Underpricing In Initial Public Offerings: Evidence From The U.S Avgoustinos Filippoupolitis

The variables tested for their relationship with the degree of underpricing are the firm size, the firm age, the industry, the offer price, the IPO gross proceeds, the time lag between o

A thesis submitted

to the Department of Accounting and Finance

of the Athens University of Economics and Business

as partial fulfillment of the requirements for the

Master‟s Degree

Athens July, 2013

We approve the thesis of

“I hereby declare that this particular thesis has been written by me, in order to obtain the

Postgraduate Degree in Accounting and Finance, and has not been submitted to or

approved by any other postgraduate or undergraduate program in Greece or abroad This

thesis presents my personal views on the subject All the resources I have used for the

preparation of this particular thesis are mentioned explicitly with references being made

either to their authors, or to the URL‟ s (if found on the internet).”

Avgoustinos Filippoupolitis

Table of Contents

Table of Contents 4

List of Figures & Tables 5

Abstract 6

Πεξίιεςε ζηα Ειιεληθά 7

Chapter 1: Introduction 8

Chapter 2: Literature Review 9

2.1 Introduction 9

2.2 Going Public: Basic Considerations 9

2.2.1 Why and When Firms Go Public 9

2.2.2 How Firms Go Public 10

2.2.3 Costs of Going Public 13

2.3 Market Problems with the Pricing of IPOs 14

2.3.1 Short-Run Underpricing 14

2.3.2 Hot & Cold IPO Markets 16

2.3.3 Long-Run Underperformance 17

2.4 IPO Underpricing Theories 17

2.4.1 Assymetric Information Models 18

2.4.2 Institutional Explanations 21

2.4.3 Ownership and Control 23

2.4.4 Behavioral Explanations 24

2.5 How & Why Underpricing Changed Over Time 26

2.6 Discussion on Literature Review 27

Chapter 3: US IPO Market 29

3.1 Introduction 29

3.2 IPO Process in the US – Statutory Framework 29

3.3 US IPO Market Activity & Underpricing in the Literature 30

Chapter 4: Data, Research Methodology & Results 32

4.1 Introduction 32

4.2 Data & Sample 32

4.3 Sample Characteristics 33

4.4 Methodology 35

4.5 Results 42

4.5.1 The Degree of Raw and Market Adjusted Underpricing 42

4.5.2 Cross Sectional Regression Results 49

Chapter 5: Discussion 54

Chapter 6: Summary & Conclusions 56

Bibliography 57

List of Figures & Tables

Figure 1: Initial IPO Returns in the US from 1980 -2002……… 31

Table 1: Equally weighted average initial returns for selected countries……….…… 15

Table 2: Number of IPOs and aggregate proceeds per year……… … 30

Table 3: Frequency of IPOs, US markets and gross proceeds per year….……… 33

Table 4: Frequency of IPOs and gross proceeds by industry….……… 34

Table 5: Selected characteristics of the IPOs ….……… … 35

Table 6: Determinants of the performance of IPOs and hypotheses tested…… ……… 38

Table 7: Descriptive statistics for dependent and independent variables.……….…… 40

Table 8: Correlation matrix.……….……… 41

Table 9: Underpricing of IPOs per year.……….……… 43

Table 10: Descriptive statistics - underpricing……….……….… 43

Table 11: Distribution of raw and adjusted underpricing……….…….…… 44

Table 12: Raw and market adjusted initial returns by industry……….…… 45

Table 13: Raw and market adjusted initial returns by age group……… …… 45

Table 14: Raw and market adjusted initial returns by size……….……….46

Table 15: Raw and market adjusted initial returns by exchange……… …… 46

Table 16: Raw and market adjusted initial returns by underwriter reputation……….….…… 46

Table 17: Raw and market adjusted initial returns by number of uses of proceeds…… …… 47

Table 18: Raw and market adjusted initial returns by hot/cold periods…….……… 47

Table 19: Raw and market adjusted initial returns before and after the crisis ………… …… 47

Table 20: Raw and market adjusted initial returns for selected groups …….……… 48

Table 21: Test of differences in mean returns …….……….… 49

Table 22: Cross sectional regression results…….……… 51

One of the findings of this study is that the average initial return for the US IPOs in the period in scope is 12% with a standard deviation of 22.4% No statistically significant difference has been found in the degree of underpricing for IPOs before and after the economic crisis in 2008, but the hypothesis that there is no statistically significant difference in underpricing between the years has been rejected at 1% level

The variables tested for their relationship with the degree of underpricing are the firm size, the firm age, the industry, the offer price, the IPO gross proceeds, the time lag between offering and listing, the exchange of listing, the number of uses of proceeds in prospectus, the underwriter reputation ranking and the market condition All these variables seem to follow existing literature apart from the number of uses, the age and the underwriter reputation, which have the opposite from the hypothesized relationship

Additionally, it has been concluded that the only statistically significant variables in influencing the degree of underpricing are the offer price, the number of uses of gross proceeds, the firm classification as high-technology and the listing on NASDAQ Using these variables together with the underwriter‟s reputation and market condition, a six factor model for the initial underpricing has been built with OLS regression analysis The model explains 9.78% of the total variation in underpricing

Περίληψη στα Ελληνικά

Σηα πιαίζηα ηεο παξνύζαο εξγαζίαο, πξαγκαηνπνηήζεθε κία αλαζθόπεζε ηεο ππάξρνπζαο βηβιηνγξαθίαο ζρεηηθά κε ην θαηλόκελν ηεο ππνηηκνιόγεζεο ζηηο αξρηθέο δεκόζηεο εγγξαθέο Ξεθηλώληαο από κία ζύληνκε παξνπζίαζε ησλ θηλήηξσλ γηα κηα αξρηθή δεκόζηα εγγξαθή θαη κία αλάιπζε ησλ πξνβιεκάησλ ηεο αγνξάο ζρεηηθά κε ηελ ηηκνιόγεζή ηεο, θαιύπηεηαη εθηελώο ην ζεσξεηηθό ππόβαζξν πνπ έρεη αλαπηπρζεί γηα ηελ εμήγεζε ηνπ θαηλνκέλνπ

Γηα ηελ εκπεηξηθή κειέηε ηνπ θαηλνκέλνπ, επηιέρζεθε ε ακεξηθάληθε αγνξά αξρηθώλ δεκόζησλ εγγξαθώλ θαη πην ζπγθεθξηκέλα ε πεξίνδνο από ην 2003 έσο ην 2012 Χξεζηκνπνηώληαο θάπνηεο από ηηο εμεγήζεηο πνπ πξνηείλεη ε βηβιηνγξαθία, κειεηήζεθε ηόζν ν βαζκόο ηεο ππνηηκνιόγεζεο γηα ηε ζπγθεθξηκέλε πεξίνδν θαη ηε ζπγθεθξηκέλε αγνξά όζν θαη νη παξάγνληεο πνπ ηνλ επεξεάδνπλ

Έλα από ηα επξήκαηα ηεο εξγαζίαο είλαη όηη ε κέζε απόδνζε θαηά ηελ πξώηε εκέξα δηαπξαγκάηεπζεο είλαη ίζε κε 12% κε ηππηθή απόθιηζε 22.4% Δελ βξέζεθε ζηαηηζηηθά ζεκαληηθή δηαθνξά κεηαμύ ηεο ππνηηκνιόγεζεο γηα αξρηθέο δεκόζηεο εγγξαθέο πξηλ θαη κεηά ηελ παγθόζκηα νηθνλνκηθή θξίζε ηνπ 2008 Παξόιαπηά, δηαπηζηώζεθε ζηαηηζηηθά ζεκαληηθή δηαθνξά ζηε κέζε ππνηηκνιόγεζε κεηαμύ ησλ δηαθνξεηηθώλ εηώλ ζε επίπεδν εκπηζηνζύλεο 1%

Οη κεηαβιεηέο ησλ νπνίσλ εμεηάζηεθε ε επίδξαζε ζηελ ππνηηκνιόγεζε είλαη ην κέγεζνο ηεο εηαηξίαο, ε ειηθία, ν θιάδνο δξαζηεξηνπνίεζεο, ε πξνζθεξόκελε ηηκή, ην ύςνο ησλ θεθαιαίσλ πνπ αληιήζεθαλ, ε ρξνληθή πζηέξεζε κεηαμύ πξνζθνξάο θαη πξώηεο δηαπξαγκάηεπζεο, ην ρξεκαηηζηήξην, ν αξηζκόο ησλ ρξήζεσλ ησλ θεθαιαίσλ όπσο αλαγξάθνληαη ζην πξνζπέθηνπο, ε θήκε ηνπ αλαδόρνπ θαη νη ζπλζήθεο ηεο αγνξάο Γηα όιεο απηέο ηηο κεηαβιεηέο επηβεβαηώλνληαη νη ππνζέζεηο ηεο ππάξρνπζαο βηβιηνγξαθίαο,

κε εμαίξεζε ηε ρξήζε ησλ θεθαιαίσλ, ηελ ειηθία θαη ηε θήκε ηνπ αλαδόρνπ γηα ηηο νπνίεο ηζρύεη ε αληίζεηε ζρέζε από απηή πνπ πξνηείλνπλ πξνεγνύκελεο έξεπλεο Επηπιένλ, έλα από ηα επξήκαηα ηεο εξγαζίαο είλαη όηη ε πξνζθεξόκελε ηηκή, ν αξηζκόο ησλ ρξήζεσλ ησλ θεθαιαίσλ, ε δξαζηεξηνπνίεζε ζηνλ θιάδν πςειήο ηερλνινγίαο θαη ε εγγξαθή ζηνλ NASDAQ είλαη νη κνλαδηθέο ζηαηηζηηθά ζεκαληηθέο κεηαβιεηέο Χξεζηκνπνηώληαο απηέο ηηο κεηαβιεηέο, δεκηνπξγήζεθε έλα κνληέιν 6 παξαγόλησλ κε ηε ρξήζε παιηλδξνκήζεσλ

Τν κνληέιν εμεγεί 9.78% ηεο ζπλνιηθήο δηαθύκαλζεο ηεο ππνηηκνιόγεζεο

Chapter 1: Introduction

Going public is an important decision in the life-cycle of a company Most companies go

public through an initial public offering (IPO) of shares to investors so as to raise capital

from the market on more favorable terms than if the companies were privately held The

pricing of the initial public offerings is a difficult task, as there is no observable market

price prior to the offering Since Ibbotson R G (1975) puzzled the researchers by first

documenting the large underpricing of the initial public offerings; great effort – both

theoretical and empirical – has been made to identify any patterns in underpricing as well

as to explain the phenomenon through classic and behavioral finance The most well

researched market is the IPO market in the United States (US) Most of the theories have

been first tested by researchers using data from the US IPO market

This thesis covers a review of the literature regarding the initial public offerings, focusing

on the underpricing phenomenon in particular Starting from a brief presentation of the

basic considerations for a firm going public, which is followed by an analysis of the market

problems for the pricing of IPOs, an extensive presentation of the theoretical explanations

of underpricing is made Given the little consistency in the reported findings in academic

literature regarding the correlates of underpricing, a discussion is attempted with the aim to

identify promising areas for future research

The empirical part of the thesis will focus on the US IPO market and more specifically on

the IPO activity from 2003 to 2012, the least researched period in the literature It is the

period following the technology bubble burst in the US (2000) and as reported by Gao,

Ritter, & Zhu (2013), it is characterized by a significant drop in the average number of

IPOs per year, with small firms having been affected the most By employing some of the

explanations suggested in the IPO literature, the extent of underpricing during this period

and the factors influencing underpricing are investigated Findings will be discussed in the

context of existing literature for the pre and post-bubble periods

In the next chapter, the literature on IPOs and the underpricing phenomenon is reviewed

Chapter 3 includes a brief presentation of the US IPO market Data and methodology are

provided in chapter 4, together with the main results of the regression analysis In chapter

5, the findings are discussed and finally, chapter 6 provides a summary and the conclusions

Chapter 2: Literature Review

2.1 Introduction

The aim of this chapter is to situate the dissertation focus within the context of the wider research regarding the initial public offerings and to exhibit what knowledge and ideas have been established on the topic At the end of the chapter, a critical review of the literature will be attempted including the identification of its strengths and weaknesses as well as gaps that could be the basis for future research

2.2 Going Public: Basic Considerations

2.2.1 Why and When Firms Go Public

The academic literature presents a number of theories to explain why firms choose to go public Most of these theories are based on the efficient market hypothesis linking the decision to go public with managers‟ aim to maximize firm value First, raising capital to fund investment opportunities in a way that minimizes company‟s weighted average cost of capital is one of the key drivers for firms to go public (Scott, 1976) (Modigliani & Miller, 1963) According to the pecking order theory (Myers, 1984), firms decide to go public to gain access to capital when other cheaper sources of capital (internal equity and debt financing) have been depleted Second, the other key driver, which is not necessarily consistent with either the market efficiency or the value-maximization hypotheses, is to allow insiders including venture capitalists to cash out and perhaps diversify their holdings (Zingales, 1995) Third, according to Zingales (1995) going public can facilitate future acquisitions as it is much easier for a potential acquirer to spot a potential takeover when it

is public In the same context, going public creates public shares which can be used as a

“currency” in a takeover deal (Brau & Fawcett, 2006) Forth, going public may have strategic drivers such as to disperse the ownership of the firm (Chemmanur & Fulghieri, 1999), improve its position in the product market and deter new entrants into the industry (Jong, Huijgen, Marra, & Roosenboom, 2012), increase reputation and inspire more faith in the firm from other investors, customers, creditors and suppliers (Maksimovic & Pichler, 2001) and attract analyst attention and coverage

In a survey of Chief Financial Officers (CFOs) in US public and private firms, performed

by Brau & Fawcett (2006) and contrary to the standard theories, neither cost of capital nor capital availability motives ranked high among the drivers of going-public The creation of

public shares for use in future acquisitions, the establishment of a market price/value of the firm, the enhancement of the reputation of the company and the dispersion of ownership were among the top drivers for IPO Bancel & Mittoo (2009) who surveyed CFOs from 12 European countries identified that enhanced visibility and financing for growth are the most important drivers with all other being similar to those of US CFOs The only difference is

on the outside monitoring which is considered a major benefit for EU CFO‟s but a major cost for US CFOs

As regards the timing of an IPO, there are two main explanations proposed by the theory; the life-cycle and the market timing theories According to the life-cycle theory, companies

go public when they need to raise equity to fund growth and create a public and liquid market for firm ownership (Brau & Fawcett, 2006) According to market timing theories, a firm goes public in a bull market so as to achieve a more favorable pricing (Lucas & McDonald, 1990), avoids issuing in periods where few other good-quality firms issue (Choe, Masulis, & Nanda, 1993) and try to benefit from “windows of opportunity”, when investors are overoptimistic (Ritter & Welch, 2002) The market timing theory suggests that during this “window of opportunity”, high investor sentiment results in temporary industry or market-wide overvaluation of shares and lower cost of equity Finally, firms decide to go public when information asymmetry between firm insiders and investors is not severe otherwise it can be very costly to issuers (Draho, 2004)

2.2.2 How Firms Go Public

For a firm to go public, there are a number of steps/decisions that need to be made which are more or less common in all equity markets These steps are summarized by Ljungqvist

A (2008) as follows: the choice of the market, the choice of the underwriter, the initial price setting, the production of initial prospectus, the marketing of the offer, the final pricing of the offer, the allocation of shares and the aftermarket services

First, the firm needs to choose the market where its stocks will be traded In the past, companies would almost always choose to have their shares traded on their domestic stock exchange but this has changed over the years with the globalization of the markets and the irrelevance of the national boundaries in investment decisions In addition, many traditional exchanges have created new segmented markets aiming, for example, at young firms with high growth potential and characterized by less stringent entry requirements

Second, the firm needs to select an IPO team consisting of a lead investment bank (the lead underwriter), a public accountant and a law firm The team will work very closely with the firm going public to perform the due diligence investigations, to produce the information required by the regulatory authorities, to define the issue price, to prepare the prospectus and to attract potential investors In other words, according to Jo, Kim, & Seok Park (2007), the economic role of underwriter is to reduce market frictions, such as “information asymmetries” and “agency problems” that otherwise increase the cost of capital This

“certification” hypothesis about the economic role of the underwriter is not evidenced by Chemmanur & Krishnan (2012) who support a “market power” hypothesis indicating that the role of the underwriter is to obtain the highest possible valuation for the IPO The underwriter – investor relationship is very important particularly in the IPOs with stronger demand, IPOs of less liquid firms, and deals by less reputable underwriters (Binay, Gatchev, & Pirinsky, 2007) Firms choose underwriters based on their reputation, expertise and quality of research in the industry the firm operates Existing theoretical literature e.g Habib & Ljungqvist (2001), has focused on the issuer‟s choice of underwriter as one sided mechanism while Fernando, Gatchev, & Spindt (2005) argue that issuers and underwriters associate by mutual choice and that underwriter ability and issuer quality are complementary

Third, a key decision in the offering is the offer price The initial offer price is defined by the analysts from the lead underwriter and represents a first “best guess” of the market value of the company Valuation techniques such as the discounted cash flow and the peer group analysis are employed There are three price setting mechanisms analyzed in the literature; the fixed-price mechanism where the price is defined before investors are approached in a formal way, the book-building method and the auctions In book building, the underwriter “builds a book” prior to finalizing the terms of the offering, in an attempt to gather information about the market demand for the issue, with participants being asked to provide nonbinding indications of interest in the issue It is this demand information that will be used to determine the size, final price and allocation of the offering and that leads Benveniste & Wilhelm (1997) to consider book building as more efficient compared to auctions In a more recent study by Sherman (2005), it is argued that book building lets underwriters manage investor access to shares, allowing them to reduce risk for both issuers and investors and to control spending on information acquisition, limiting either underpricing or aftermarket volatility In addition to the above, compared to auctions, book-

building reduces issue costs for large issuers while auctioning is less costly for small issuers (Kutsuna & Smith, 2004)

Forth, the initial information gathering phase ends with the publication of the initial prospectus (“red herrings” in the US and “preliminary or pathfinder prospectuses” in other countries) Only when the price and issue size are defined is the final prospectus produced According to the guidelines issued by PwC‟s in its “Roadmap for an IPO” in the US (PwC, 2010), a typical prospectus should include a summary describing the company and its business, the risks associated with the business, the use of proceeds, the dividend policy and its restrictions, the capital structure of the company prior to and after the offering, any dilution results when there is a disparity between the IPO price and the net book value per share, the underwriting and distribution of securities issued, company‟s financial information, the corporate governance and executive compensation and the management‟s assessment of company‟s financial condition

Fifth, after having produced the initial or final prospectus, the issue is marketed to investors through road shows, investment banks contacting their clients, press briefings, internet alerting services and advertising For fixed price offers, the marketing phase helps to elicit bids from investors while for book building offers, the marketing phase gathers demand information which will be used for determining the final price and offer size

Sixth, after having defined the final price for the offer, the final stage involves the allocation of the shares The allocation of the shares depends on the type of the price setting mechanism used and on the oversubscription or undersubscription for the offering Many countries have “fair” allocation rules with all bids to be scaled down pro rata until supply equals demand (especially when a fixed price method has been used) while in the case of book building, share allocation is at the complete discretion of the investment bank or reflects the preferences of the issuing company regarding the types of initial investors it wants to attract On the other hand, in auctions, both pricing and allocation decision rules are usually announced ex ante

Finally, in most cases, the role of the underwriter finishes when stocks start trading In some markets, though, the investment bank can provide further services such as the share price stabilization in the aftermarket in the event of a pressure or excess demand for the shares This service may be provided up to 30 days after the stocks were first traded in the stock market The investment bank may also commit to ensure liquidity for the stock in the

aftermarket or may provide continued analyst coverage for a specified period of time after the IPO

2.2.3 Costs of Going Public

According to Ljungqvist A (2008), the costs of going public can be grouped into two major categories; direct and indirect costs

Direct costs are borne by the company in preparing for the IPO and include fees paid to auditors and lawyers, the cost of conducting the marketing road show and the opportunity cost of management time Costs also depend on the IPO method itself as direct costs of book-building are twice as large as direct costs for fixed price offers (Ljungqvist, Jenkinson, & Wilhelm, 2000) Some of these costs such as the cost of marketing, legal and auditing work are essentially fixed and other costs such as selling commissions and underwriting fees are variable and proportional to the amounts raised although for larger issues lower fees are negotiated As a result, the costs of going public (expressed as a percentage of the amounts raised) fall with the size of the issue

IPO direct costs differ significantly across countries mainly due to the considerably different average size of IPOs, the issue methods and required marketing activities According to Ljungqvist, Jenkinson, & Wilhelm (2000), marketing in the US adds around 1.4 per cent on average to the costs and listing in the US adds a further 0.4 per cent to the gross spread In addition, the “Anglo” capital markets – including the UK, Singapore, Malaysia and South Africa, have significantly lower costs than average, while Germany, Sweden, Italy, Canada and Israel significantly higher These differences are most likely related to the state of competition among the financial intermediaries, the regulatory and legal environment

Looking inside the US IPO market, in their article titled “The Seven Percent Solution” Chen & Ritter (2000) report that 90 per cent of the deals from 1995 to 1998 raising 20-80 million dollars have variable costs (including management fee, underwriting fee and selling concession) of exactly 7% of the gross proceeds This raises questions about the competitiveness of the investment banking industry in the US

The major indirect cost of going public is the underpricing which will be extensively discussed in the next section Another indirect cost of going public is the loss of confidentiality According to Pagano, Panetta, & Zingales (1998) the disclosure rules of

stock exchanges require companies to disclose information such as R&D projects and marketing or corporate strategies that are crucial for their competitive advantage

2.3 Market Problems with the Pricing of IPOs

According to Ibbotson, Sindelar, & Ritter (1994), the pricing of IPOs is difficult This is because there is no market price prior to the offering and because many of the firms going public have little or no operating history If the price is set too low, the issuer “leaves money on the table” as he does not fully exploit the opportunity to raise capital If the price

is set to high, then the investor would not get sufficient return and might not accept to participate in the offering He would also avoid purchasing offerings from an underwriter with a record of overpriced offerings

The academic literature presents three anomalies associated with the IPOs; high initial returns (also known as short-run underpricing), cycles in both the volume of new issues and the magnitude (also described as hot and cold IPO markets) and long-run underperformance

2.3.1 Short-Run Underpricing

The most well documented pattern associated with the process of going public is the large initial returns for those investing in IPOs of common stock The large initial returns of the IPOs are generated by the shares of the companies being offered to investors at prices considerably below the prices at which they trade on the stock market Underpricing is an important implication of an IPO, given that it involves a cost for the issuer as the latter essentially “leaves money on the table” for the benefit of those who are willing to invest in the company

There are two plausible reasons that can explain, at some extent, the underpricing of IPOs; transaction costs and risk According to Ljungqvist A (2008), some underpricing is required as a form of initial discount so that investors are enticed to change their portfolio and bear all the respective transaction costs that this change entails The underpricing can also be interpreted as a return for bearing the risk that the market price falls below the issue price after the shares start trading on the stock exchange With no underpricing, investors would prefer to buy shares in the aftermarket In addition to the above, various explanations can be found in the literature regarding IPO performance These will be extensively discussed in the next section of this study

Underpricing can be estimated in two ways First, as the percentage difference between the price the share is offered to investors and the price at the end of the first day of trading Second, underpricing can also be estimated as the “money left on the table” This is the difference between the price at the end of the first day of trading and the offer price, multiplied by the number of shares sold at the IPO

Table 1:

Equally weighted average initial returns for selected countries

Source: Loughran, Ritter, & Rydqvist, Initial Public Offerings: International Insights, 1994 -

Table updated in 2013 by Jay Ritter and retrieved from his website: http://bear.warrington.ufl.edu/ritter/

Literature provides evidence of underpricing in almost every country with a stock market, although the extent of underpricing varies from country to country Underpricing in US averages between 10-20 percent The extent of underpricing in Europe is also substantial

As observed in Table 1, there are cross-country differences which are most probably related

to differences in the institutional framework Shi, Pukthuanthong, & Walker (2013) show that the extent of underpricing is negatively associated with the stringency of IPO disclosure requirements, after controlling for various country and firm specific factors In general, countries with developed capital markets have more moderate underpricing than emerging markets (Ritter, 1998)

2.3.2 Hot & Cold IPO Markets

As academic literature suggests, there are patterns in both the volume and the average initial returns of IPOs Underpricing is auto correlated, the level of activity is also highly auto correlated and the two series, underpricing and level of activity, are positively auto correlated This apparently implies that firms prefer to go public when they are less able to obtain full pricing The periods of high average initial return and rising volume are known

as “Hot Issue” markets

Ibbotson, Sindelar, & Ritter (1994) propose three possible explanations for the hot issue markets; changes in firm risk, positive feedback or “momentum” strategies and windows of opportunities First, if there are periods during which riskier firms go public, these periods will be characterized by higher average initial returns, as riskier firms tend to be more underpriced than less risky firms Second, investors may follow “positive feedback” strategies assuming that there is positive autocorrelation in the initial returns on IPOs and biding up the price of an issue once it is listed if the price of previous issues has also risen Finally, hot issue market cycles may follow the cycles in investor sentiment During periods when investors are especially optimistic about the growth potential of the firms, firms would try to take advantage of it and offer the new issues at a price below the aftermarket price (Ljungqvist, Nanda, & Singh, 2006)

Yung, Colak, & Wang (2008 ) propose an alternative explanation After a positive shock in

an economy, improving investment opportunities raise the price at which a fixed group of firms would be able to sell securities These higher prices increase the willingness of firms

of relatively low quality to go public In a recent study though, Helwege & Liang (2004) argue that hot and cold IPO markets do not differ in the characteristics (profits, age, growth

potential) of the firms going public Hot & cold markets are not related to changes in adverse selection costs and managerial opportunism but are mainly driven by greater investor opportunism

2.3.3 Long-Run Underperformance

The long-run underperformance of IPOs has received considerable attention in the literature leading to controversial results and findings As a general rule, the change from excess positive to negative returns appears to take place within a few months after the listing (Thomadakis, Nounis, & Gounopoulos, 2012)

Three main explanations of the long-run underperformance problem can be found in the literature The “divergence of opinion hypothesis” suggests that buyers of the offering are the most optimistic about the value of the IPO and their valuations are higher than those of the pessimistic buyers In the long run, more information becomes available and the divergence of opinion between optimistic can pessimistic investors will narrow and the market price will drop (Miller, 1977) The “impresario hypothesis” suggests that IPOs are underpriced by investment bankers (the impresarios) so that they convey a message of excess demand to the market Subsequently, the long-run performance of IPOs should be negatively related to short-run underpricing (Shiller, 1990) Finally, the “windows of opportunity hypothesis” Loughran & Ritter (1995) suggest that firms going public in hot periods (characterized by investors‟ optimism regarding their prospects) are more likely to

be overvalued at the time of the IPO and present lower returns in the long-run

In addition to the above, other less popular hypotheses have emerged According to Teoh, Welch, & Wong (1998), firms manipulate their accounting numbers making their offerings more attractive and achieving higher that fair price which is not sustainable in the long-run

Ma & Shen (2003) suggest that underperformance of IPOs is not a puzzle The low probability outcomes of achieving high returns are valued more than in the standard expected utility setting As a consequence, even though long-run underperformance is evidenced, investors will still invest in IPOs given that they will be compensated by the prospect of gaining very high positive initial returns

2.4 IPO Underpricing Theories

There are four major categories of explanations/theories regarding the initial underpricing phenomenon and most theories have undergone rigorous empirical testing These are the

asymmetric information models, institutional explanations, ownership/control and behavioral explanations They mainly focus on the different aspects of the relations between investors, issuers and underwriters As per Ritter (1998), “these theories are not mutually exclusive” and “a given reason can be more important for some IPOs than for others” In addition, he argues that most of the theories are related to rational strategies by buyers while there are several other explanations involving irrational strategies

2.4.1 Assymetric Information Models

The term information asymmetry refers to the difference in the information available to the various groups involved in an IPO, but mainly the investors

The Winner’s Curse

The Winner‟s Curse hypothesis has been proposed by Rock (1986) Rock assumes that some investors are better informed about the true value of the shares As a consequence, they participate only in attractively priced IPOs with uninformed investors ending up with the less successful IPOs Keeping the uninformed investors in the market requires an additional premium which is the average underpricing of all IPOs In the same context, Ibbotson R G (1975) suggests that new issues are underpriced so as to “leave a good taste

in investors‟ mouths”

The main empirical extensions from Rock‟s Winner‟s Curse model are summarized by Ljungqvist A (2008) as follows:

(a) Adjusted for rationing, uninformed investors earn zero initial returns on average This

is what keeps them in the market Informed investors‟ returns just cover their costs of acquiring the information In this context, the share allocation rules applying in each market and their implications in underpricing has been extensively researched

(b) Underpricing is lower if information is distributed more homogeneously across investor groups Institutional and retail investors are the two major groups involved (c) The greater is the ex-ante uncertainty, the higher is the expected underpricing (Beatty

& Ritter, 1986) Various proxies have been used in the literature for ex-ante uncertainty These can be grouped into five categories: company characteristics (age, size, turnover, and industry), offering characteristics (gross proceeds, offer price, underwriting fee), prospectus disclosure (earning forecast, number of uses of IPO proceeds, number of risk factors), certification (established credit relationships, venture

backing, reputation of underwriter, auditor reputation) and aftermarket variables (daily after market returns, daily trade volume)

(d) Underwriters that underprice too much (too little) lose business from issuers (investors) (Beatty & Ritter, 1986) Therefore, underwriters should reach an average optimal level

of underpricing which will keep both issuers and investors happy

(e) Underpricing can be reduced by reducing the information asymmetry between informed and uninformed investors

Issuers, for whom underpricing is a cost, can take costly actions to reduce underpricing as long as marginal cost of reducing underpricing equals the marginal benefit Hiring prestigious intermediaries (underwriters or auditors) that certifies the quality of the issue and in this way, reduce the investors‟ incentives to produce their own information has been suggested as a way to reduce information asymmetry (Carter & Manaster, 1990) Chemmanur & Fulghieri (1994) support that investment bank‟s credibility as defined by their equity-marketing history is what makes it prestigious The role of prestigious underwriters, however, is still under discussion as some theories propose the opposite In another context, Booth & Smith (1986) suggest that promotional activity and underpricing are substitutes as promoting the issue can decrease the extent of adverse selection models with the increase in the fraction of uninformed investors participating in the issue

Signaling Theory

According to signaling theory, underpricing may be used as a means to convince potential investors for the “true” value of the firm A high-value company can be benefited from signaling its “true” value as this ensures that a higher price will be achieved at a sale at a later stage As the company‟s „true‟ value is revealed after the IPO, low-quality companies are deterred from mimicking and signaling due to the implications of the detection of their cheating in the aftermarket The risk of detection and the reduction in IPO proceeds due to underpricing are enough to deter low-quality firms from mimicking and signalling and separate them from high-quality firms

Signaling based models have been developed by Allen & Faulhaber (1989), Welch (1989) and Grinblatt & Hwang (1989) According to Allen & Faulhaber (1989), firms that underprice more are more likely to have higher dividends and the market reacts more favorably to their dividend announcements For Grinblatt & Hwang (1989), insider ownership as well as underpricing signal the firm‟s value This is because when owners

retain equity, they are supposed to expect high future cash flows For Welch (1989), high quality firms underprice at the IPO in order to obtain a higher price at a seasoned offering The basic assumption behind the latter is that underpricing is used by high –quality firms to add sufficient signalling costs to the imitation costs low-quality firms have to bear in order

With underpricing, underwriters try to reduce the risk of failing to sell all available shares

to investors as well as cover their marketing costs At the same time, since their fees are proportional to the gross proceeds, high level of underpricing will reduce their earnings from the IPO Therefore, the underwriter needs to take into consideration the resulting benefits and costs as well as investors‟ demand so as to determine the optimal level of underpricing The greater the uncertainty about the value of a firm, the greater the asymmetry of information between the issuer and the underwriter, the more important the role of the underwriter and the higher the level of underpricing

According to Baron & Holmstrom (1980), underpricing is a way for issuers to compensate underwriters for the use of their superior information In the same context, Michaely & Shaw (1994) argue that larger offerings require higher marketing and distribution effort, which, as a consequence, require higher underpricing Saunders (1990) suggests that the underwriters increase the level of underpricing so that they decrease the risk of undersubscription Cliff & Denis (2004) found that high reputation underwriters will be compensated with greater underpricing as a compensation for analyst coverage

Information Revelation Theories

In the context of the market feedback hypothesis, Ritter (1998) argues that in the case of bookbuilding, underpricing may be used by underwriters in order to make regular investors reveal information during the bookbuilding period In order for this mechanism to be effective, more underpricing is used in IPOs for which favorable information is revealed compared to IPOs for which unfavorable information is revealed

This has been initially suggested by Benveniste & Spindt (1989) who supports that there is

a secondary information asymmetry between institutional investors and the issuer, given that the former may have more information about the prospects for the company‟s industry, the prospects of the economy in which the firm operates, the demand in the IPO market as well as deals by similar companies One of the underwriter‟s key responsibilities is to ensure that the investor is no better off lying than telling the truth and her reward -in terms

of underpricing- will just reflect the marginal cost of her private information Additionally,

as investors and underwriters are in a constant co-operation for various IPOs, investors has less incentive to lie as lying may affect share allocation and underpricing in future IPOs This relationship, however, may raise concerns about underwriters treating some group of investors more favorably and gives auctions (where bidders remain anonymous) an advantage over bookbuilding Binay, Gatchev, & Pirinsky (2007) share this concern by arguing that underwriters favor institutions they have previously worked with and that regular investors benefit more from underpriced issues compared to casual investors

capital) as well as the amount of damages in the event of a lawsuit (the deterrence effect)

As per the US Securities Act of 1933, damages for investors in the IPO are based on the difference between the offer price and the share price at the time of the lawsuit

Ljungqvist A (2008) claim that the lawsuit avoidance hypothesis is mainly a US IPO market phenomenon, where strict laws exist More specifically, between 1988 and 1995 almost 6% of companies were sued for violations relating to the IPO and the damages awarded were about 13.3% of IPO proceeds The empirical results presented by Tinic (1988) based on a US sample IPOs that were brought to the market before and after the Securities Act of 1933 (which increased the control and introduced higher legal liabilities) also support the hypothesis In a more recent study using US IPO data from 1988 to 1995, Lowry & Shu (2002) found strong evidence that firms with higher litigation risk underprice their IPOs by a greater amount as a form of insurance and higher underpricing lowers expected litigation costs, without however being able to say if litigation risk has a first-order effect on underpricing (Ljungqvist A , 2008)

Researchers rejecting the hypothesis based on the average lawsuit settlement costs and low historical lawsuit frequency do not take into consideration that frequency and costs are low because most of the firms use underpricing as a form of insurance (Lowry & Shu, 2002)

Price Stabilization

In her paper “Underwriter Price Support and the IPO Underpricing Puzzle”, Ruud (1993) challenges the previous research on the grounds that positive initial IPO returns are primarily related to intentional underpricing and presents the underwriter price support hypothesis According to this hypothesis, since underwriters can support IPO prices by making transactions to prevent or delay a decline in the market price of a share, the number

of negative initial returns is reduced The distribution of the initial return can produce a positive mean initial return even if the offer price was set at the expected market value with

no underpricing at all

With price stabilization, underwriters can price an IPO aggressively, subsequently increase their fees and at the same time, convince investors that the issue has not been intentionally overpriced Additionally, price stabilization is an effective way to reduce the uninformed investor‟s winner curse (Ljungqvist A , 2008)

Tax-Based Motives

Dandapani, Prakash, Reside, & Dossani (1992) propose a model by which underpricing is dependent on issuer‟s personal tax rate on ordinary income, ability to defer capital gains and the proportion of retained ownership in the firm By underpricing, the issuer benefits from the deferment of the tax of a currently unrealized capital gain The tax will be deferred until the gain is realized In his study on Swedish IPOs, Rydqvist (1997) argues that the drop of underpricing from 41% (1980-1989) to 8% (1990-1994) is related to a tax introduced in 1990 which made underpricing related gains subject to income tax In this way, it removed the incentive to allocate underpriced stock to employees, which was a by-product of the pre-1990 law which taxed employment income much more that capital gains

In the US, capital gain taxes are typically lower than income taxes This can also generate

an incentive to underprice, especially for companies that rely on managerial and employee stock options (Ljungqvist A , 2008)

2.4.3 Ownership and Control

The two principal models to explain underpricing in the context of ownership and control are: underpricing as a means to retain control and underpricing as a means to reduce agency costs

Underpricing as a means to retain control

Also known as ownership dispersion hypothesis, underpricing may be used by firm‟s management or pre-existing shareholders as a means to retain control With underpricing, firms manage to ensure oversubscription and rationing in the share allocation process This allows them to discriminate between investors and, in this way, reduce the block size of new shareholdings The disperse ownership will make stock more liquid and make it more difficult to outsiders to gain control of the firm through the stock market (Brennan & Franks, 1997) This argument is particularly strong among firms with increased family involvement which underprice more at IPOs in order to avoid outside blockholders (Yu & Zheng, 2012) There are recent research studies (Field & Sheehan, 2004) (Hill, 2006), though, which reject the hypothesis suggesting that the future research is directed elsewhere

to find answers to the underpricing phenomenon

Underpricing as a means to reduce agency costs

Underpricing and rationing can also be seen as a means to reduce agency costs According

to Stoughton & Zechner (1998), underpricing & rationing can provide a mechanism for different classes of investors to be treated differentially In this way, institutional investors can be favored in share allocation process and consequently, be able to monitor the firm and thus lead to a higher intrinsic value The latter is supposed to reduce the amount of underpricing

2.4.4 Behavioral Explanations

Behavioral explanations are based on the assumption that underpricing is related to irrational investors who, by investing in the IPO shares, substantially increase the shares market value above their intrinsic value This can also be related to the fact that most IPO firms are young and therefore, hard to value by investors Another assumption which behavioral theories are based upon is that issuers, being subject to biases, fail to put pressure on the underwriters to have underpricing reduced Informational cascades, the investor sentiment and the prospect theory have been used in the academic literature to explain IPO initial underpricing

Informational Cascades

According to Welch (1992), the IPO market may be subject to informational cascades or bandwagon effects This means that investors do not act based only on the information they possess about the “true” value of the company issuing new shares in an IPO but are also influenced by the purchasing decisions of earlier investors As a consequence, issuers use underpricing so that they induce the first few potential investors to buy They also expect that later investors‟ bids will be influenced by the bids of earlier investors, rationally disregarding their own information This model is particularly effective if the distribution channels of investment banks are limited, as it will take the underwriter time to approach potential investors and give the late investors the time to gather information regarding earlier investors‟ decisions From investors‟ point of view, underpricing is required in order for them to early commit to the IPO and in this way, to start a positive cascade Cascades cannot exist in a market with no information asymmetry or for an IPO using bookbuilding

Investor Sentiment

The investor sentiment has been first introduced in IPO underpricing literature by Ljungqvist, Nanda, & Singh (2006) The mechanism explaining underpricing as well as its connections to the long-run IPO underperformance is as follows: Through the underwriter, the issuer allocates shares to “regular” institutional investors for gradual sale to sentiment investors who are driven by overconfidence or self-attribution or the belief that their knowledge is more accurate about the prospects of an IPO Assuming constraints on short sales, “regular” institutional investors maintain stock prices by restricting the availability of shares but take advantage of the surplus under the sentiment investors demand curve Through underpricing, they are compensated for the case of losses arising from the possibility that sentiment demand may drop It is evident that Ljungqvist, Nanda, & Singh‟s model does not focus on asset pricing anomalies but on the fact that firms act strategically

to take advantage of market‟s mispricing or misconceptions According to Derrien (2005), the IPO price chosen by the underwriter depends on both the intrinsic value of the company revealed by institutional investors and noise trader sentiment, with the latter being partially incorporated in the IPO price as the underwriter is concerned about the aftermarket behavior of IPO shares The level of initial return is positively related to the noise trader sentiment

Prospect Theory and Mental Accounting

Loughran & Ritter (2002) use prospect theory and mental accounting to explain why issuers don‟t get upset about leaving millions of dollars “on the table” They argue that issuers in an IPO will sum the wealth loss from leaving money on the table with the larger wealth gain on the retained shares from a share price increase This will produce a net wealth increase for the issuer As prospect theory suggests, issuers care about the change in their wealth rather than the level of wealth itself In the literature, the “partial adjustment phenomenon” is also used to explain underpricing Benveniste & Spindt (1989) argue that regular investors in order to reveal good information - through high demand for the issue - must be rewarded with more underpricing High demand, at the same time, means that there will be positive revision of the offer price and volume between the filing of the preliminary prospectus and the offer date However, underwriters only partially adjust the price upwards towards the true market price as they need to please regular investors So, the bad news that a lot of money was left on the table arrives together with the good news of higher

proceeds and a high market price That is why issuers don‟t complain In this context, Hanley (1993) documents that the relation of the final price to the range in anticipated prices disclosed in the preliminary prospectus is a good predictor of initial returns

2.5 How & Why Underpricing Changed Over Time

Apart from the variability of first day returns that has been associated with hot and cold markets, there has been empirical evidence of lower frequency movements Loughran & Ritter (2004) examine three hypotheses for the changes in underpricing over time: the changing risk composition hypothesis, the re-alignment of incentives hypothesis and the changing issuer objective function hypothesis

The changing risk composition hypothesis, introduced by Ritter, assumes that riskier IPOs will be underpriced more that less risky IPOs, with risk being related to either technological

or valuation uncertainty It is however considered very weak to explain much of the variation in underpricing over time given that there have been no much changes in the characteristics of the firms going public (Loughran & Ritter, 2004) The realignment of incentives hypothesis, introduced by Ljungqvist & Wilhelm (2003), assumes that variation over time is associated with issuers willingness to accept more underpricing or to bargain for a higher offer price Changes in CEO ownership, ownership fragmentation and frequency of “friends & family” share allocation are among the reasons for this Little support is also found in the literature for this hypothesis Finally, the changing issuer objective function hypothesis assumes that, with no changes in the level of managerial ownership and other characteristics, issuing firms become more willing to leave money on the table as (a) they are more concerned about hiring an underwriter that will offer high analyst coverage (analyst lust hypothesis) and (b) they actively seek underwriters with a reputation of underpricing due to spinning (underwriters setting up personal account to venture capitalist and issuing firm executives so that they allocate hot IPOs to them) The latter is supposed to explain the increased underpricing during the 1990s and the bubble periods in the US

Changes in the regulatory environment play an important role in the extent of underpricing and low frequency variations over time In a recent article, Johnston & Madura (2009) argue that initial returns of IPOs in the US has declines since the introduction of Sarbanes-Oxley act in 2002, as the latter imposes new requirements for firms going public which improved transparence and reduced the uncertainty surrounding their valuation Similar

conclusions can be drawn by the study of Rydqvist (1997) on Swedish IPO returns pre and post the tax legislation changes in 1990 This has been already analyzed in a previous section Furthermore, low frequency variations in IPO short-run underpricing can be influenced by restrictions in daily stock price fluctuation These restrictions are common mostly in emerging markets and aim at protecting the stock market and investors from irregularities and speculation attacks As an example, the Athens stock exchange had imposed a limit of ±8% for all shares including new from 1992 to 1996 After 1996, the regulation changed and there was no limit for the first three days of trading for a new stock The ceiling was made applicable again after the third day of trading and its level had been revised a couple of times till completely removed in 2000

In a recent study by Dolvin & Jordan (2008) and contrary to what has been already discussed regarding underpricing low frequency variations over time, it is argued that the percentage of shareholder wealth lost is surprisingly stable over time, unlike underpricing This is mainly associated with the fact that in periods of high underpricing, increased share retention has been observed, which consequently offsets much of the potential loss

2.6 Discussion on Literature Review

As extensively analyzed in the previous sections, the academic literature is abundant with theories explaining why IPOs are undervalued at the offer price compared to the first-trading day closing price, which is by definition the indication of its fair value Review of existing research suggests little consistency in findings across studies regarding the factors associated with underpricing

As summarized by Ljungqvist A (2008) in the “Handbook of Empirical Corporate Finance”, the empirical evidence generally supports that: (a) informed or institutional investors are those who gain from IPOs underpricing, all the rest experience little or no excess returns, (b) underpricing increases with the ex-ante uncertainty regarding the firm going public and (c) some investors are informed and they can influence the underwriter‟s choice for the offer price However, he suggests that information-related explanations cannot fully explain the low-frequency variation of underpricing over time

In addition to the classic theories regarding underpricing, new theoretical explanations have also been proposed An example is the divergence of opinion explanation by Booth & Booth (2010) They argue that initial underpricing is a natural by-product of liquidity-

motivated ownership dispersion requirements and divergence of opinion Given that liquidity is one of the main reasons a firm goes public, shares have to be widely distributed

at the IPO This is achieved by underpricing (to attract pessimistic investors) and rationing

of optimistic investors In the secondary market, optimistic investors who were rationed will bid up the share price by buying shares from pessimistic investors

There have also been studies which cast doubt upon the underpricing phenomenon itself Purnanandam & Swaminathan (2004) argue that the issuers and the underwriters underprice their IPO relative to some maximum price they could achieve, given the demand they observe The “true” value of the share is not identical to the first-trading day price, but to the long-run fair value of the firm In this context, using a sample of more than 2,000 IPOs from 1980 to 1997, they conclude that the median IPO was significantly overvalued at the offer price by 14% to 50% compared to its industry peers if valuation price multiples were employed

Although some of the theoretical frameworks developed are criticized by recent studies, one can conclude from the literature reviewed that there is no single theory that can fully explain the phenomenon Depending on the market and the time period analyzed, more than one explanation can apply What has not been sufficiently covered by the literature yet is the quantification of the relative importance of the different explanations for underpricing for different firm categories, countries or periods

Driven by the debate between the Benveniste & Spindt (1989) and the agency model attributed to Jay Ritter, Ljungqvist A (2008) suggests that future literature should focus on the behavioral approaches to explain underpricing variations over time, cross-country differences in institutional framework and their impact on underpricing, the conflicted behavior of investment banks during the market boom and auction mechanisms to price and allocate IPOs In the same context, Ritter & Welch (2002) argue that information asymmetry theories cannot explain the very high levels of initial returns and therefore the future literature should focus more on agency conflicts as well as share allocation issues and behavioral explanations

Chapter 3: US IPO Market

3.1 Introduction

The aim of this chapter is to briefly describe the main characteristics of the US IPO market, highlight any difference against the basic IPO processes described in a previous section, as well as any market specific legal requirements or restrictions In this way, implications for underpricing will be identified and conclusions useful for the empirical part of this study will be drawn

3.2 IPO Process in the US – Statutory Framework

There are two principal US federal statutes which govern securities transactions in the US; the amended Securities Act of 1933 and the amended Securities Exchange Act of 1934 The

1933 Act requires that a registration statement (Form S-1 and Prospectus) is filed with the Securities and Exchange Commission (SEC) before any offer of securities is made and is declared effective by the SEC before any sale is made (Securities and Exchange Commission, 2013) In order to declare the offer effective, SEC concentrates on disclosures that appear to conflict with the rules or the applicable accounting standards and on disclosures that appear to be deficient It does not advise whether the security is fairly priced or not

There are three critical periods during an IPO process First, the pre-filing period during which no security can be offered or promoted unless registration statement with the SEC has been filed Second, the waiting period (post-filing but pro-effective) during which certain kings of offers but no sales can be made Third, the post-effective or selling period during which sales can be made

The key documents that the issuer should prepare during the IPO process are: issuer‟s registration statement and prospectus, issuer‟s NASDAQ or NYSE listing application, issuer‟s board approval of the IPO, issuer‟s post-IPO articles of association, underwriting agreement, road show material etc Apart from the preparation of the registration statement

as previously described, there are many regulatory and disclosure issues that the firm has to address, including issues related to the Sarbanes-Oxley Act of 2002 (internal controls, audit committee, board of directors, auditor relationships and code of ethics), current accounting procedures (stock-based compensation, liability versus equity classification, preferred stock and debt, revenue recognition, consolidation issues etc.) and taxations issues (PwC, 2010)

US IPOs use bookbuilding for pricing and volume setting During and immediately after the roadshows, the underwriter contacts potential buyers and records who is interested in buying how much and at what price Auctions are very low in number totaling in 22 cases since 1999

3.3 US IPO Market Activity & Underpricing in the Literature

US IPO market has drawn the attention of the researchers as it has been considerably active More specifically, during 1980-2012 an average of 240 companies per year went public in the US Total gross proceeds during this period reach 726 billion dollars (Ritter, Initial Public Offerings: Updated Statistics, 2013)

Table 2:

Number of IPOs and aggregate proceeds per year (1980-2012)

Year

Number of Initial Public Offerings

Aggregate Proceeds (billion dollars)

Year

Number of Initial Public Offerings

Aggregate Proceeds ( billion dollars)

Source: Table reproduced using IPO data available by Jay Ritter (Ritter, 2013)

As presented in Table 2 and observed by Gao, Ritter, & Zhu (2013), the IPO activity has dropped considerably after the technology bubble in 2000 More specifically, from an average of 310 companies per year that went public in the US during 1980-2000, the average number of new issues dropped to 99 during 2001-2012 The drop has been more intense among the small firms It has been attributed to the introduction of the Sarbanes-