market trading prices, business climate over the period of one month prior the IPO, usage of the Greenshoe option which consists in issuing extra shares due to excess demand, and issue volume (number of issued shares including the exercised Greenshoe multiplied by issue price). The authors come to the conclusion that investor sentiment and demand have the dominant influence and impact on initial returns on the first trading day and thus using many proxies for sentiment which they find very significant in their results. It is changes in investor sentiment which have the dominant influence on the fluctuations of initial returns.

They also find another important result that the length of the subscription period and width of the bookbuilding range have negative effects on underpricing.

Aggarwal, Krigman, and Womack (2002) introduce another explanation to first day IPO anomaly. They relate the aftermarket price path to momentum traders. They focus on the role of research analysts and the media in creating momentum and in inducing high investors' sentiment. So to value the investors' sentiment, they focus on the role of media and research analysts: when they are important and favourable to the issuing company and to the IPO, this induces high and positive sentiment among individual investors, and then initial returns are higher and underpricing is more important.

Section 3- The model and empirical implications

The effect of sentiment investors has been advocated particularly strongly for the Initial Public Offerings' patterns, since by definition, IPO firms have no prior share price history and tend to be young, immature, and relatively informationally opaque. So they are very sensitive to the state of mind of the investors and to investors' feelings and beliefs. A large literature over the past decades, both theoretical and empirical, has attributed the IPO pattern and the short run IPO puzzle to the presence of this type of investors.

The works of researchers who are interested in the behavioral approach and in the investor sentiment explanation to clarify the short run IPO puzzle and to understand the IPO underpricing anomaly are numerous. The list is long and is in continuously growth.

I presented in the previous paragraphs the most important studies and papers that have considered the sentiment explanation and the investor's behaviour as the most convincing and relevant driver and determinant of IPO underpricing. I presented the application of the behavioral and sentiment approach to clarify and to understand the IPO patterns by numerous researchers to shed light on the importance of the sentiment investors in the IPO market.

I continue in the same direction of the behavioral approach and the investor sentiment explanation to the short run IPO puzzle, believing in the relevance of this explanation to clarify and to understand the underpricing phenomenon and in the importance of sentiment investors in the IPO market to explain its anomalies. I presume that there are periods when capital providers (individual and institutional investors) become irrationally enthusiastic about new investment opportunities. In this thesis, we are in the case of the new Initial Public Offerings. They become irrationally overly optimistic about the IPO market. This over optimism is based on the favourable recent history of the IPO market and the positive initial returns recently observed for the new issues. These sentiment investors project this positive tendency on the recent future, believing that the IPO market will continue in the same direction in the recent future for sure. This period is known as «Hot IPO market». The investors' exuberance translates into periods of high demand for Initial Public Offerings, called «sentiment demand» since it is based on sentiment and beliefs. It is far from a perfectly rational demand based on fundamentals, and these sentiment investors are willing to pay higher prices to have IPO shares. As I presented in a previous paragraph, the IPO market is cyclical. There are periods of «hot market», but there are also periods of «cold IPO market» and sentiment investors become overly pessimistic about IPOs.

There are two types of investors: there are individual or retail investors and institutional investors who are more informed and important clients. A question may arise: what type of investors is driving first day closing prices and the underpricing anomaly?

The empirical works and studies investigating in IPO issue activity and particularly in the short run IPO anomaly, using the «traditional explanations» based on the asymmetric or symmetric theories, or using the behavioral approach and the sentiment explanation, draw numerous and different conclusions.

The biggest limitation is that none conduct a comprehensive analysis that evaluates all of the explanations in a systematic manner. None try to present in the same model all the explanations: asymmetric, symmetric and behavioral, to determine which is the most relevant and convincing to explain the short run IPO puzzle, and none try to distinguish between the sentiment of the two types of investors: individual and institutional, to verify which type of investors exactly is driving the underpricing anomaly.

In this study, an effort to regroup the most important explanations that have been advanced in the same model to determine which of these explanations characterizes best the data in the context of a unified framework and model. I introduce the three theories:

And since there are two types of investors in the IPO market: individual and institutional investors, one of the goals of this study, is to identify which type is driving the first day closing prices and underpricing by distinguishing between the individual investors' sentiment and the institutional investors' sentiment.

To this end, I use direct measure of sentiment for the two types of investors and this represents another contribution of this work.

The aim in this thesis is to show how sentiment investors and their irrationally overly optimism can lead to a first day price run up and therefore to underpricing, and can explain this short run anomaly with a distinction between the two types of investors in the IPO market to clarify and to understand which type is more conducting the short run IPO puzzle, and using a direct measure of sentiment for each category of investors: the investor sentiment index.

I- The model and explanatory variables:

I-1\ The model:

As I presented in the previous sections, there are numerous explanations advanced to understand and to clarify the short run IPO anomaly. These explanations can be classified in three main categories:

Ø Explanations based on asymmetric information between the key IPO parties and have been considered the most convincing explanations for decades.

Ø Explanations asserting the informational transparency and lucidity and asserting the IPO market efficiency.

Ø And explanations based on Behavioral Approach and on investors' sentiments and beliefs.

I regroup the three theories advanced in the same model to determine which of these explanations characterizes best the data in the context of a unified model and presents the most relevant and reliable explanation to underpricing phenomenon. I present every theory by one or more indicators and determinants.

v For the informational asymmetry theory, I use the firm quality as a determinant with many proxies: the Overhang Ratio, Venture Capital backed, Underwriter reputation and R&D intensity.

v For the theory asserting the IPO market efficiency, I use the risk determinant: age of the issuing firm, firm size (sales and assets), firm profitability, ROA and the issue risk (if the firm operates in a technological and risky sector). I use also the issue size (Ln (expected proceeds)), and the issuer bargaining power.

v Finally, for the behavioral approach, I use direct measures of investor sentiment, and I distinguish between the two types of investors: the individual investor sentiment and the institutional investor sentiment. For this, I use the individual investor sentiment index (AAII) and the institutional investor sentiment index (II).

Underpricing = a₀ + a₁Underwriter Reputation Dummy + a₂Overhang + a₃R&D Intensity + a₄VC Dummy + a₅Ln (1+age) + a₆Ln (assets) + a₇Ln (sales) + a₈Firm Profitability + a₉ROA + a₁₀Issue Risk Dummy + a₁₁Ln (expected proceeds) + a₁₂Insiders Ownership + a₁₃Institutional Ownership + a₁₄Blockholders Ownership + a₁₅Individual Investor Sentiment + a₁₆Institutional Investor Sentiment+ a₁₇Time Dummy + å_i

I-2\ The explanatory variables:

In this paragraph, I try to justify the use of these determinants and explanatory variables in the model for each theory and to present briefly the previous results reached by researchers that used the same indicators and control variables in their models.

I-2-1\ Informational Asymmetry Theory:

The issuer is the most and best informed party about the issuing firm quality. There are many proxies of firm quality that can be employed . I introduce the most important ones often used by researchers:

The percentage ownership retained by insiders serves as a signal for firm quality, and two different relations are observed between firm quality and underpricing with two different explanations. The first relation is that, for high overhang ratio and so for high quality issues, the issue price is lower a mean to demonstrate their quality and to distinguish themselves from the pool of low quality issuers, then a higher level of underpricing is observed. In the same direction, we can justify the use of overhang ratio and its importance to underpricing phenomenon by the fact that only the shares sold to the public in the IPO are undervalued. The shares retained by insiders are valued at market. Thus, for a given degree of underpricing, the economic cost per retained share (the dilution) declines as overhang rises. Because the cost of underpricing to the issuer declines as overhang rises, it is natural to conjecture that firms with greater overhang will have greater underpricing and we find the same positive relation between overhang ratio and underpricing.

The second relation is completely different in that issuing firm quality is negatively correlated with underpricing. Issuers with high quality firms bargain for higher offer prices for their IPOs and then a lower degree of underpricing is observed at the first day of trading.

R&D expenditures are the intangible investment most extensively researched in economics, accounting and finance, they have to be disclosed in the corporate financial reports. R&D contributes to information asymmetry and Guo, Lev and Shi (2006) consider R&D activities as the major source of asymmetry. Pre-IPO R&D intensity of the issuer is strongly and positively related to the first day underpricing. R&D-intensive firms are often undervalued by investors. That is why R&D intensive issuers can not set a high offer price for their IPO. Besides, they are more willing to forgo money on the IPO table than are no R&D issuers, because they expect to recoup money left on the table by subsequent issues of seasoned stocks when the market realizes over time the positive outcomes of their R&D. So the relation between R&D intensity and underpricing is positive but many studies find no relation between underpricing and SEO, which refute the last point of recouping money left on the table by SEO in the future.

Another relation can be found between R&D intensity and underpricing, mainly since no relation between underpricing and SEO: R&D intensive issuers believing in the high quality of their firms and in the importance of their R&D and its positive outcomes in the recent future require high offer prices for their IPOs which induces a negative relation between R&D intensity and underpricing.

Ø Venture Capital backed: a dummy variable taking the value of one if the issue firm is backed by a venture capital and zero otherwise.

Since venture capitalists have expertise in particular industries, they are expected to make superior investments relative to other investors. In essence, VCs certify the quality of an IPO and their presence signals that asymmetric information is relatively low for this issue and that the issuing firm quality is high. So, issuers can bargain for a high offer price which declines underpricing. Another explanation is found, that VC backed firms face larger underpricing, since high quality issuers will continue demonstrating the firm quality by throwing money on the IPO table and not requiring high issue prices. However, in many studies using the venture capital as a determinant of firm quality, this variable is found insignificant. A question may arise: «Is Venture Capital backing really a signal of firm quality and an explanatory variable to underpricing anomaly?». I introduce this variable in the model to verify these results.

Ø Underwriter Reputation: a dummy variable taking the value of one if the lead underwriter has a rank = 8 (zero otherwise).

Some issuers use it as a signal of high quality and want to hire a prestigious underwriter, since by agreeing to be associated with an offering, prestigious intermediaries «certify» the quality of the issue. When an issuer chooses a prestigious underwriter for the book-building mechanism, he sets a low offer price conducting a high underpricing on one hand, as compensation to the underwriter and on the other hand, he is sure about full subscription, he is concerned by quantity rather than price and a lower offering price increases the probability of full subscription. Then, there is a positive relation between underwriter reputation and underpricing.

But, there is another explanation completely different: when an issuer chooses a prestigious underwriter who certifies the high quality of the issue, he can bargain for a higher offer price and then a lower level of underpricing is observed: A negative relation between underwriter reputation and underpricing.

I-2-2\ Theory asserting informational symmetry and IPO market efficiency:

I use the main characteristic of Initial Public Offerings, «risk» related to technological or valuation uncertainty and I use many measures:

Ø The issuing firm size reflects the valuation uncertainty risk: Ln (sales), Ln (assets).

Ø The issue risk, a dummy variable taking a value of one if the firm operates in a risky industry and zero otherwise, a variable reflecting technological risk which induces a valuation uncertainty.

Ø The age of the issuing firm: Ln (1+age) (age is number of years since the firm's founding date to the date of going public and the date of introduction in IPO market).

Ø I use also some financial data as proxies for valuation uncertainty risk: firm profitability and ROA.

Almost the empirical studies introducing risk find a positive relation between risk and underpricing: Riskier firms set a low offer price to incite investors to participate in the IPO market and to buy the IPO risky shares, and a high offer price will dissuade them. However, a study of Bartov, Mohanram and Seethamraju (2003) reports that there is no correlation between risk and underpricing. This finding refutes the prior researches and results about the suitability of the risk as an explanation to the underpricing anomaly. And I use this variable with many measures to verify if risk can be considered as a relevant explanation to short run IPO anomaly.

I use another determinant of the symmetric information theory: the issue size calculated as the natural logarithm of the net expected proceeds. Previous researches report a negative and statistically significant relation between size and offer price, so we can talk about a positive and significant relation between issue size and underpricing (Cornelli, Goldreich and Ljungqvist (2004)). Other researches find a negative relation between issue size and underpricing, explaining this by the fact that sizable firms are generally less risky than those making smaller issues, then issuers can bargain a higher offer price conducting a lower level of underpricing.

I also introduce the issuer bargaining power as a variable in the theory asserting the informational transparency and lucidity, with three proxies to have an idea about the ownership structure: insiders' ownership, institutional ownership and blockholders' ownership prior to the offering.

The main result found in the earlier studies, the greater is the ownership concentration, the greater is the issuing firm's bargaining power, the higher is the offer price and the lower is the first-day return and vice-versa. But Loughran and Ritter (2004) argue that this argument has little support as an explanation for underpricing.

I introduce another control variable «Time dummy»: a dummy variable taking the value of 1 if the IPO date is after July 2007, and zero otherwise. This variable is used to control the beginning of financial crisis period and its impact on underpricing.

I-2-3\ Investor sentiment and Behavioral approach:

The list of proxies used by researchers to value investor sentiment and to study its impact on underpricing anomaly is very long. The most important proxies often used in researches and empirical studies are: grey market prices, market conditions, demand submitted by individual investors, discounts on closed-end funds and market-to-book ratio. Researches and empirical works report that the higher is the investor sentiment and his optimism, the higher is the level of underpricing.

Ø The first measure is obtained from the survey data of the American Association of Individual Investor (AAII): In July 1987, AAII started conducting a weekly sentiment survey asking for the likely direction of the stock market during the next six months (up, down or the same). The participants are randomly chosen from approximately 100,000 AAII members, only subscribers to AAII are eligible to vote and they can only vote once during the survey period.

Each week, AAII mails the questionnaires, and members fill them out and return them via US mail. Each week AAII collects responses from Friday to the following Thursday, compiles the results based on survey answers and labels them as bullish, bearish or neutral and reports the results on Thursday or Friday. Since this survey is targeted towards individual investors, it is primarily a measure of individual investors' sentiment.

Ø The second measure is obtained from the survey data of Investors Intelligence (II), an investment service based in Larchmont, New York. II compiles its sentiment data weekly by categorizing approximately 150 market newsletters since 1964. Data is based on a survey of investment advisory newsletters. To overcome the potential bias problem towards buy recommendations, letters from brokerage houses are excluded. Newsletters are read and marked starting on Friday each week. The results are reported as percent bullish, bearish, or neutral on the following Wednesday. Since authors of these newsletters are current or past market professionals, the II series is interpreted as a measure for institutional investors' sentiment.

II- Data Description:

The sample consists of American Initial Public Offerings underpriced in the first day of trading for 2006 and 2007. The years 2008 and 2009 have not been included in the sample due to lack of available information. The data base includes 348 offerings, after excluding the issues having an underpricing less than 5%, the sample is reduced to 225 IPOs, and I require the firms to be available in the U.S Securities and Exchange Commission website because I use the SEC database for the company filings and prospectuses which further reduce the sample to 217 IPOs.

My primary data source for the IPOs is http://www.iposcoop.com database. I have collected the missing information from prospectuses available in the SEC's Electronic Data Gathering and Retrieval (EDGAR) system for IPOs (final prospectuses are identified on EDGAR as document 424B at http://www.sec.gov), and from a number of other sources.

Underwriter rankings are on a 0 to 9 scale. Information comes from the Jay Ritter website, a file presenting the rankings of all underwriters from 1980-2007. All the rankings in this file are integers followed by a 0.1 (1.1 up to 9.1). Loughran and Ritter attach a 0.1 to all rankings so that other researchers can easily distinguish between their rankings and those from Carter and Manaster and Carter, Dark, and Singh, which never end with a 0.1. Because Loughran and Ritter to make this underwriter rankings' file start with the Carter and Manaster (1990) and Carter et al. (1998) rankings, but with some changes, and they make their own methodology to complete rankings for the other years.

To measure the Overhang Ratio, Pre-IPO shares retained are calculated as the difference between shares post offer and shares offered to the public. Information on shares post offer come from the prospectuses and completed from: http://biz.yahoo.com/ipo and http://moneycentral.hoovers.com .

Issuing firm age is calculated as Ln (1+age) and age is the difference between the founding year and the issue year. Information on founding year comes from Jay Ritter website, a file presenting founding years for firms from 1975 to 2008.

Information on expected net proceeds to measure the issue size, on ownership structure (insiders' ownership, institutional ownership, blockholders ownership) and on financial data in the year prior to going public (assets, sales, net income, EBITDA) to measure the different proxies of risk and on R&D expenditures come from the final prospectuses of issuing firms.

If a firm has zero sales, I assign a sales value of 1 thousand $, since in my empirical work I use logarithms, and the logarithm of zero is undefined.

R&D intensity is calculated as the ratio of R&D expenditures (a year prior to offering) to the expected market value. To calculate the expected market value of an IPO, I use the offer price multiplied by the post-issue number of shares outstanding.

For sales and R&D expenditures, if I am not sure whether they are zero or are missing, I treat the value as missing.

Unfortunately, missing values on R&D expenditures (underwriter reputation for 6 firms and sales for 4 firms) reduce the sample to 117 observations. Trying to eliminate R&D intensity as an explanatory variable to increase the number of observations has deteriorated the model, so R&D intensity is a fundamental explanatory variable to underpricing phenomenon.

Issue risk is a dummy variable taking the value of one if the issue is operating in a technological risky sector and zero otherwise. I use the same method of Loughran and Ritter (2004) to categorize IPOs as a technology firm or not.

The two sentiment indicators used in this model: individual investor sentiment and institutional investor sentiment are available on a weekly basis that are compiled from surveys by the American Association of Individual Investors and Investors Intelligence.

If I want to look at the true relationship between underpricing and sentiment, for an IPO at the week t, I should work with the sentiment indicator of the week t-1 because it more accurately reflects the investor sentiment for the week t, but since there is a time lag of 1 week between responses and reporting (as I presented earlier the procedure of reporting these indicators), I should actually work with the sentiment of the same week t to overcome this reporting lag.

Time dummy: a dummy variable taking the value of 1 if the IPO date is after July 2007, and zero otherwise. This variable is used to control the beginning of financial crisis period and its impact on underpricing.

Table 2 provides descriptive statistics of the full sample: The overall mean underpricing of the sample (117 observations) 29.54%, the overall mean for overhang ratio is 9.706, 0.022 for R&D intensity, the overall mean firm age of the sample is approximately 11 years (10 years and 10 months). The overall means of firm profitability and ROA are negative and the overall mean of insiders' ownership is 55.48%. The overall means of investors bullish bearish ratios (individuals and institutionals) are higher than 1, which indicates that individual and institutional investors in this period and for this sample are rather optimistic than pessimistic.

*Table 2. Descriptive Statistics of the Full Sample*
The sample size is 217, R&D expenditures is missing for 90 firms, Underwriter ranking is missing for 6 firms and sales is missing for 4 firms, reducing the number of observations to 117. The dependent variable is the percentage first-day return from the offer price to the first-day closing price. The underwriter reputation dummy takes a value of one if the lead underwriter has an updated Carter and Manaster (1990) rank of 8 or more, and zero otherwise. Share overhang is the ratio of retained shares to the public float (the number of shares issued). R&D intensity is the ratio of Pre-IPO R&D expenditures to the expected market value.
The VC dummy takes a value of one (zero otherwise) if the IPO is backed by venture capital. Ln(1 + age) is the natural log of 1 plus the years since the firm's founding date as of the IPO. The Ln(assets) is the natural logarithm of the pre-issue book value of assets expressed in millions $. Ln(sales) is the natural log of the firm sales a year prior the offering, expressed in millions $. Firm profitability is the ratio of Pre-IPO EBITDA to total assets value.
ROA is the ratio of Pre-IPO net income to total assets value. The issue risk is a tech dummy takes a value of one (zero otherwise) if the firm is in the technology business. Insiders' ownership and blockholders' ownership are proxies to ownership structure and issuer bargaining power. AAII and II bullish bearish Ratios are the % of bullish investors divided by the % of bearish investors. Time dummy takes a value of one (zero otherwise) if the IPO occurred after July 2007.

	Mean	Median	Max	Min	Std Dev	Skewness	Kurtosis	Jarque-Bera	Probability
UNDERPRICING	0,295	0,219	1,254	0,055	0,235	1,408	4,831	55,027	0,000
UNDERWRITER REPUTATION	0,846	1,000	1,000	0,000	0,362	-1,919	4,682	85,584	0,000
OVERHANG RATIO	9,706	3,412	399,289	0,000	38,750	8,993	89,246	37839,240	0,000
R D INTENSITY	0,022	0,015	0,145	0,000	0,025	2,096	8,566	236,685	0,000
VC BACKED DUMMY	0,726	1,000	1,000	0,000	0,448	-1,016	2,033	24,699	0,000
FIRM AGE	2,475	2,303	4,905	0,000	0,871	0,612	3,988	12,057	0,002
FIRM SIZE LN ASSETS	4,300	3,985	8,640	0,272	1,506	0,900	3,788	18,828	0,000
FIRM SIZE LN SALES	3,473	3,859	9,225	-6,908	2,827	-2,051	8,830	247,683	0,000
FIRM PROFITABILITY	-0,064	0,053	2,062	-3,527	0,509	-2,448	22,255	1924,288	0,000
ROA	-0,068	-0,001	3,648	-1,279	0,497	3,222	28,728	3429,200	0,000
ISSUE RISK TECH DUMMY	0,556	1,000	1,000	0,000	0,499	-0,224	1,050	19,512	0,000
INSIDERS OWNERSHIP	0,555	0,599	1,000	0,000	0,269	-0,440	2,320	6,029	0,049
BLOCKHOLDERS OWNERSHIP	0,694	0,731	2,091	0,000	0,342	0,784	5,861	51,888	0,000
AAII BULL BEAR RATIO	1,263	1,219	2,148	0,491	0,416	0,325	2,401	3,812	0,149
INVI BULL BEAR RATIO	2,169	2,260	3,212	1,223	0,518	-0,171	1,865	6,852	0,033
TIME DUMMY	0,248	0,000	1,000	0,000	0,434	1,168	2,364	28,570	0,000

From the Jarque-Bera test, we can deduce that the dependant variable and all the explanatory variables are not normally distributed except for individual investors sentiment (p.value = 0.1486).

I classified the sample into two subsamples: firms with low underpricing (=sample median 21.85% (59 obs)) and firms with high underpricing (>sample median (58 obs)). (Descriptive statistics of the two subsamples are in tables 3 and 4)

For firms with high underpricing, I notice that the means of overhang ratio, R&D intensity and insiders' ownership are lower than those of low underpricing, but the mean of AAII bullish bearish ratio is higher: these remarks indicate that underpricing increases monotonically with higher individual investors' sentiment and decreases with higher firm quality and higher issuer bargaining power.

Table 5 indicates that firms with highest underpricing at the 1^st day of trading are almost risky firms (young firms with small assets and sales and operating in a technological sector). From the underwriter prestige segmentation, I remark that almost firms look for hiring prestigious underwriters and these firms have the highest underpricing degree. Higher issuer bargaining power leads to lower underpricing.

*Table 3. Descriptive Statistics of Firms with Low Underpricing*
Firms with low underpricing are firms with underpricing less than the median of the full sample 21.85%.

		Mean	Median		Max		Min	Std Dev	Skewness		Kurtosis			Jarque-Bera		Probability
UNDERPRICING		0,125	0,125		0,219		0,055	0,047	0,124		1,942			2,902		0,234
UNDERWRITER REPUTATION		0,814	1,000		1,000		0,000	0,393	-1,610		3,593			26,360		0,000
OVERHANG RATIO		14,228	2,895		399,289		0,160	54,189	6,352		44,887			4709,997		0,000
R D INTENSITY		0,028	0,022		0,145		0,000	0,029	1,743		6,773			64,893		0,000
VC BACKED DUMMY		0,712	1,000		1,000		0,000	0,457	-0,936		1,875			11,717		0,003
FIRM AGE		2,458	2,303		4,868		0,693	0,878	0,746		3,898			7,455		0,024
FIRM SIZE LN ASSETS		4,510	4,144		8,640		0,272	1,719	0,412		2,832			1,735		0,420
FIRM SIZE LN SALES		3,172	3,832		8,961		-6,908	3,436	-1,574		5,758			43,048		0,000
FIRM PROFITABILITY		-0,147	0,015		0,604		-3,527	0,589	-3,506		19,713			807,607		0,000
ROA		-0,076	-0,008		3,648		-1,279	0,623	3,309		23,149			1105,687		0,000
ISSUE RISK TECH DUMMY		0,576	1,000		1,000		0,000	0,498	-0,309		1,095			9,856		0,007
INSIDERS OWNERSHIP		0,603	0,649		1,000		0,000	0,268	-0,644		2,833			4,150		0,126
BLOCKHOLDERS OWNERSHIP		0,736	0,729		2,091		0,000	0,389	1,131		5,682			30,269		0,000
AAII BULL BEAR RATIO		1,229	1,216		2,087		0,537	0,403	0,394		2,446			2,277		0,320
INVI BULL BEAR RATIO		2,174	2,123		3,212		1,223	0,518	-0,271		2,022			3,074		0,215
TIME DUMMY		0,153	0,000		1,000		0,000	0,363	1,933		4,736			44,138		0,000
*Table 4. Descriptive Statistics of Firms with High Underpricing*
Firms with high underpricing are firms with underpricing more than the median of the full sample 21.85%.

	Mean		Median		Max	Min		Std Dev		Skewness			Kurtosis		Jarque-Bera		Probability
UNDERPRICING	0,468		0,412		1,254	0,221		0,223		1,107			4,198		15,324		0,000
UNDERWRITER REPUTATION	0,879		1,000		1,000	0,000		0,329		-2,329			6,423		80,737		0,000
OVERHANG RATIO	5,105		3,843		30,441	0,000		5,007		3,356			15,321		475,732		0,000
R D INTENSITY	0,016		0,011		0,109	0,000		0,020		2,489			10,947		212,535		0,000
VC BACKED DUMMY	0,741		1,000		1,000	0,000		0,442		-1,102			2,216		13,237		0,001
FIRM AGE	2,492		2,303		4,905	0,000		0,871		0,474			4,106		5,126		0,077
FIRM SIZE LN ASSETS	4,086		3,812		8,553	2,323		1,231		1,727			6,334		55,706		0,000
FIRM SIZE LN SALES	3,778		3,874		9,225	-6,908		2,015		-2,756			17,031		549,203		0,000
FIRM PROFITABILITY	0,021		0,064		2,062	-1,026		0,400		1,819			13,611		304,105		0,000
ROA	-0,059		0,024		1,143	-1,095		0,328		0,016			6,215		24,986		0,000
ISSUE RISK TECH DUMMY	0,534		1,000		1,000	0,000		0,503		-0,138			1,019		9,668		0,008
INSIDERS OWNERSHIP	0,506		0,548		0,914	0,000		0,262		-0,290			1,953		3,463		0,177
BLOCKHOLDERS OWNERSHIP	0,651		0,746		1,095	0,000		0,282		-0,593			2,329		4,481		0,106
AAII BULL BEAR RATIO	1,297		1,265		2,148	0,491		0,430		0,244			2,359		1,567		0,457
INVI BULL BEAR RATIO	2,164		2,260		3,212	1,282		0,522		-0,072			1,717		4,027		0,133
TIME DUMMY	0,345		0,000		1,000	0,000		0,479		0,653			1,426		10,106		0,006
Table 5. Average First-day Returns Categorized by Underwriter prestige, Share Overhang, R&D Intensity, VC Backing, Age, Assets, Sales, Firm profitability, Industry, Bargaining Power, Individual Investors' sentiment and Time
Data are from iposcoop, Securities and Exchange Commissions data and other sources; the sample size is 217, R&D expenditures is missing for 90 firms, Underwriter ranking is missing for 6 firms and sales is missing for 4 firms. High prestige underwriters are those with a ranking of 8 or higher on 0-9 scale. Low share overhang IPOs have an overhang ratio of 3.412 (median) or lower. Firms with low R&D intensity are those with R&D intensity lower than 1.5% (median). Firms with 0-7 years old are classified as young. Firms with 12 months pre-IPO assets less than $53.8 million are classified as small. Firms with 12 months pre-IPO sales less than $47.4 million are classified as low sales firms. Firms with insiders' ownership less than 59.88% (median) are classified as firms with low bargaining power.
Segmented by				Average 1st day return								N
Underwriter prestige
Low prestige				17.59%								39
High prestige				27.77%								164
Share Overhang
Low				20.53%								130
High				31.93%								87
R&D intensity
Low				35.02%								63
High				21.74%								64
Venture Capital backing
Non VC backed				20.70%								95
VC backed				28.53%								122
Age
Young ( 0-7 years old )				25.98%								68
Old ( 8 years and older )				24.7%								149
Assets
Small				31.18%								72
Large				22.08%								145
Sales
Small				28.21%								76
Large				23.97%								135
Firm Profitability
<0				24.88%								67
>=0				25.20%								150
Industry
Non Technology				23.50%								145
Technology				28.33%								72
Bargaining power
Low bargaining power				26.23%								107
High bargaining power				24%								110
Individual Investors' sentiment
Pessimistic sentiment				25.07%								162
Optimistic sentiment				25.21%								55
Time
Before crisis period				22.57%								168
After crisis period				33.80%								49
ALL				25.10%								217

III- Empirical implications and analysis:

Before presenting the results and empirical implications of the model, I should describe the different steps and the methodology I used to obtain the final specification:

Ø 1^st Step: using the AAII and II weekly sentiment indicators, I calculate three investors sentiment measures: bullish proportion (% of bullish investors / % of bullish and bearish investors), bullish-bearish Spread (% of bullish investors - % of bearish investors) and the bullish bearish Ratio (% of bullish investors/ % of bearish investors). From these 3 possible measures, I should find the best measure of sentiment and the model that best explains the data. I try to construct 3 models, each one contains all the variables with one of the 3 possible sentiment measures, so every specification contains 17 explanatory variables with a constant: Underwriter reputation, Overhang ratio, R&D intensity, VC backed, Firm age, Firm size Ln(assets), Firm size Ln(sales), Issue size, Firm profitability, ROA, Issue risk (tech dummy), Insiders' ownership, Institutional ownership, Blockholders' ownership, AAII sentiment indicator, II sentiment indicator and Time dummy.

The regression results and the models adjustment quality and different measures of goodness of fit allow the choice of the model containing the bullish bearish Ratio as the model that best explains underpricing within the sample and using these control variables: the highest R-squared and adjusted R-squared which measure the success of the regression in predicting the values of the dependant variable within the sample, and the lowest Akaike Information Criterion and Schwarz Criterion which are measures of the goodness of fit of an estimated statistical model and they measure the efficiency of the parameterized model in terms of predicting the data. So in a first step, I continue working with the model using the Bullish Bearish Ratio as measure of investor sentiment with all the other explanatory variables.

The tables 6 and 7 present the regression results of the models containing all the variables with respectively the bullish proportion and the bullish-bearish spread as measures of investors' sentiment. The tables also contain criterion used to decide which model explains best the dependant variable within the sample.

*Table 6. Ordinary Least Squares Regression Results (with Bullish proportion as investors' sentiment measure)*
Dependent Variable: UNDERPRICING
Method: Least Squares
Sample (adjusted): 1 214 Included observations: 117 after adjustments
White Heteroskedasticity-Consistent Standard Errors & Covariance
Variable	Coefficient		Std. Error		t-Statistic		Prob.
C	0.069774		0.309901		0.225149		0.8223
UNDREWRITER REPUTATION	0.095300		0.071413		1.334489		0.1851
OVERHANG RATIO	-0.000577		0.000260		-2.218402		0.0288
R D INTENSITY	-2.857168		0.928221		-3.078112		0.0027
VC BACKED DUMMY	0.014893		0.062503		0.238276		0.8122
FIRM AGE	0.014673		0.031730		0.462419		0.6448
FIRM SIZE LN ASSETS	-0.028986		0.026733		-1.084293		0.2809
FIRM SIZE LN SALES	-0.000771		0.009336		-0.082608		0.9343
ISSUE SIZE	0.007820		0.039760		0.196668		0.8445
FIRM PROFITABILITY	-0.067138		0.053643		-1.251562		0.2137
ROA	0.055279		0.059343		0.931524		0.3538
ISSUE RISK TEH DUMMY	0.000610		0.042783		0.014249		0.9887
INSIDER OWNERSHIP	-0.170075		0.086312		-1.970474		0.0516
INSTITUTIONAL OWNERSHIP	0.233858		0.231054		1.012134		0.3139
BLOCKHOLDERS OWNERSHIP	-0.368274		0.220834		-1.667651		0.0985
AAII BULL PROPORTION	0.477753		0.238433		2.003722		0.0478
INVI BULL PROPORTION	0.226409		0.369040		0.613508		0.5409
TIME DUMMY	0.146285		0.053135		2,742582		0.0070
*R-squared*	0.286636		*Mean dependent var*				0.295424
*Adjusted R-squared*	0.164139		*S.D. dependent var*				0.234851
*S.E. of regression*	0.214713		*Akaike info criterion*				-0.098386
*Sum squared resid*	4.564085		*Schwarz criterion*				0.326563
*Log likelihood*	23.75561		*Hannan-Quinn criter.*				0.074138
*F-statistic*	2.339949		*Durbin-Watson stat*				2.326671
*Prob(F-statistic)*	0.004784
*Table 7. Ordinary Least Squares Regression Results (with Bullish Bearish Spread as investors' sentiment measure)*
Dependent Variable: UNDERPRICING
Method: Least Squares
Sample (adjusted): 1 214 Included observations: 117 after adjustments
White Heteroskedasticity-Consistent Standard Errors & Covariance
Variable		Coefficient		Std. Error		t-Statistic		Prob.
C		0.411271		0.171480		2.398364		0.0183
UNDREWRITER REPUTATION		0.095188		0.071482		1.331636		0.1860
OVERHANG RATIO		-0.000600		0.000259		-2.317958		0.0225
R D INTENSITY		-2.842995		0.930156		-3.056472		0.0029
VC BACKED DUMMY		0.015700		0.062839		0.249853		0.8032
FIRM AGE		0.014385		0.031341		0.458990		0.6472
FIRM SIZE LN ASSETS		-0.027984		0.026783		-1.044824		0.2986
FIRM SIZE LN SALES		-0.000854		0.009332		-0.091548		0.9272
ISSUE SIZE		0.006576		0.040015		0.164343		0.8698
FIRM PROFITABILITY		-0.067658		0.053506		-1.264489		0.2090
ROA		0.058442		0.059784		0.977542		0.3307
ISSUE RISK TECH DUMMY		0.001121		0.042737		0.026232		0.9791
INSIDER OWNERSHIP		-0.167241		0.085970		-1.945335		0.0546
INSTITUTIONAL OWNERSHIP		0.246116		0.228617		1.076544		0.2843
BLOCKHOLDERS OWNERSHIP		-0.381754		0.219424		-1.739803		0.0850
AAII BULL BEAR SPREAD		0.003062		0.001532		1.998317		0.0484
INVI BULL BEAR SPREAD		0.001949		0.002459		0.792373		0.4300
TIME DUMMY		0.144385		0.053118		2.718194		0.0078
*R-squared*		0.288315		*Mean dependent var*				0.295424
*Adjusted R-squared*		0.166106		*S.D. dependent var*				0.234851
*S.E. of regression*		0.214461		*Akaike info criterion*				-0.100743
*Sum squared resid*		4.553344		*Schwarz criterion*				0.324207
*Log likelihood*		23.89344		*Hannan-Quinn criter.*				0.071782
*F-statistic*		2.359205		*Durbin-Watson stat*				2.328713
*Prob(F-statistic)*		0.004430

Ø 2^nd Step: I check the correlation between the explanatory variables, I use Klein correlation detection criterion: comparison between R-squared (R^²) and the different squared simple correlation coefficients (r^²) with r^²=Cov^²(X_i,X_j)/ Var(X_i)*Var(X_j)

For this, I should use the correlation matrix to determine the different correlation coefficients between all the explanatory variables. For R^²= 0.2897, I find a correlation between issue size and firm size Ln (assets) (r^²= 0.4361) and a correlation between institutional ownership and blockholders' ownership (r^²= 0.8326). I construct all the specifications necessary to decide which control variables I should eliminate and which explanatory variables I should keep in the model to give the best explanation of underpricing phenomenon (the dependant variable of the model). Founding my decision on measures of goodness of fit and of adjustment quality for the model, I eliminate issue size and institutional ownership, and I keep all the other control variables. The final specification contains 15 explanatory variables: Underwriter Reputation, Overhang Ratio, R&D Intensity, VC Backed, Firm Age, Firm Size Ln(assets), Firm Size Ln(sales), Firm Profitability, ROA, Issue Risk (tech dummy), Insiders' Ownership, Blockholders' Ownership, AAII Bullish Bearish Ratio, II Bullish Bearish Ratio and Time Dummy.

With a new correlation matrix, I check the correlation between these 15 control variables: r^²_Xi,Xj< R^²= 0.2833 for i?j

There is no correlation between the control variables after eliminating issue size and institutional ownership.

Underpricing = a₀ + a₁Underwriter Reputation Dummy + a₂Overhang Ratio+ a₃R&D Intensity + a₄VC Dummy + a₅Ln (1+age) + a₆Ln (assets) + a₇Ln (sales) + a₈Firm Profitability + a₉ROA + a₁₀Issue Risk Dummy + a₁₁Insiders' Ownership + a₁₂ Blockholders' Ownership + a₁₃Individual Bullish Bearish Ratio+ a₁₄Institutional Bullish Bearish Ratio + a₁₅Time Dummy + å_i

With underwriter reputation, overhang ratio, R&D intensity and VC backing used to measure the informational asymmetry and exactly are proxies for the firm quality. Age, firm size, firm profitability, ROA and issue risk are measures of risk. Insiders' ownership and blockholders' ownership are proxies for issuer bargaining power. Individual and Institutional bullish bearish ratios are measures of investors' sentiment. Finally, time dummy is a control variable to verify the impact of the beginning of the financial crisis on IPO underpricing.

I present the Ordinary Least Squares regression results in table 8. The first result is that the model is statistically significant (p.value= 0.001935), which means the relevance of the control variables used in this model to explain the underpricing phenomenon (the dependant variable). I find acceptable values of Akaike Information Criterion and Schwarz Criterion which measure the goodness of fit and the efficiency of the parameterized model in terms of predicting the data.

Let's verify the individual significance of the explanatory variables. 5 statistically significant variables: Overhang Ratio, R&D Intensity, Insiders' Ownership, Individual Bullish Bearish Ratio and Time Dummy.

§ The negative and significant coefficient of insiders' ownership is consistent with the hypothesis that issue firms with high bargaining power request higher offer prices conducting lower degree of underpricing. This negative relation between issuer bargaining power and underpricing is consistent with the findings of Ljungqvist, Nanda and Singh (2003), Ljungqvist and Wilhelm (2003), Loughran and Ritter (2004) and many other researchers who used this variable in their studies.

§ The positive and significant coefficient of Individual Bullish Bearish Ratio is consistent with the hypothesis that investors with high and optimistic sentiment are willing to pay higher prices to acquire the IPO shares leading to higher underpricing. This result is consistent with the findings of researchers who are interested in the behavioral explanation of IPO first day price run up using different proxies for sentiment: Lee, Shleifer and Thaler (1991) with discount on closed-end funds, Cornelli, Goldreich and Ljungqvist (2004) with grey market prices and many other researchers using the behavioral approach and the investor sentiment.

§ The positive and significant coefficient of Time dummy presents the positive relation between underpricing and the beginning of the financial crisis which has an impact on the IPO first day returns.

§ The negative and significant coefficients of overhang ratio and R&D intensity (both are measures of firm quality) can be explained by the fact that high quality issue firms request higher offer prices for their IPOs conducting lower level of underpricing.

§ The lack of significance of the 6 measures of risk (age, Ln(assets), Ln(sales), firm profitability, ROA and issue Risk (tech dummy)) is consistent with the findings of Bartov, Mohanram and Seethamraju (2003): using a dummy variable for risky IPOs, they find that there is no correlation between risk and underpricing and the notion of risk can not explain the setting of a lower offer price and then the underpricing phenomenon.

§ The lack of significance of the VC backing coefficient confirms the findings of many researchers that used VC dummy as a proxy and as a signal for firm quality. A question can be asked about the explanatory power of this variable and its introduction in explanatory models of underpricing phenomenon.

By regrouping the three theories advanced earlier as explanations for the underpricing phenomenon in the same model and by using different proxies for each theory, I show that:

In this model, I distinguish between the two types of investors: individual and institutional investors using a direct measure of sentiment for each type calculated from the American Association of Individual Investors and the Investors Intelligence survey data, these are two other contributions in this study. The positive and significant coefficient of individual investors' sentiment and the positive but insignificant coefficient of institutional investors' sentiment lead to an important finding: individual investors are those driving the first day closing prices and then underpricing anomaly and are more conducting the short run IPO puzzle than the institutional investors. The individual investors' sentiment and their over optimism and enthusiasm is more important and relevant in explaining the IPO underpricing anomaly than does the institutional investors' sentiment. Individual investors are the type more conducting the IPO first day returns and the short run IPO puzzle.

This finding can be explained by the fact that in some researches and studies, institutional investors are defined as rational investors. This can be a possible explanation for the irrelevance of institutional investors' sentiment in explaining the underpricing anomaly.

*Table 8. Ordinary Least Squares Regression Results*
The sample size is 217, R&D expenditures is missing for 90 firms, Underwriter ranking is missing for 6 firms and sales is missing for 4 firms, reducing the number of observations to 117. The dependent variable is the percentage first-day return from the offer price to the first-day closing price. The underwriter reputation dummy takes a value of one if the lead underwriter has an updated Carter and Manaster (1990) rank of 8 or more, and zero otherwise. Share overhang is the ratio of retained shares to the public float (the number of shares issued). R&D intensity is the ratio of Pre-IPO R&D expenditures to the expected market value.
The VC dummy takes a value of one (zero otherwise) if the IPO is backed by venture capital. Ln(1 + age) is the natural log of 1 plus the years since the firm's founding date as of the IPO. The Ln(assets) is the natural logarithm of the pre-issue book value of assets expressed in millions $. Ln(sales) is the natural log of the firm sales a year prior the offering, expressed in millions $. Firm profitability is the ratio of Pre-IPO EBITDA to total assets value.
ROA is the ratio of Pre-IPO net income to total assets value. The issue risk is a tech dummy takes a value of one (zero otherwise) if the firm is in the technology business. Insiders' ownership and blockholders ownership are proxies to ownership structure and issuer bargaining power.AAII and II bullish bearish Ratios are the % of bullish investors divided by the % of bearish investors. Time dummy takes a value of one (zero otherwise) if the IPO occurred after July 2007. The t-statistics are calculated using White's (1980) heteroskedasticity-consistent method.
Dependent Variable: UNDERPRICING
Method: Least Squares
Sample (adjusted): 1 214 Included observations: 117 after adjustments
White Heteroskedasticity-Consistent Standard Errors & Covariance

Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.292331	0.166898	1.75155	0.0829
UNDERWRITER REPUTATION	0.093909	0.069317	1.354779	0.1785
OVERHANG RATIO	-0.000488	0.000221	-2.205774	0.0297
R D INTENSITY	-2.946844	0.867721	-3.396074	0.001
VC BACKED DUMMY	0.029582	0.057948	0.510493	0.6108
FIRM AGE	0.010542	0.030908	0.341071	0.7338
FIRM SIZE LN ASSETS	-0.02194	0.02136	-1.027144	0.3068
FIRM SIZE LN SALES	-0.000269	0.009324	-0.028826	0.9771
FIRM PROFITABILITY	-0.034211	0.049397	-0.692579	0.4902
ROA	0.017825	0.051876	0.343617	0.7318
ISSUE RISK TECH DUMMY	-0.004296	0.0414	-0.103774	0.9176
INSIDERS OWNERSHIP	-0.177436	0.082075	-2.161876	0.033
BLOCKHOLDERS OWNERSHIP	-0.137413	0.09812	-1.400456	0.1644
AAII BULL BEAR RATIO	0.098838	0.04782	2.066853	0.0413
INVI BULL BEAR RATIO	0.034061	0.039904	0.853586	0.3954
TIME DUMMY	0.147677	0.053652	2.752493	0.007
*R-squared*	0.283352	*Mean dependent var*		0.295424
*Adjusted R-squared*	0.176919	*S.D. dependent var*		0.234851
*S.E. of regression*	0.213066	*Akaike info criterion*		-0.127981
*Sum squared resid*	4.585098	*Schwarz criterion*		0.249752
*Log likelihood*	23.48689	*Hannan-Quinn criter.*		0.025374
*F-statistic*	2.662259	*Durbin-Watson stat*		2.364719
*Prob(F-statistic)*	0.001935

CONCLUSION

he first and the most important observation in the IPO market comes back to the early writers that have been interested in IPO market, notably Stoll and Curley (1970), Logue (1973), Reilly (1973) and Ibbotson (1975), who documented that when companies go public, the price of shares they sell tends to jump substantially on the first day of trading and the first day closing price is systematically higher than the issue price at which the public offering was introduced in the market. This phenomenon is called «underpricing» from the issuer's point of view, he thinks that he has not correctly valued the IPO shares. He underpriced the real value of the shares of his company, it is money left on the IPO table which could have been raised if the offer price had been set at an appropriate and higher level.

IPOs exhibit positive first day returns on average with no exception to the period and date of going public, to the country and to the industry to which the IPO belongs. Underpricing is a persistent anomaly characterizing the short run IPO market behaviour. It is a puzzle to resolve.

Underpricing anomaly has intrigued academics and practitioners over the past three decades and has generated considerable research which provided numerous analytical advances and empirical insights aimed at explaining this short run phenomenon.

The list of explanations that were advanced is very long, but it can be classified in three main categories:

Ø Theories and explanations based on the informational asymmetry between the key parties of an IPO.

Ø Theories asserting the informational transparency and lucidity and the IPO market efficiency.

Ø Behavioral explanations based on the importance of investors' sentiment as a primary driver to underpricing and to first day returns.

Concerning the theories asserting the informational asymmetry, we can begin by the fact that issuers are more informed than the other parties, and then we find the theory of signalling based on the importance of the issuing firm quality. Issuers are the most informed about the true value and quality of their firms, and the high quality issuers may attempt to signal this quality and value and to distinguish themselves from the pool of low quality issuers. They may voluntarily sell their shares at a lower price than the market beliefs to signal their high quality. They leave deliberately money on the IPO table to deter lower quality issuers from imitating. However, another conduct can be observed, high quality issuer may bargain harder for higher offer price since he believes in the quality of his firm and its true value conducting lower level of underpricing.

The most important proxies of firm quality used in almost studies and researches to verify the relevance of the relation between firm quality and underpricing are:

R&D Intensity, Overhang Ratio, Venture Capital Backing as a dummy variable, Underwriter reputation,... The result found in almost studies is a positive and significant relation between firm quality and underpricing: Higher firm quality issuers do not bargain for higher offer prices to signal the quality of their firms and to deter lower quality firms from imitating and then a higher level of underpricing is observed at the first day of trading. And the explanations for not requiring high offer prices vary from believing in high returns on future SEO, to increasing the probability of a full subscription of the first offering, this is dependant on issuers' beliefs and future strategies.

As I said the relation between firm quality and underpricing can be negative, high quality issuers bargain for higher offer prices and then a low level of underpricing is observed at the end of the first day of trading. However, some researchers also find this relation insignificant.

Other explanations were advanced going in the same direction of informational asymmetry but asserting investors as the most informed party of an IPO. We find the «Information Revelation Theory» in which underpricing is compensation for revealing private information and interest indications, the Winner's Curse explanation (Rock 1986) assuming an imbalance of information between the potential investors themselves and the importance of underpricing to encourage the best informed investors to participate in an unattractive offering, and the agency conflicts between underwriter and issuing firm.

The second category of explanations is assuming the informational transparency and lucidity between the key parties and the IPO market efficiency, and then underpricing phenomenon can be explained by the characteristics of the offering: risk, issue size, or issuer bargaining power.

Riskier issuing firms can not require high offer prices and high underpricing is then observed. Risk can be proxied by issuing firm age, issuing firm size (Ln(assets), Ln(sales)), or by the business and the activity sector of the firm. Besides, Bartov, Mohanram and Seethamraju (2003) report that a dummy variable for risky IPOs has no effect on the setting of the final offer price and that is no correlation between risk and underpricing.

Issue size is also an important determinant: when the issue size is large, the issue price should reflect the greater difficulty of selling the shares in the aftermarket, and then the issue price should be lower leading to higher level of underpricing. The relation can be seen of another side, sizable issue is less risky and issuer can request higher offer price.

Issuer bargaining power can also be an explanation to this short run anomaly. Issuing firms with concentrated ownership structure and so with high bargaining power require high offer prices for their issues conducting lower underpricing.

Some researches were advanced and based on other explanations of underpricing by lawsuit avoidance, underpricing as a substitute of Marketing expenditures, or by favourable market conditions.

The third category of explanations is the most important and promising area of research, since all the «traditional» theories advanced earlier are unlikely to clarify and to give a relevant, a reliable and a convincing explanation to the underpricing anomaly, mainly after the surprisingly and severe level of underpricing of the dot-com bubble. Turning to behavioral approach and to investors' sentiment seems to be a necessity to clarify the mystery of underpricing. Behavioral approach and investor sentiment is not a new field, the investor sentiment was introduced earlier in the 1990's by Welch who presented the Informational Cascade theory. However, this approach has attracted more attention, has intrigued more and more researchers and has taken all its impetus in this decade. Many researchers tried to introduce this behavioral approach to explain the short run IPO anomaly and the research effort is continuing.

Informational Cascade Theory (Welch 1992) assumes that issuers should succeed the first sales and should underprice to induce the first few potential buyers and later induce a cascade: bandwagon effects. All subsequent investors want to buy irrespective of their own information, they will imitate the first potential investors. Then we find, Loughran and Ritter (2002) who try to apply the prospect theory of Kahneman and Tversky (1979) to IPO market to argue that issuers are more tolerant of excessive underpricing if they simultaneously learn about an aftermarket valuation that is higher than expected.

But, we can say that the importance of investor sentiment was introduced and analyzed in the context of the underpricing phenomenon for the first time by Ljungqvist, Nanda and Singh (2004) in their article «Hot markets, Investor sentiment and IPO pricing».

Many researchers try to value the investors' sentiment to study its impact on underpricing and to verify the significance of the relation between underpricing and investors' sentiment. The list of proxies used by researchers to value investor sentiment is very long. I present the most important: grey market prices, market conditions, demand submitted by individual investors and discounts on closed-end funds. And the main result is that high investors' sentiment induces high underpricing. The optimism and enthusiasm of investors is positively related to first day returns and to underpricing.

In the third section of this work, I regroup the most important explanations that have been advanced in the same model to determine which of these explanations characterizes best a sample of 217 U.S IPOs for 2006 and 2007. In the context of a unified framework and model, I present the three theories: asymmetric, symmetric and behavioral approach. I use many variables as proxies for the same theory and I even introduce some variables found insignificant in earlier studies to verify the relevance of earlier results for this sample. I also distinguish between the individual investors' sentiment and the institutional investors' sentiment and I use a direct measure of sentiment. In the model, I use 15 explanatory variables: Underwriter Reputation, Overhang Ratio, R&D Intensity and VC Backing used to measure the informational asymmetry and exactly are proxies for the firm quality. Age, Firm Size, Firm Profitability, ROA and Issue Risk are measures of risk. Insiders' Ownership and Blockholders' Ownership are proxies for issuer bargaining power. Individual and Institutional Bullish Bearish Ratios are measures of investors' sentiment. Finally, Time Dummy is a control variable to verify the impact of the beginning of the financial crisis on IPO underpricing.

The negative and significant coefficient of insiders' ownership is consistent with the hypothesis that issue firms with high bargaining power request higher offer prices conducting lower degree of underpricing. The positive and significant coefficient of Individual bullish bearish ratio is consistent with the hypothesis that investors with high and optimistic sentiment are willing to pay higher prices to acquire the IPO shares leading to higher underpricing. The positive and significant coefficient of Time dummy presents the positive relation between underpricing and the beginning of the financial crisis which has an impact on the IPO first day returns. The negative and significant coefficients of overhang ratio and R&D intensity (both are measures of firm quality) can be explained by the fact that high quality issue firms request higher offer prices for their IPOs conducting lower level of underpricing. And the lack of significance of the 6 measures of risk and of the VC backing is consistent with the findings of many researchers.

Finally, the positive and significant coefficient of individual investors' sentiment and the positive but insignificant coefficient of institutional investors' sentiment lead to an important finding: Individual investors are those driving the first day closing prices and then underpricing anomaly and are more conducting the short run IPO puzzle than the institutional investors.

Underpricing phenomenon can be explained by the informational asymmetry theories based on issuer as the best informed party of an IPO about its firm quality (a negative relation), by issuer bargaining power (a negative relation) and by the importance of the individual investors' sentiment: the higher the individual investors' optimism, the higher are the first day closing prices and the higher are the first day returns and Underpricing anomaly.

References

Baker, M., and J. Wurgler (2006), «Investor Sentiment and the Cross-Section of Stock Returns» Journal of Finance, Vol. LXI, No. 4, pp. 1645-1680.

Baker, M., and J. Wurgler (2007), «Investor Sentiment in the Stock Market», Journal of Economic Perspectives, Vol. 21, No. 2, pp. 129-151.

Bradley, D.J., and B. Jordan (2002), «Partial Adjustment to Public Information and IPO Underpricing», Journal of Financial and Quantitative Analysis, Vol. 37, No. 4, pp. 595-616.

Brown, G., and M. Cliff (2005), «Investor Sentiment and Asset Valuation», Journal of Business, Vol. 78, No. 2, pp.405-440.

Chung, K.H., M. Li, and L. Yu (2005), «Assets in Place, Growth Opportunities, and IPO Returns», Financial Management, pp. 65-88.

Cornelli, F., D. Goldreich and A.P. Ljungqvist (2004), «Investor Sentiment and Pre-IPO Markets», Working Paper, London Business School, pp. 1-45.

Doukas, J., and N.T. Milonas (2004), «Investor Sentiment and the Closed-end Fund Puzzle: Out-of-sample Evidence», European Financial Management, Vol.10, No.2, pp. 235-266.

Filatochev, I., and K. Bishop (2002), «Board composition, share ownership, and «underpricing» of U.K. IPO firms», Strategic Management Journal 23, pp. 941-955.

Glushkov, D. «Investor sentiment, firm characteristics and institutional behaviour», pp. 1-55.

Guo, R., B. Lev, and C. Shi (2006), «Explaining the Short- and Long-Term IPO Anomalies in the US by R&D», Journal of Business Finance & Accounting, Vol. 33, pp. 550-579.

Ivanov, V., and C. Lewis (2008), «The determinants of market-wide issue cycles for initial public offerings», Journal of Corporate Finance 14, pp. 567-583.

Kaserer, C., and D. Schiereck (October, 2008), «Primary Market Activity and the Cost of Going and Being Public», pp.1-73.

Kaustia, M., and S. Knüpfer (December 2008), «Do Investors Overweight Personal Experience? Evidence from IPO Subscriptions», Journal of Finance, Vol. LXIII, No. 6, pp. 2679-2702.

Leite, T. (2007), «Adverse selection, public information, and underpricing in IPOs», Journal of Corporate Finance 13, pp. 813-828.

Leoni, P. (2008), «A market microstructure explanation of IPOs underpricing», Economics Letters 100, pp. 47-48.

Ljungqvist, A.P. (2004), «IPO Underpricing», Handbooks in Finance: Empirical Corporate Finance, Chapter III.4, pp. 1-74.

Ljungqvist, A.P., V. Nanda and R. Singh (2003), «Hot Markets, Investor Sentiment, and IPO Pricing», pp. 1-49.

Ljungqvist, A.P., and W.J. Wilhelm, (December 2003), «What makes issuers happy? Testing the Prospect Theory of IPO Underpricing», NYU Working Paper, pp.1-38.

Loughran, T., and J. Ritter (2004), «Why Has IPO Underpricing Changed Over Time?», Financial Management, pp. 5-37.

Loughran, T., J. Ritter and K. Rydqvist (1994), «Initial Public Offerings: International Insights», Pacific-Basin Finance Journal, Vol. 2, pp. 165-199.

Lowry, M. (2003), «Why Does IPO Volume Fluctuate So Much?», Journal of Financial Economics 67, pp. 3-40.

Mantecon, T., and P. Poon (2009), «An analysis of the liquidity benefits provided by secondary markets», Journal of Banking & Finance 33, pp. 335-346.

Oehler, A., M. Rummer, and P. N. Smith, (Revised Version, July 2005), «IPO Pricing and the Relative Importance of Investor Sentiment - Evidence from Germany», pp. 1-42.

Ritter, J., and I. Welch (2002), «A Review of IPO Activity, Pricing, and Allocations», Journal of Finance, Vol.57, pp. 1795-1828.

Verma, R., and P. Verma (2008), «Are survey forecasts of individual and institutional investor sentiments rational?», International Review of Financial Analysis 17, pp. 1139-1155.

Yong, O. (2007), «A review of IPO research in Asia: What's next?», Pacific-Basin Finance Journal 15, pp. 253-275.

Zheng, S.X., and M. Li (2008), «Underpricing, ownership dispersion, and aftermarket liquidity of IPO stocks», Journal of Empirical Finance 15, pp. 436-454.

Investor sentiment and short run IPO anomaly: a behavioral explanation of underpricing

Table of contents