Empirical Evidence on IPO-Underpricing

Table of Contents

List of Tables

List of Abbreviations

1 Introduction

2 Initial Public Offerings
2.1 The IPO Process
2.1.1 Institutions and Roles in IPOs
2.1.2 IPO Transactions
2.1.3 International Equity Issuing Conditions
2.2 Empirical Evidence on IPO-Underpricing
2.2.1 Definition of IPO-Underpricing
2.2.2 Evidence on Underpricing and Country Differences
2.2.3 Long-Term IPO Performance Studies

3 Theories explaining IPO-Underpricing
3.1 Institutional Explanations
3.1.1 Reduction of Legal Liability
3.1.2 Analyst Compensation Theory
3.1.3 Banking Relationships
3.2 Models based on Ownership and Control
3.2.1 Reduced Monitoring Theory
3.2.2 Increased Monitoring Theory
3.3 Models based on Information Asymmetry
3.3.1 Information Asymmetry between Issuer and Investor
3.3.2 Information Asymmetry between Investor Groups
3.3.3 Information Asymmetry between Underwriter and Investor
3.4 Theories related to Individual Rent-Seeking Behaviour
3.4.1 Underpricing to the Benefit of the Management
3.4.2 Underpricing to the Benefit of the Underwriter

4 The Model
4.1 Sample Description
4.1.1 Investigation of Data
4.1.2 Descriptive Characteristics of Sample Firms
4.2 The Regression Model
4.3 Results of the Regression Analysis
4.4 Interpretation of Regression Results
4.5 Alternative Interpretation

5 Conclusion

List of Appendices

Appendix

List of References

List of Tables

Table 1: Derivation of sample size

Table 2: Industry distribution of sample firms

Table 3: Company characteristics

Table 4: Transaction characteristics

Table 5: Underpricing for different time periods

Table 6: Weekly breakdown of underpricing, closing spread and turnover values

Table 7: Explanatory variables and assumed relation to underpricing

Table 8: Regression results

List of Abbreviations

Abbildung in dieser Leseprobe nicht enthalten

1 Introduction

Objective of the paper

The decision to go public marks a milestone in the life of a company. Its implications for the firm are manifold, as it provides access to public equity capital, enables existing shareholders to diversify their investments and raises the level of public attention. However, the company also becomes obligated to conform to certain transparency standards and disclosure requirements, which impose a challenge on the management.^[1] Most companies entering the equity market do so via an initial public offering (IPO), wherein the valuation of the newly issued shares has been subject to academic debate for decades.^[2] Researchers empirically discovered that offer prices tend to be undervalued, resulting in a price jump on the first day of trading.^[3] This phenomenon, termed IPO-underpricing, is consistently observable across countries and time periods, but with varying magnitudes.^[4] Underpricing reached a peak during the years 1998 to 2001, commonly known as the dot-com period.^[5] Seemingly, underpricing comes as a cost to existing shareholders, because it fundamentally dilutes the value of the pre-IPO shares. Based on the supposition that this is not a favourable outcome for these shareholders, the search for a satisfactory answer to this apparent incongruity has led to a broad array of theories explaining its occurrence.

This paper has the objective to relate the most well-established theories to an empirical analysis performed on the German IPO market with the intention to obtain a greater understanding for the motives of underpricing in Germany. In doing this, it outlines empirical findings on underpricing and performs a thorough literature review on the different approaches explaining these observations. With those findings, it builds a regression model based on a comprehensive dataset of IPOs in Germany and tests theoretical models based on two forms of information asymmetry. In addition, this paper provides an alternative explanation based on behavioural patterns of market participants during the turn of the century, which seems to be applicable for this period.

Organisation of the paper

The paper starts with outlining the essential steps in an IPO process and describes the functions and motives of its participants. A comparison of international equity issuing regulations and practices helps the reader to determine in how far underpricing observations can be compared across countries. Next, the term underpricing is defined, and evidence on underpricing across countries and time periods is compiled. In order to distinguish between the concepts of underpricing and overvaluation, the results of various long-term IPO performance studies are investigated and placed into the context.

In the following step, different theories explaining IPO-underpricing are presented to the reader. The theories are clustered into four groups which differ in their underlying fundamental assumptions. All of them strictly refer to rational behaviour of the IPO participants. Note that it is not attempted to take account of all underpricing theories, as this would exceed the scope of this paper.

Subsequent to this academic section, a data sample of 371 IPOs on the Frankfurter Börse is analysed with reference to its company and transaction characteristics, especially to the level of underpricing. This data is then used in a regression analysis, relating information asymmetry-based underpricing explanations to the findings in the sample transactions. The regression is tested for consistency by various statistical measures and the results are described.

In its last section, the paper links the findings of the regression analysis to the previously described underpricing theories and interprets the results. Because initial returns of IPOs at the turn of the century exhibit rather peculiar patterns, alternative explanations referring to the area of behavioural finance are given.

2 Initial Public Offerings

In going public, an issuing firm undergoes a rather complex process which must meet the requirements and expectancies of the capital markets, and at the same time maximise the desired benefits for the issuer. However, the listing process is associated with several risk factors, which must be evaluated before the IPO initialisation. In the following, the IPO process is be presented with respect to its major participants and country particularities. Then, the underpricing phenomenon is described and put into a temporal and cross-country context.

2.1 The IPO Process

2.1.1 Institutions and Roles in IPOs

Issuing company

An initial public offering, also named primary offering, describes the first issuance of stock for public sale of a private company.^[6] Prior to an IPO, the issuing firm has to comply with explicit listing regulations such as legal and formal requirements set by the respective stock exchange authority. In addition, market participants expect non-explicit qualities like the comprehensibility of the equity story as well as a clear organizational and ownership structure.^[7]

An IPO has certain implications for the issuing firm, which can be grouped into financial and non-financial aspects. Empirical studies have shown that the main financial reason to go public is to strengthen the equity base with the purpose of financing company growth. In many cases, this is a prerequisite for higher future leverage. A second financial benefit in going public is in enabling the company to make use of capital market instruments like corporate bonds and to potentiate restructuring measures like Spin-Offs or Equity Carve-Outs. Noteworthy with reference to non-financial benefits is the increased public exposure, which yields positive effects on the company’s attractiveness to its employees, customers and suppliers.^[8]

Besides mentioned benefits, going public bears disadvantages and sources of risk: augmented disclosure requirements, one-time costs (provisions and marketing costs) as well as recurring costs (preparation of shareholder documents and organization of meetings) are the most relevant drawbacks. These costs can make up to 10% of the whole issue volume.^[9]

Investment bank

The main task of an investment bank in an IPO transaction is the development and execution of the issuing strategy.^[10] In its intermediary role, the bank communicates between the investor community and the issuer, merging both parties’ interests. Besides the generation of information about investor demand, it consults the company management in aspects about the preparation of the IPO, elects consortium banks and actively manages the entire offering process.^[11] Concerning the internal bank structure, the Equity Capital Market Division (ECM) is responsible for the fundamental pricing and placement of the shares as it is in close contact to institutional and large private investors. The Research Division closely monitors the market and must therefore also be integrated into the IPO process.^[12]

Investors

For the investment bank, the selection and analysis of potential investor groups is essential, as investor demand ultimately determines the success or failure of the offering. The specific risk-return relationship of the respective share needs to be evaluated and matched to the proper investor segment.

In general, there are two investor types: the first type, the institutional investor, professionally manages large capital volumes, gives pricing feedback to the investment bank and represents the major trading partner at the time of issuance.^[13] As a result, investment banks have a tendency to allocate IPO shares favourably to institutional investors.^[14] The second type, the retail investor, typically invests its own funds and has limited expertise and experience in doing so. Retail investors oftentimes neglect their voting rights and are typically longer-term shareholders. Their demand is more transparent, as it is strongly influenced by analyst reports and public media.^[15]

2.1.2 IPO Transactions

The course of initial public offerings

Having decided upon the step to perform an IPO, the issuing firm selects an investment bank to prepare the IPO process. This selection takes place via a so-called beauty contest and a pitch, in which various investment banks compete against each other to take the lead position or to become one of the consortium banks.^[16]

In a first step towards the listing process, the investment bank consults the company about equity financing implications, the current market condition and files a registration statement with the stock exchange supervisory authority, for example the Security Exchange Commission for placements inside the United States.^[17] This document includes financial information on proposed financing measures, business-related data, accounting statements, and a management forecast. While the authority then analyses this document during a so-called waiting period, the investment bank distributes a preliminary prospectus to few institutional investors for an initial evaluation of their level of interest.^[18] After the exchange authority declares the registration effective, the investment bank performs a roadshow, which is marketing the offer to institutional investors and generating feedback in form of indications of interest.^[19] This process is called bookbuilding, as the indications are collected in a book. Besides the current capital market condition, this information is the basis for setting the price range, which is finally done by the ECM Division of the bank. The bookbuilding phase serves the important function of obtaining information about the demand for the IPO shares, effectively reducing the level of risk both for the investment banks and the issuer.^[20] The proposed offer price is then presented to the company management. In a last step, the shares are allocated to the investors and trading begins.^[21]

Mechanisms of share allocations

Ultimately, there are two methods of share allocations, namely the underwriting method and the best effort method.^[22] In the underwriting method, “[…] the investment bank (or a group of investment banks) buys the securities for less than the offering price and accepts the risk of not being able to sell them.”^[23] Herein, the difference between the underwriter’s buying price and the offering price is called the gross spread and reflects the basic compensation of the bank. In the best efforts method, the investment bank does not purchase the shares, but acts as an agent receiving a certain commission for each share sold. The bank is legally obligated to sell the issue using its best efforts.^[24]

Greenshoe provision and lockup agreement

Greenshoe provisions have become common clauses in underwriting contracts. They give the underwriter the option to cover excess demand or oversubscription by issuing additional shares, typically 15% of the total volume.^[25] Greenshoe provisions can be seen as a contingent compensation to the investment bank because it will only exercise the option if the share price increases, and by arbitrarily selling the shares making a riskless profit. The greenshoe normally extends to a period of 30 days.^[26]

Besides the greenshoe provision, lockup agreements are regularly featured in IPOs. They restrict insiders and pre-IPO shareholders from selling their shares for a certain time frame, typically 180 days.^[27] The details of the agreements including its duration are important articles in the IPO prospectus.^[28] Lockup agreements have the functions of obliging key employees to the firm, signalling trust in the company’s long-term performance, and of temporarily constraining the supply of shares.^[29] On the unlock day, so the apprehension of investors, lockup agreements may cause a drastic increase of trading given the information asymmetry between the traders.^[30]

2.1.3 International Equity Issuing Conditions

Accounting standards

Companies listed on the German stock exchange usually base their reporting on the International Financial Reporting Standards (IFRS), whereas in the U.S. the Generally Accepted Accounting Principles (U.S. GAAP) are used.^[31] In order to detect the influence of the respective accounting standard on the IPO process, the degree of disclosure and the disclosure quality which these standards generate need to be examined, as they determine the information asymmetry between the issuer and the analyst.^[32] The majority of academics studying the effect of accounting standards on disclosure state that IFRS are basically equivalent to U.S. GAAP.^[33] Only few researchers oppose to this view. Hence, one can conclude that no significant relation between the choice of the accounting standard and information quality exists, two important factors for information asymmetry in the IPO-underpricing discussion.^[34]

Practices and legal restrictions

With respect to the allocation and pricing process, one could observe an international harmonisation of procedures in which bookbuilding replaced fixed price offerings and IPO auctions.^[35] However, prior to 1995 bookbuilding was not used in Germany and all issues were performed using fixed pricing.^[36] Fixed pricing is an approach, in which the offer price is set before indications of interest are collected.^[37] Placing a buying order, investors engage into an obligation to purchase the respective amount of shares. In case the aggregated demand for the shares exceeds the issue volume, the underwriter scales the orders, a process called rationing. Fixed price offerings have lost acceptance because the information exchange is low, increasing ex-ante investor uncertainty.^[38]

In the U.S., bookbuilding is based on a pre-set price range, which is always 2 US$ (irrespective the issuing volume and price). However, the issuer may decide to fix the final offer price 20% above or below the range without further announcements. In Germany, the price range is not set before first indications of interest are collected. Also, the price range is not restricted to any currency value.^[39]

U.S. stock exchange regulations require a so-called quiet period, which prohibits affiliated and non-affiliated analysts to publish reports or trading recommendations for up to 40 days after the IPO date.^[40] In Europe, quiet periods do not exist and coverage before and directly after the IPO is permitted.^[41]

With respect to the legal systems, it is noteworthy to point out that strike suits are permitted by law in the U.S.. These are lawsuits, in which a plaintiff (a shareholder in this case) represents the joint shareholder base, and acts on their behalf. In case the lawsuit is successful, the sued party has to reimburse all disadvantaged shareholders at once.^[42] In the U.S., most of the listed firms invest in lawsuit insurance,^[43] which limits the damages in case of an unsuccessful litigation.^[44]

Concluding, one can state that despite certain dissimilarities, equity issuing conditions in the U.S. and Germany are alike, and this insight enables the author to transfer findings from the U.S. to the German equity market.

2.2 Empirical Evidence on IPO-Underpricing

Given the assumption that investment banks have certain expertise in the pricing of new issues, they should be able to find the fair value of a firm and set the offer price accordingly. Furthermore, from the investors’ point of view, there is no rational incentive to pay any other than this value on the secondary market. However, numerous academic researchers have shown that an abnormal return on the first trading day, as well as a certain pattern on the long-term performance of IPOs has been consistent over time.^[45]

2.2.1 Definition of IPO-Underpricing

The term underpricing originates from U.S. academic literature dating back to the 70s and basically describes the phenomenon of a price movement of newly issued shares after the offering. However, in order to operationalise the term underpricing, a more detailed definition had to be developed. Most literature describes underpricing as the difference between the fair value of the company and the offer price, making the price movements on the first trading day the adaptation process to the fair valuation. Based on the assumption that the actual firm value is known to the issuer, the concept implies the issuer’s deliberateness to underprice.^[46] However, the actual scope of this assumption is relativised in chapter three, where potential motives for this behaviour are explored.

For studies exposed in the following, underpricing is defined as the closing price on the first trading day^[47] (P1) minus the offer price (P0), divided by the offer price.^[48] The term Money on the Table describes the spread between the aftermarket trading price and the offer price, multiplied by the number of shares sold in the IPO and provides a currency value rather than a percentage value.^[49]

Abbildung in dieser Leseprobe nicht enthalten

2.2.2 Evidence on Underpricing and Country Differences

IPO activity

IPO activity tends to cluster in certain time periods, thus it appears in waves.^[50] In these waves, so-called hot IPO markets, IPOs show distinct particularities.^[51] The first ones who related IPO waves to underpricing were Ibbotson/Jaffe (1975), detecting certain years in the 60s and 70s where IPOs generated very high initial returns. They conclude that underpricing follows a distinct pattern.^[52] Based on their discovery, Ritter (1984) advances with this finding and applies it to the year 1980 in his article “The “Hot Issue” Market of 1980”, which gives ground-braking insights and is still referenced to in recent works.^[53] Ritter points out that underpricing is not solely a function of risk, but it depends on the time frame chosen, in which one industry type strongly participates in the IPO market.^[54]

On the U.S. market, the number of IPOs between 1980 and 2001 exceeded one per business day. These 6.249 offerings generated a total of over 448 billion US$. However, focusing on averages impedes the view for patterns. In the years 1996 to 2000 there was a drastic peak, in which up to 621 (1996) companies became listed per year. In this time frame, a total of 2.123 listings generated over 224 billion US$ in gross proceeds.^[55]

In Germany, 790 IPOs were performed in the time frame from March 1997 until January 2007, whereby 367 of these apply only to the years 1999 and 2000.^[56] Firms going public in these two years were predominantly technology-related companies (92%).^[57] In an international comparison, firms going public in Germany have a higher average firm age and a larger market capitalisation than their U.S. counterparts.^[58]

Evidence on underpricing in Germany

Consistent with the periodical pattern of IPO activity demonstrated above, the level of underpricing is also heavily influenced by the time period chosen in the respective research study. In a fairly recent and frequently cited analysis, Stehle/Ehrhardt (1999) investigate IPOs on the Frankfurter Börse between 1960 and 1995. One peculiarity about this study is that it only includes fixed price IPOs, as bookbuilding was not applied in Germany prior to 1995. The authors distinguish the listings with respect to their market segment (official market, regulated market, and open market). Because the time lag between the underwriting and the first trading is relatively small, Stehle/Ehrhardt disregard market movements as they assume their ultimate effect to be marginally small. As a result, they find an average initial return on the first trading day over the entire period of 15,79%. They observe the highest level of underpricing on the official market (16,04%), followed by the open market (12,86%) and the regulated market (8,35%).^[59] Also, the study reveals that very high returns are predominantly applicable to smaller and relatively unknown companies.^[60]

The effect of the time period chosen in a study can be pointed out by holding it against a study investigating the years 1996 to 2000 by Franzke (2003).^[61] For these years, she finds that IPO shares are underpriced by 49,81% and that the average amount of money left on the table amounts to 27.916 million €.^[62] Franzke split her data set in venture capital financed and non-venture capital financed IPOs and finds heavier underpricing for venture capital financed issues. The different results between the studies of Franzke and Stehle/Ehrhardt indicate the previously mentioned importance of the time frame on underpricing observations.

Evidence on underpricing in the United States

For the U.S. primary market, Loughran/Ritter (2002) investigate underpricing on a sample size of 6.169 firms for different time frames and find average initial returns of 7% between 1980 and 1989, 15% for 1990-1998 and up to 65% (1999) for the following years during the dot-com period. For the U.S. market, they conclude that underpricing has continuously increased over time, contradicting to Ritter (1984) stating that underpricing comes in waves.^[63]

Ljungqvist/Wilhelm (2003) detect even higher values for these years. For 1999, they estimate the mean initial return to be 73,3% with a standard deviation of 96,3% and a median of 39,5%. These values indicate that only few IPOs are underpriced to a great extend, leaving huge amounts of money on the table. Even though it is shown that the initial returns in the U.S. are consistently higher than in Germany, the periodical patterns pointed out are similar.

2.2.3 Long-Term IPO Performance Studies

Differentiation to underpricing research

Long-term IPO performance studies aim at detecting abnormal return patterns of IPO companies. As underpricing only concentrates on the price adjustments on the first trading days, the underperformance of IPO stocks is assumed to be observable in a longer time frame, typically 36 months.^[64] The importance in analyzing underperformance stems from the fact that statements on underpricing always imply its persistence. In other words, if post-IPO share prices are mean-reverting, this longer-term underperformance compensates the underpricing effect.^[65] The challenge in this kind of data analysis is to elaborate an adequate benchmark in order to distinguish between ex-ante overvaluation and actual underperformance.^[66]

To show IPO stock movement anomalies, long-term studies comprise the calculation of market-adjusted abnormal returns. The key is to create an adequate market portfolio, and to subtract its returns from the return of the stock under observation.^[67] Generally, researchers pick an index or create a comparable portfolio on size, market or other similar characteristics.^[68] Also, it is necessary to take dividend payments, purchasing rights and other pecuniary advantages into consideration.^[69]

Empirical evidence on the long-term performance of IPO companies

One of the most frequently cited and well accepted studies performed by Ritter/Welch (2002) investigates the average three-year buy and hold return of IPOs in the U.S. in the period from 1980 to 2001. The returns are computed by two different adjustment processes. The market adjusted returns are measured against a value-weighted (CSRP) index of AMEX, NASDAQ and NYSE firms. The style-adjusted returns are calculated as the difference between the return of an IPO company and a style-matched firm.^[70] As a result of this analysis, Ritter/Welch identify an average 3-year buy-and-hold return of 22,6% on the IPOs, a market-adjusted return of -23,4% and a style-adjusted return of -5,1%.^[71] However, the averages hide the fact that specific periods had negative returns drastically above these average numbers. In 1999, the style-adjusted return amounted to -74,2%, which can predominantly be attributed to the strong underperformance and failure of technology-related IPOs during the dot-com period.^[72]

These results, which indicate a clear underperformance of IPO stock, are increasingly questioned by studies in recent years. Sapusek (2000) detects no significant level of underperformance of IPO shares on the German market in the time period from 1983 to 1993. Using the same method as Ritter/Welch (2002), he compares the stock returns with major market indices, price indices and value-weighted small cap and blue chip performance indices.^[73] However, he reveals that if initial returns on the first trading days are deducted from the calculation, a statistically significant but very low (1%) underperformance is observable. He concludes that the relative performance of IPO shares depends on the time period chosen, the benchmark applied and the calculation method.^[74]

In another recent study Stehle/Ehrhardt (1999) come to a similar conclusion. They observe a style-adjusted 3-year abnormal return of -5,04% and a market-adjusted abnormal return of 1,54%^[75] The authors support their results of the higher IPO performance in Germany than in the U.S. with the fact that German IPO firms tend to be more mature and larger in size.^[76]

These results unleashed a discussion about the influence of company characteristics on long-term performance. Stehle (1999) analyses existing data on the size effect and comes to the conclusion that their results need to be revaluated and adjusted to the appropriate benchmark. Companies with a market capitalisation of under 50 million US$ have a lower average buy-and-hold return than larger companies.^[77] Beyond the sizing issue, Brav/Gompers (1998) argue that the market-to-book values need to be considered for an appropriate benchmark because they realistically reflect the market’s expectation.^[78]

Concluding, if one accepted the view that IPO shares underperformed in the long-run, most explanations on underpricing were hard to substantiate. Mentioned findings suggest that the post-issue performance of IPOs depends both on the benchmark used for market adjustment and the period investigated. If adjusted appropriately, long-term underperformance of IPO shares is not consistently observable.

3 Theories explaining IPO-Underpricing

Finding an appropriate explanation for why companies underprice in IPOs can be done from three perspectives: Firstly, one can assume that the offer price is unintentionally set wrong, which is potentially applicable due to the imprecision of the “Stuttgarter Verfahren” in Germany until the mid 80s, but not in recent years.^[79] Secondly, assuming that investors overestimate the value of the shares implies that shares consistently underperform on the long-run, a supposition disproved previously.^[80] As the only remaining alternative, shares must deliberately be underpriced, a framework developed in depth in the following.

3.1 Institutional Explanations

The following models attributed to the analysis of the underpricing phenomenon are based on “institutional” factors. Firstly, the litigiousness of investors has inspired the legal liability theory. Even though this theory appears to be highly plausible, its applicability must be strongly relativised. Alternatively, underpricing can be legitimated as a form of compensation for analyst coverage. Furthermore, banking relationships are shown to have an impact on the level of initial returns.

3.1.1 Reduction of Legal Liability

The legal liability theory states that underwriters and issuers are exposed to serious risk of litigation if they overestimate or leave out material facts in the IPO prospectus.^[81] In case of a legal defeat, costs to the defendant comprise direct costs such as loss of compensation, legal fees, and omission of management time, but also indirect costs such as reputation impairment.^[82] This especially holds true if class action lawsuits are embedded in the legal system. A recent study exposed the fact that in the period from 1988 to 1995 about 6% of companies floated in the U.S. were sued for violations in the IPOs with claims averaging 13% of IPO proceeds.^[83]

In this surrounding, intentionally underpricing new issue has the effect of acting like insurance against litigation. However, because underpricing reduces the underwriter’s spread and the gross proceeds of the issuer, a trade-off between the potential costs of litigation on the one hand and a reduction of underwriter compensation and lower proceeds on the other hand is necessary. Therefore, investment banks tend to increase their gross spread with higher ex-ante uncertainty about the offer price and it may be argued that the legal liability is a component of this uncertainty.^[84]

Evidence on the legal liability theory is ambiguous. The coherence that more prestigious underwriters with greater due diligence skills have less need for underpricing as a form of protection against law suits can be empirically proven.^[85] However, one can argue that more prestigious underwriters, which are assumed to have greater ability to assess the proper offer price, also have higher financial resources and are therefore more prone to lawsuits.^[86] Ljungqvist (2005) exposes data stating that due to liability reasons, prestigious underwriters are less likely to take risky firms public at all.^[87]

A direct relation from underpricing to the probability of being sued has not yet been empirically demonstrated due to the following simultaneity problem: firms underprice to reduce the probability of lawsuits, but the level of underpricing depends on the probability of being sued. In other words: IPOs likely to be sued later are also underpriced more beforehand.^[88] Neglecting this causal chain, it can be shown that firms which were sued in the past are inclined to under- rather than overstate the value.^[89]

Recapitulating, the legal liability theory provides a motive to intentionally underprice in IPOs but its impact is hard to assess. Furthermore, the legal liability theory has strict liability laws as a prerequisite to the model, which do not hold true in many countries other than the U.S. Nevertheless, the levels of underpricing are on a similar scale.^[90] This indicates that other motives for underpricing have to be persistent.

3.1.2 Analyst Compensation Theory

The role of research analysts has gained a lot of importance in the issuing process of securities in recent years.^[91] Star analysts who are highly ranked in the “Annual Institutional Investor” magazine are found to increase the IPO-market share of the respective investment bank.^[92] Concerning the direct compensation of underwriters in the form of percentage spread, there is surprisingly little variation. Besides the standard 15% greenshoe provision in most IPO contracts, spreads circle around the 7% mark.^[93] Assuming the issuing company sets value on analyst coverage, it has reason to deliberately pay for better coverage accordingly. One academic viewpoint is that underpricing is used to indirectly compensate the investment bank for its services.^[94]

At first sight, underpricing reduces the underwriter’s cash flow because it lowers the spread. Yet, since investment banks are at the same time market makers, they can allocate underpriced IPO shares to preferred clients^[95], thereby creating future investment banking business and receiving trading commissions. This process is also known as spinning. In addition, underpricing has a boosting effect on the post-IPO trading volume, which is served by the market maker, the underwriter.^[96]

If issuers are content with indirectly compensating investment banks by deliberately allowing underpricing, it remains questionable if the loss of proceeds will be compensated by the benefits of better analyst coverage. Also, it is not clear if the underwriter’s benefits will be greater than a raised spread. One can argue that investment banks can generate higher revenues if the form of payment is less transparent, because direct fees are more observed by the issuer.^[97]

Using an extensive set of proxy variables, Cliff/Denis (2004) empirically test this model and find a positive relationship between underpricing and the degree of analyst coverage as well as analyst quality.^[98] They assume that the underpricing is part of the underwriter compensation.^[99] Yet one can critically argue that underpricing also attracts the attention of non-affiliated analysts because they then give out buy-recommendations, facilitating their commission business, and this external effect is not taken into consideration by the authors.

Concluding, the analyst compensation theory leads to the proposition that increased IPO analyst coverage positively relates to the level of underpricing.

[...]

---

^[1] Cp. Ljungqvist (2005), p. 2.

^[2] Cp. Ritter (1984), p. 223.

^[3] Cp. Loughran/Ritter (2003), p. 413.

^[4] Cp. Loughran/Ritter (2003), p. 415.

^[5] Cp. Ljungqvist/Wilhelm (2003), p. 723.

^[6] Cp. Achleitner (2000), p. 241.

^[7] Cp. Achleitner (2000), p. 241.

^[8] Cp. Ritter/Welch (2002), p.1798.

^[9] Cp. Ritter (2003), p. 427.

^[10] Cp. Achleitner (2000), p. 242.

^[11] A more detailed explanation of the offering process will be given in the following chapter.

^[12] Cp. Achleitner (2000), p. 250.

^[13] Cp. Achleitner (2000), p. 266.

^[14] Cp. Aggraval/Prabhalla (2002), p. 1423.

^[15] Cp. Aggraval/Prabhalla (2002), p. 1421.

^[16] Cp. Achleitner (2000), p. 249.

^[17] Cp. Ljungqvist (2005), p. 3.

^[18] Cp. Ross/Westerfield/Jaffe (2002), p. 540.

^[19] Cp. Ljungqvist (2005), p. 4.

^[20] Cp. Sherman (2004), p. 2.

^[21] Cp. Ljungqvist (2005), p. 4.

^[22] Cp. Ross/Westerfield/Jaffe (2002), p. 544.

^[23] Ross/Westerfield/Jaffe (2002), p. 544.

^[24] Cp. Ross/Westerfield/Jaffe (2002), p. 545.

^[25] Cp. Ross/Westerfield/Jaffe (2002), p. 545.

^[26] Cp. Garfinkel/Malkiel/Bontas (2002), p. 51.

^[27] Cp. Garfinkel/Malkiel/Bontas (2002), p. 53.

^[28] Cp. Field/Sheehan (2002), p. 1.

^[29] Cp. Field/Sheehan (2002), p. 2.

^[30] Cp. Field/Sheehan (2002), p. 1.

^[31] Cp. Leutz (2003), p. 453.

^[32] Cp. Leutz (2003), p. 455.

^[33] Cp. Leutz (2003), p. 448.

^[34] Cp. Leutz (2003), p. 449.

^[35] Cp. Ritter (2003), p. 428.

^[36] Cp. Stehle/Ehrhardt (1999), p. 1396.

^[37] Cp. Ritter (2003), p. 426.

^[38] Cp. Stehle/Ehrhardt (1999), p. 1396.

^[39] Cp. Ritter (2003), p. 427.

^[40] Before 2002, the quiet period ended 25 days after going public.

^[41] Cp. Ritter (2003), p. 427.

^[42] Cp. Ritter (2003), p. 428.

^[43] Also known as directors and officers insurance.

^[44] Cp. Ritter (2003), p. 430.

^[45] Cp. Loughran/Ritter (2003), p. 2.

^[46] Cp. Lowry/Murphy (2006), p. 4.

^[47] A small number of researchers note that significant underpricing can also be observed in the course of the first week of trading, therefore including this time frame into their studies, so for example Schrand/Verrecchia (2005), p. 6.

^[48] Cp. Loughran/Ritter (2002), p. 417.

^[49] Cp. Loughran/Ritter (2002), p. 414.

^[50] IPO activity refers to the number of IPOs performed in a specific time frame, thus the frequency of IPOs.

^[51] Cp. Ritter (1984), p. 215.

^[52] Cp. Ibbotson/Jaffe (1975), p. 235.

^[53] Cp. Ritter (1984), p. 223; Ljungqvist (2005), p. 4.

^[54] Cp. Ritter (1984), pp. 215-218.

^[55] Cp. Ritter (2002), p. 1797.

^[56] Deutsche Börse Group, Primärmarktstatistik.

^[57] For an analysis of the IPO activity, company characteristics and underpricing levels during that time frame, see chapter 4.

^[58] Cp. Stehle/Ehrhardt (1999), p. 1397.

^[59] Cp. Stehle/Ehrhardt (1999), p. 1400.

^[60] Cp. Stehle/Ehrhardt (1999), p. 1397.

^[61] Cp. Franzke (2003), p. 12.

^[62] Franzke (2003) calculated the money left on the table by multiplying the total issuing volume with the initial return.

^[63] Loughran/Ritter (2002), p. 419.

^[64] Cp. Drobetz/Kammermann/Wälchli (2005), p. 255.

^[65] Cp. Stehle/Ehrhardt (1999), p. 1409.

^[66] Cp. Drobetz/Kammermann/Wälchli (2005), p. 253.

^[67] Cp. Stehle/Ehrhardt (1999), p. 1401.

^[68] Cp. Ritter/Welch (2002), p. 1797.

^[69] Cp. Stehle/Ehrhardt (1999), p. 1402.

^[70] The firm chosen had the closest market capitalization and the closest market-to-book ratio.

^[71] Cp. Ritter/Welch (2002), p. 1797.

^[72] Cp. Ritter/Welch (2002), p. 1798.

^[73] Cp. Sapusek (2000), p. 377.

^[74] Cp. Sapusek (2000), p. 375.

^[75] Cp. Stehle/Ehrhardt (1999), p. 1414.

^[76] Cp. Stehle/Ehrhardt (1999), p. 1415.

^[77] Cp. Ritter (2006), p. 19.

^[78] Cp. Brav/Gompers (1998), p. 13.

^[79] Cp. Stehle/Ehrhardt (1999), pp. 1400-1402.

^[80] Cp. Stehle/Ehrhardt (1999), p. 1400.

^[81] Cp. Ljungqvist (2005), p. 42.

^[82] Cp. Ljungqvist (2005), p. 41.

^[83] Cp. Lowry/Murphy (2006), p.41.

^[84] Cp. Chen/Mohan (2002), pp. 5-7.

^[85] Cp. Ljungqvist (2005), p. 42.

^[86] Cp. Ljungqvist (2005), p. 3.

^[87] Cp. Ritter (2002), p. 429.

^[88] Cp. Ljungqvist (2005), p. 43.

^[89] Cp. Lowry/Murphy (2006), p.3.

^[90] Cp. Ljungqvist (2005), p. 40.

^[91] Cp. Cliff/Denis (2004), p. 2871.

^[92] Cp. Cliff/Denis (2004), pp. 2870-2874

^[93] Cp. Cliff/Denis (2004), p. 2871.

^[94] Cp. Ritter/Welch (2002), p. 1810.

^[95] So-called friends and family-accounts.

^[96] Cp. Cliff/Denis (2004), p. 2872.

^[97] Cp. Loughran/Ritter (2002), p. 417.

^[98] Cp. Cliff/Denis (2004), pp. 2871-2874.

^[99] Cp. Cliff/Denis (2004), p. 2872.