Long-run IPO stock performance and the European financial crisis

(1)

BACHELOR THESIS

UNIVERSITY OF AMSTERDAM

COLLEGE OF ECONOMICS AND BUSINESS BSc Economics & Business

Bachelor Specialization Economics and Finance

LONG-RUN IPO STOCK PERFORMANCE AND THE

EUROPEAN FINANCIAL CRISIS

Author: F. Gökalp

Student Number: 10814701 Thesis Supervisor: dr. M.I. Dröes Date of Publication: August 15th, 2017

(2)

Abstract

This study explores the effect of the recent financial crisis on the long-run performance of Initial Public Offerings. The thesis conducts an analysis of 806 IPOs within the European Union over the period 2003-2013. No evidence has been found of a difference between the long run performance of IPOs issued during the financial crisis and IPOs issued outside the financial crisis period. By measuring the three-year buy-and-hold abnormal returns, while benchmarking the IPO performance to the local stock performances, and using an equally-weighted scheme, the results show a positive non-significant effect of the crisis on

performance in the long run. However, the raw results show a 11.69% difference between the average abnormal returns of the crisis period and the non-crisis period. The fact that the buy-and-hold abnormal return approach and the equally weighted schemes have shortcomings, as well as the fact that there are alternative ways of benchmarking, deserve further attention.

JEL Classification: G14, G15

Statement of originality

This document is written by student F. Gökalp, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

1. Introduction

The first sale of a private company’s shares to the general public is referred to as an Initial Public Offering (IPO). An IPO gives a firm the opportunity to finance its investments and operations at a lower cost and its shareholders the opportunity to diversify their investments. The way IPOs perform is relevant for a couple of reasons. For the issuer, going public is an important event in the life cycle of the company. On the other hand, from the perspective of the entrepreneur an IPO is the reward of years of building the firm. Moreover, for current venture capitalists and private shareholders an IPO is a chance to make a benefit on investment and leave the firm. The process of going public and the post-IPO performance became of particular interest. IPOs have therefore developed to one of the most popular matters among research in the field of corporate finance.

In the past there has already been done lots of research on Initial Public Offerings. The first anomaly that has been proven multiple times is the underpricing of IPOs, which leads to large returns on the first trading day (Ibbotson’s, 1975). Secondly, it has been documented that IPOs are overpriced in the long run (Loughran and Ritter, 1995). Lastly, it has been documented that there is an existence of hot and cold issue markets. When initial returns rise above the IPO price and an increasing amount of new stocks issues take place, hot issue markets exist (Helwege and Liang, 2004).

This thesis will seek to contribute to existent literature by investigating whether a fourth anomaly, regarding the performance of IPOs during crisis years, exists. Therefore, this thesis’ main research question will be:

Is the long run performance of IPOs issued during periods of financial crises different from IPOs issued outside periods of financial crises?

To narrow the research, the stock performance of 806 IPOs that took place during the time frame of 2003 until 2013 within the European Union will be analyzed. The crisis that will be used for this research is the financial crisis that started in 2008. By looking at the number of IPOs, it is possible to see that the crisis period was particularly prevalent during 2008 and 2009. After defining the period of crisis, data of IPOs has been used to determine whether the performance of going public during these years was different from non-crisis years. This analysis has been done by making use of the three-year buy-and-hold abnormal

(5)

return and controlling for company characteristics, which, in previous literature, are described as prognosticators of IPOs in the long run.

Our raw results show a difference in average return of crisis and non-crisis years. The average return of the crisis years is 7.92%, while the average of the non-crisis years is -3.77%. However, when testing this statistically, this paper does not find significant difference between the two groups. This implies that we cannot conclude that the long-run stock performance of IPOs within the European Union was different during the crisis of 2008-2009.

This papers aims to contribute to the existing body of literature by researching whether there is a difference in long run performance between firms that went public during and outside periods of crisis. Additionally, this paper aims to contribute to the literature by making use of more recent data which is useful for further research in the field of IPO performance.

In order to answer the main question, the remainder of this thesis is structured as follows. A review of the existing literature will be given regarding views on IPOs, motives to go public and IPO performance. After that, the data selection and analysis will be provided. Next, the methodology and results will be discussed. Finally, in the conclusion, there will be given an answer to the main question.

(6)

2. Theoretical Framework

This chapter will discuss the existing relevant research done on Initial Public Offerings. Besides explaining the background of IPOs and the financial crisis, this chapter will examine the previous literature about the three main anomalies of IPO’s, namely underpricing, the long-term underperformance and the hot issue market phenomenon.

2.1 Initial Public Offering

There are two main company categories: private companies and public companies. A private company’s first time of selling stock to the general public is referred to as an Initial Public Offering (IPO). Firms can sell existing shares or issue new shares. The benefits earned by selling new shares make it easier for companies to attract capital, while the sale of existing stock bring revenue to existing shareholders (Jenkinson and Ljungqvist, 2001).

2.1.1 Companies’ motives to go public

There are a couple of reasons for a firm to have an IPO. Bancel and Mittoo (2009) stated that the most important reason for a firm to issue stock is because it provides funds for business growth. At some point in the organizational life cycle of a company it will naturally consider an IPO in order to grow and develop by attracting additional equity capital. Furthermore, Zingales (1995) found that turning to the stock market gives the firm’s founders and its other shareholders the opportunity to convert their investment into cash and diversify their investments.

Another explanation is documented by Pagano et al. (1998). In the sample of 66 Italian IPO firms they came to the conclusion that future investment needs were not the main reason for a company to go public. According to their findings the main factor regarding the decision of having an IPO is the industry market to book ratio. Their result implies that companies operating in industries with more investment opportunities and needs, are the ones that go public. Another possible implication of their results is that companies operating in over-valued industries are more likely to go public.

Moreover, going public brings a firm legitimacy. Other companies get to know the firm that goes public and therefore the chances of potential mergers and acquisitions increase.

(7)

Brau and Fawcett (2006) state that the most important reason for a company to go public is that an IPO creates public stock for use in future acquisitions.

Furthermore, a firm’s choice to go public brings along that it has to follow strict disclosure and regulatory requirements of stock exchanges (Bancel and Mittoo, 2009). The writers examine the benefits and costs of an IPO by surveying chief financial officers. Their conclusion is that the decision to go public is complex, and therefore it is not possible to be explained by one single theory and that the benefits of going public outweigh the costs, which explains why many private firms decide to go public.

2.1.2 The IPO Process

The common first step that a firm has to take when it wishes to go public is to select an investment bank. The investment bank performs as an underwriter and gives the company advice regarding the determination of the best offer price. This operation of selecting is a two-way affair; investment banks investigate their clients at least as thoroughly as firms that select investment banks.

The process of going public can be managed by either one underwriter (sole managed) or multiple underwriters. According to Corwin and Schultz (2005) the tendency is to have a greater number of underwriters that form a syndicate. One bank takes the role of lead manager, and is the most important player during the transaction. The main responsibilities of this lead manager are making plans with the issuing firm, creating an issue schedule, managing the process of the due diligence and pricing of the firm stock. The underwriters are compensated a part of the gross spread in exchange for their support.

Firms that wish to go public have to come up with a Registration Statement, which subsists of two components: the prospectus and information regarding issuance and distribution. The first component has to be shared with every buyer of the shares, while the second component is not meant to be shared with the general public. The goal of the Registration Statement is to provide the public with satisfactory and dependable information.

After processing the issues stated by the inspector regarding the Registration Statement, the offerings marketing starts. The firm and the underwriters pitch the IPO to institutional investors as well as sales people. In the end, the underwriters make expectations of investor interest. The initial price bounds are often comprehended in one of the prospectuses, which is filed after most of the comments of the inspector are processed.

(8)

One day before the offer date the firm and underwriter determine the offer price and the amount of shares to be offered. Ritter (1991) concluded that IPOs are underpriced on the offer day. For the firm, this implies that it does not get full value for its offer. However, this can still be beneficial for the firm when this underpricing assures that the IPO succeeds. After determining the offer price and number of shares, the firm and underwriter sign the Underwriting Agreement and complete the final prospectus. The morning of the offering, the underwriter sends the price amendment. This day, the trade of firm stocks starts. Often, two or three days after the effective date, the firm hands over its stock and the underwriter makes the deposit of the IPO net proceeds to the company. The final phase of the offering starts after the quiet period, which is in general a month after the effective date.

2.1.3 Financial crisis in Europe

In August 2007, the crisis first came to light. As a result of the sudden sale of assets at fire-sale prices, the European Central Bank’s first intervention in the interbank market took place (Block, 2010). Next, in September 2007, the Bank of England made public that it was going to financially back up Northern Rock, a British investment bank. The aim of the central banks was to cool down the markets. Nonetheless, the opposite happened; the public interpreted these interventions as an indication of solvability problems (Acharya et al., 2010).

During the summer of 2008, more problems with investment banks and insurance companies surfaced. One of these banks was Bear Stearns, whose failure made the so-called counterpart risk clear. Counterpart risk presumes that individual companies are interconnected with the rest of the financial system, and losses of one firms cause losses of more firms and therefore leads to a ripple effect around the financial system (Acharya et al., 2010). The first to fail, due to counterpart risk, were the United States’ largest mortgage lenders Freddie Mac and Fannie Mae. As a result, the United States government decided to place both under government conservatorship. According to Acharya et al. (2010) this government action was unavoidable because more than half of the lenders’ debt was owned by financial institutions who were dealing with liquidity problems.

On the 14th of September, the climax of the crisis took place with the failure of investment bank Lehman Brothers due to liquidity problems. In less than two months five US banks declared bankruptcy or merged with other banks. Moreover, AIG, world’s largest insurance firm was nationalized. A ripple effect appeared, causing more financial institutions to lose value or go bankrupt. Because of the inability of these institutions to repay debts, markets all

(9)

over the world collapsed, credit markets froze and the largest crisis since the great depression occurred (Acharya et al., 2010).

2.2 IPO Performance

2.2.1 Short-run performance

There has been done extensive research on the phenomenon called IPO underpricing. According to this notion, shares of firms that go public show substantially positive returns on the first trading day.

The first author that documented this anomaly is Ibbotson (1975). In his dataset of 2650 American IPOs in the 1960s, he found an average initial return of 11.4%. Jenkinson and Ljungqvist (2001) studied 38 IPOs in 35 countries and conclude that the level of underpricing differs across markets. According to their results industrialized countries have an underpricing level of 15%, while the phenomenon performs stronger in emerging markets with a level of 60%. Loughran and Ritter (2004) note that the level of underpricing differs over time. During the 1980s they find relatively low first day returns of 7%, whereas between 1991-1998 this level doubled before reaching a level of 65% during the bubble-period of 1999-2000. During the three years following the bubble period the average underpricing level dropped to a level of 12%. There are several theories explaining underpricing.

The first theory is the adverse selection theory. Rock (1986) investigated a market with two groups of investors; one of the groups had superior information of the issuing firm, while the other group was uninformed about the issuing firm and invested randomly. Subsequently, the informed ones crowd out the uninformed ones when issue prices are below or at true value. Correspondingly, informed investors pull back when they know that issue prices are higher than true value. This results in uninformed investors ending up with shares not wanted by informed investors, which is also known as ‘the winner’s curse’ (Ibbotson et al., 1994). As a result, uninformed investors do not engage anymore, implying a drop in IPO demand, which leads to undersubscribed issuing firms. To compensate less demand, and in order for uninformed investors to at worst break even, IPO companies provide a discount, also known as the underpricing of stock.

The second theory is regarding the relationship between signaling and IPO price. The assumption here is that firms that go public have more information than investors when it

(10)

comes to the true value of the shares issued. This leads to rational investors being afraid of the ‘lemons problem’, implying that low-quality IPO-firms will be the only ones to sell their stock at the average price. On the other hand, high-quality issuers will try to differentiate themselves by underpricing their stock to make a benefit. Investors see through that only the best firms are able to give a signal by lowering their prices now and compensating this in the future. Another reason for firms to intentionally underprice themselves is to receive large returns on the first day and therefore gain positive publicity and media exposure. The theory’s conclusion is that better firms underprice their issue price for the sake of getting higher prices during a future seasoned offering (Chemmanur, 1993).

The third theory explaining underpricing is based on allocations of IPOs. Benveniste and Spindt (1989) note that, by underpricing, firms insure themselves against undersubscription and generate excess demand in order to have discretion when allocating shares to investors. This is valuable since some investors are more desirable than other investors. Moreover, it could be favorable for IPO firms to subdivide allocation among a larger amount of investors to lower new shareholders block size (Brennan and Franks, 1997).

2.2.2 Long-run performance

Even though issued shares on average show high returns on their first day of trading, they commonly show an underperformance in the long run. The first author that reported this phenomenon was Ibbotson (1975). In his sample of US IPOs he found a below average performance in the four years after issuing. However, his results were not statistically significant because of a small sample size and high standard deviation. Ritter (1991) also reported an underperformance. In his sample of 1526 IPOs between 1975 and 1984, he found an average three-year return of 34%, while his benchmark had a value of nearly 62%. Comparable results have been found for other markets. In the United Kingdom, Levis (1993) documented an average IPO underperformance of -23% three years after going public. Furthermore, Stehle et al. (2000) found a three-year abnormal performance of -6% in Germany, while Chahine (2004) reported a two-year underperformance in France of 9.94%.

The long-run underperformance phenomenon is contradictory to the Efficient Market Hypothesis, which would claim that stocks always perform according to their real intrinsic value. Several theories explain the long-run underperformance of issuing stocks.

(11)

The first theory explains excessive optimism of investors. Miller (1977) as well as Ibbotson et al. (1994) state that the most optimistic investors will be the ones who buy the shares. When the value of an issuing firm is unclear, the difference between the valuation of optimistic and pessimistic investors will be large. Nonetheless, this difference will small down when more information becomes clear. In the end, the marginal valuation will converge to the mean; implying share prices to drop and a long-run IPO underperformance. Daniel, Cornelli et al. (2006) also conclude that investors are over-optimistic and exuberant, and therefore cause underperformance.

The second theory is explained by Ritter (1991) and Loughran and Ritter (1995). They note that the windows of opportunity theory is the reason of IPO long-run underperformance. The authors state that overvaluation and exorbitant investor optimism occur more in periods of high IPO volumes. The theory implicates that periods of high IPO volume show the lowest returns and aftermarket performance.

On the other hand Aggarwal and Rivoli (1990) came up with a contradicting view. They stated that the consistent underpricing of IPOs by underwriters, results in the fact that early investors make positive abnormal returns not only in the short run, but also in the long run. However, this statement is not supported by their data.

Furthermore, the Fads Hypothesis, also known as “The Impresario”, gives another explanation to the long-run underperformance of IPOs. It states that investment bankers are impresarios, who underprice the issuing stocks to create excess demand (Shiller, 1988). The forecast is that greater initial first day returns lead to greater correction of overpricing, which results in lower returns.

2.2.3 Hot and cold markets

Another anomaly of IPOs is the occurrence of hot and cold issue markets. This anomaly implies that the volume of IPOs as well as the extent of the first day return are cyclical and exposed to variation. The hot market and underpricing anomaly are related to each other. The first to state the great issuance variation were Ibbotson and Jae (1975). The authors defined their findings as ‘periods with abnormally high aftermarket performance of IPOs’. Their conclusion was that the offer prices of IPOs that took place in hot issue market were further above their efficient prices than offer prices of cold issue IPOs. Moreover, Helwege and Liang (2004) state that hot issue markets contain high IPO volumes, greater underpricing and

(12)

offerings oversubscription. While on the other hand, cold markets are characterized by low offerings volume, less underpricing and little offerings oversubscription. Companies try to go public during times of stock market peaks in order to benefit from investor optimism and high valuations (Baker and Wurgler, 2000).

Throughout the literature, several models and theories are formed to find out what the differences are between hot period issuing firms and cold period issuing firms. This difference is of importance for this paper because crisis periods can be characterized as cold markets in terms of low IPO volumes. Even though Helwege and Liang (2004) did not find characteristic differences between firms going public in the different markets, I present two theories below that are relevant for my thesis.

The first theory is regarding the differences in industry. The observation is that in hot periods of IPO activity, IPOs are more concentrated in particular industries. Several theories say that the process of an IPO is informative regarding the company’s industry, implying that companies from the same industry are encouraged to offer in the same period. Firms learn about their own valuation, when they see other firms from the same industry go public (Benveniste et al. (2002). When companies find out that the prospects of their industry are high, it becomes more costly to stay private, leading to more companies within the same industry to go public. The second theory is about the size of firms. Jaskiewicz et al. (2005) and Brav and Gompers (1997) state that size of a firm has a positive effect on the long-term abnormal returns of IPOs. Choe et al. (1993) identify hot market firms with higher quality and size.

2.3 Hypothesis

With the literature review taken into account, this part formulates the hypothesis in order to achieve this thesis’ research aims and analyze the relation between the crisis and European IPO market.

Two contradicting arguments will be set up to explain the relationship between the crisis and the long-run IPO performance. The first theory is in favor of a higher long-term performance during crisis years, while the second provides an explanation for lower long-term performance during crisis years.

As mentioned in Section 2.2.2, Ibbotson et al. (1994) state that the underperformance of IPOs in the long-run is caused by extremely optimistic investors. These investors are the ones that buy the shares, while their valuation differs a lot from more pessimistic investors.

(13)

According to the theory this difference in valuation between investors narrows when more information becomes available. This leads to falling stock prices because the marginal valuation converges to the mean. However, during crisis years the optimism of investors is in general expected to be lower. This results in a smaller valuation gap between more optimistic and more pessimistic investors during crisis years, implying less conversion of the marginal valuation to the main. Consequently, the long-term IPO performance of firms issued during crisis years will be better than firms issued during non-crisis years due to lower initial prices.

On the other hand, there are theories that suggest the opposite. As stated in section 2.2.3, Helwege and Liang (2004) mention that risk averse investors will trade less frequently during cold periods, implying that this will occur during crisis. Consequently, the performance of firms that issue during crisis years will be worse than firms that go public in non-crisis years.

Hypothesis:

Hypothesis: The three-year Buy-and-Hold Abnormal Returns for firms that went public during the crisis years 2008 and 2009 will be different from firms that went public outside crisis years.

If the hypothesis holds, it gives an answer to the question whether the long run performance of IPOs issued during periods of crises is different than the performance of IPOs issued outside periods of financial crises.

(14)

3. Data

This chapter presents the data that is used in this research. Specifically, the data selection procedure, data sources and the motivation behind the choice of data will be described. Moreover, the descriptive statistics will be presented.

3.1 Data selection

The thesis’ sample will contain all IPOs completed between 2003 and 2013 within the European Union (EU)1. The sample will include IPOs on both the main and secondary markets. By choosing the time frame 2003-2013, the data before and after the crisis of 2008-2009 can be captured. To be able to measure the three-year post IPO performance of data collected until the end of 2016, the time frame is set until 2013.

Furthermore, the sample is refined to all IPOs with a known offer price that is higher than €1, as proposed by Ritter (1991). Also, in line with Ritter (1991), real estate firms, financial firms, equity, trust or close-end funds and unit offers or spin-offs are excluded from my data sample. Moreover, the selected IPOs had available monthly stock prices for at least 3 years after the IPO.

The data is found through the Zephyr of Bureau van Dijk Database. The database provided company names, countries, ISIN numbers, SIC codes, dates of IPO completions and post-deal company total assets.

After obtaining the IPO information from Zephyr, the three-year post IPO returns could be retrieved through the Datastream database by matching company ISIN numbers. A small amount of IPOs with insufficient data were eliminated from the dataset. The resulting data sample consists of 806 IPOs. A chart overview of the refinement of the sample is given in the Figure 1 below.

Figure 1: Sample refinement

1_{Member states of the European Union are: Austria, Belgium, Bulgaria, Croatia, Czech Republic, Denmark, Estonia, Finland, France,}

Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Maltha, Netherlands, Poland, Portugal, Romania, Slovenia, Slovak Republic, Spain, Sweden and the United Kingdom

Initial Public Offerings within the EU Time set:

2003-2013 Deal offer price > €1

Non-Financial sectors Datastream restriction The research sample

(15)

3.2 Data Analysis

The first variable I analyze is the three-year buy-and-hold abnormal return. This is derived from the Datastream database by finding the monthly stock and local index prices of the IPOs. Figure 2 provides the three-year average IPO buy-and-hold abnormal return for each year. Both of the crisis years 2008 and 2009 gave positive abnormal returns, namely 13% and 2.84%. Economically, this means that firms that went public during crisis years, on average outperformed the market. The highest abnormal return of the time frame was recorded in 2008. However, 2009 is ranked fifth in terms of highest abnormal return. The average return of the crisis years is 7.92%, while the average of the non-crisis years is -3.77%. Which means that my raw results are in line with the hypothesis stated in section 2.

Six out of nine non-crisis years displayed negative abnormal returns. However, the non-crisis years 2003 and 2010 are ranked second and third in the highest abnormal return ranking with abnormal returns of respectively 7.34% and 4.75%. The average three-year abnormal return over the whole period is -1.65%, which means that the sample on average underperformed against the market. This is in line with Ritter (1991) and Loughran and Ritter (1995).

Figure 2: Long-term performance of IPOs per year

Secondly, the variable of interest of this thesis, crisis, is explained. As mentioned before, the time set is divided into crisis years and non-crisis years, with 2008 and 2009 being crisis years. The sample consists of 751 IPOs that took place in non-crisis years and 55 IPOs that took place in crisis years. Figure 3 gives an overview of the IPO volumes per year. As can be seen in the figure, the years after the dot-com bubble can be defined as years of low volumes. These numbers increased to peak volumes right before the crisis; with 191 firms going public in 2006 and 164 in 2007. During the first year of the crisis, this number of 164 collapsed to 34

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 BHAR 7.34% -5.87% 4.22% -5.91% -3.72% 13% 2.84% 4.75% -16.48 -6.22% -12.06 7.34% -5.87% 4.22% -5.91% -3.72% 13% 2.84% 4.75% -16.48% -6.22% -12.06% -20.00% -15.00% -10.00% -5.00% 0.00% 5.00% 10.00% 15.00% 3-Ye ar B HA R

(16)

and in the second year of the crisis it lowered even more to 21. After the crisis, volumes did not raise above 60 firms going public per year.

Figure 3: Number of IPOs per year

Thirdly, the control variable assets is being discussed. The Zephyr database provided me the total asset value of IPO firms in euros right after going public. The average amount of assets of non-crisis is 1.775.802, while the firms that went public during the crisis years had an average amount of assets of 1.549.583. This implies that companies that went public outside the crisis period had a larger size in terms of assets.

Figure 4: Average number of assets per year

Fourthly, another control variable, industry will be discussed. The firm’s industries are derived from the Standard Industrial Classification (SIC) codes that were provided by Zephyr. Table 1 provides an overview of the IPOs per sector. It can be seen that IPOs tend to be more present in some particular industries. The dominant industries for IPOs within the European

23 64 104 191 164 34 21 60 54 34 57 0 50 100 150 200 250 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 N umb er o f I PO s Year 357,472 452,947 1,936,790 2,849,004 241,536 278,810 3,607,024 732,486 6,647,594 684,189 1,250,502 0 5,000,000 10,000,000 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Assets

(17)

Union are Services (33.8%) and Manufacturing (31.9%). Together, these two industries make up for nearly 66% of my sample.

Table 1: IPOs distribution per sector

Industry / year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Total

Agriculture, Forestry and Fishing 0 0 0 0.5 2.4 0 4.8 1.7 3.7 0 0 1.1

Mining 17.4 4.7 7.7 5.2 3.0 17.6 0 6.7 9.9 2.9 0 5.7

Construction 0 0 2.9 4.7 3.0 0 0 0 0 2.9 1.8 2.4

Manufacturing 34.8 31.3 30.8 33.5 34.1 26.5 9.5 21.7 35.2 35.3 38.6 31.9

Transportation, Communication, Electric, Gas & Sanitary 21.7 20.3 7.7 8.4 11.6 14.7 19.0 16.7 0 11.8 7.0 10.9

Wholesale Trade 8.7 4.7 0 2.6 2.4 5.9 9.5 5.0 3.7 5.9 0 3.1

Retail Trade 4.3 1.6 3.8 2.6 1.8 2.9 4.8 8.3 1.9 2.9 8.8 3.5

Finance, Insurance and Real Estate 0 4.7 9.6 7.9 8.5 0 19.0 1.7 7.4 2.9 14.0 7.4

Services 13 32.8 36.5 34.6 32.3 32.4 33.3 38.3 38.9 35.3 29.8 33.8

Public Administration 0 0 1.0 0 0.6 0 0 0 0 0 0 0.3

Grand Total 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

Fifthly and lastly, the control variables Hot periods and Cold periods are explained. Both binary dummy variables are determined by using information regarding IPO volumes per month provided by Zephyr. To determine whether a month was a hot or a cold period month, the IPO volume per month is been measured over the sample period. After counting the number of IPOs per month, each month has been ranked accordingly. The top quartile of this ranking are labeled as hot months, while the lowest third has been ranked as cold months. Moreover, a hot period is described as a period with at least three hot months in a row. For cold periods the same goes up. Table 2 gives an overview of hot and cold periods over the sample period. The sample period consists of 132 months, 22 months were part of hot periods while 33 months were part of cold periods. In the first year of the time set, one of the years after the dot-com bubble a cold period of several months is being measured. The two years before the crisis can be described as years with long hot periods. Furthermore, more than half of the months within the crisis years are labeled as part of cold periods.

(18)

Table 2: Overview of hot and cold periods

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan

Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb Feb

Mar Mar Mar Mar Mar Mar Mar Mar Mar Mar Mar

Apr Apr Apr Apr Apr Apr Apr Apr Apr Apr Apr

May May May May May May May May May May May

Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun

Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul

Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug

Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep

Oct Oct Oct Oct Oct Oct Oct Oct Oct Oct Oct

Nov Nov Nov Nov Nov Nov Nov Nov Nov Nov Nov

Dec Dec Dec Dec Dec Dec Dec Dec Dec Dec Dec

In addition to the data provided above, Table 3 shows the number of observations, mean, standard deviation, minimum and maximum values of the variables. The mean of variable Crisis is given in column 1, implying that 6.8% of the firms in our sample went public during crisis years. Furthermore, column 4 shows that in our sample the firm with the most assets has 340 million assets. Moreover, this variable has a high standard deviation indicating that there is a high variation across the sample when it comes to number of assets. Furthermore, it can be seen that over 25% of our sample firms went public after the crisis, 45.8% went public during hot periods and 4.5% went public during cold periods.

Table 3: Descriptive statistics

(1) (2) (3) (4) (5)

VARIABLE Mean Std. Dev. Min Max N

Crisis .0682382 .2523108 0 1 806

Assets 1743593 1.69e+07 21 3.43e+08 806

Industry 6.104218 2.489043 1 10 806 PostCrisis .2543424 .4357613 0 1 806 HotPeriod .4578164 .4985267 0 1 806 ColdPeriod .044665 .2066956 0 1 806 Hot period Cold period No hot or cold period

(19)

4. Methodology

This chapter starts with providing an overview of the literature on methodology. Subsequently, the applied methodology is being explained and motivated.

4.1 Literature on methodology

Throughout the literature two time regimes are used when looking at the long-term IPO performance. Writers either use the Event Time or the Calendar Time approach.

Schöber (2008) states that most authors use the Event Time approach. In this approach, researchers start with calculating the excess stock returns of issuing firms for a chosen time from the first day of the IPO on. Next, the average of the excess stock return is calculated and compared to a benchmark. The two most used types of the Event Time approach are Buy-and-Hold Abnormal Returns (BHAR) and Cumulative Abnormal Returns (CAR).

The BHAR method calculates compounded returns over a chosen time period for the company as well as the benchmark. Subsequently, these two are being subtracted from each other. Fama (1998) mentions that compounding is the vulnerable aspect of BHAR because one single year with extreme results can disproportionally affect the results. Schöber (2008) also notes that considerable researchers thought that BHAR has a probability of carrying out unreliable statistical testing. Therefore, several researchers decided to make use of the CAR method instead. This method calculates the excess returns of IPOs on a daily basis and subsequently sums up these returns for the chosen time period. However, the CAR method has shortcomings as well. Schöber (2008) mentions that CARs tend to show upward bias. Furthermore, he states that CARs, unlike the BHARs, do not represent the investor’s point of view when it comes to representing returns.

The Calendar Time approach forms a portfolio of companies who recently went public for each month. Next, the method creates a regression by using these monthly portfolio returns in an asset pricing model. Examples of asset pricing models are the Fama-French three-factor model and the Carhart four-factor model. Subsequently, the approach tests whether the regression’s intercept, also known as Jensen’s Alpha, is different from zero.

Since, both of the approaches have their weak spots, there is discussion throughout the literature regarding the choice of approach. The main criticism expressed about the Event Time approach is that it has the possibility of being biased by cross-sectional dependence,

(20)

on the lack of sufficient criteria for the matching benchmark. On the other hand, imperfections of the Calendar Time approach are heteroscedasticity of the portfolio returns and the probability of underestimating periods of clustering. Furthermore, Ang and Zhang (2015) note that the Calendar Time’s asset-pricing models are not capable of fully reproducing the returns and risks of market stocks.

Moreover, when using BHAR, there are two weighting schemes to calculate the average sample abnormal return. The first scheme is value-weighted, while the second scheme is equally weighted. The value-weighted version calculates average abnormal returns by using for example total assets or market capitalization. The equally weighted gives all firms the same value, this results in higher abnormal returns for small firms in comparison to the value weighted method. Fama (1998) states that smaller firms have more anomalies compared to bigger firms. In order to compensate this effect, several researchers use the value weighted method. However, Schöber (2008) mentions that the use of value-weighted method causes bigger companies to bias the results by seriously affecting the sample.

4.2 Applied methodology

In line with Ritter (1991) this paper takes three years following an IPO to measure long-term performance, this allows to look at firms that went public from 2003 till 2013.

This thesis makes use of Buy-and-Hold Abnormal Returns (BHARs) to determine the performance of IPOs because this approach is more often used in literature and, compared to CAR, it gives a better view of the perspective of investors.

BHARs are calculated by determining the difference between individual firm returns and a selected benchmark for a particular period of time. The BHAR formula can be written as follows:

(1) 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 = ∏_𝑡𝑡=1𝜏𝜏 [1 + 𝐵𝐵_{𝑖𝑖𝑡𝑡}] − ∏_𝑡𝑡=1𝜏𝜏 [1 + 𝐸𝐸(𝐵𝐵_{𝑏𝑏𝑡𝑡})]

𝐵𝐵𝑖𝑖𝑡𝑡 and 𝐸𝐸(𝐵𝐵𝑖𝑖𝑡𝑡) are respectively the returns on IPO 𝑖𝑖 and benchmark 𝑏𝑏 at period 𝑡𝑡. Furthermore, ∏_𝑡𝑡=1𝜏𝜏 stands for compounded returns: the product of all returns for the taken period.

The three-year buy and hold abnormal returns are computed by the use of Datastream’s monthly stock prices and market indices. In order to account for

(21)

country-specific risk, the local stock market indices are used as benchmark. Besller and Seim (2012) note that this is needed to rule out country-specific return characteristics. Moreover, in the light of accounting for survivor bias, all IPOs are kept within the data sample, disregarding whether they delisted before the end of the three year time set. In line with Levis (2011), Viviani et al. (2008) and Cao and Lerner (2009), among others, this thesis uses a time set of three years to measure the long-run abnormal returns.

Furthermore, this paper makes use of the equally weighted scheme instead of the value-weighted scheme. As mentioned in Section 4.1 this eliminates the possibility of large firms to bias the sample. The formula of the equally weighted BHAR can be written as follows:

(2)

𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝑤𝑤𝑤𝑤𝑖𝑖𝑤𝑤ℎ𝑡𝑡𝑤𝑤𝑡𝑡 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 = 1_{𝑛𝑛 �[𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵}𝑖𝑖] 𝑁𝑁

𝑖𝑖=1

By using this method and running a couple of statistical tests, the hypothesis stated in Section 2 can be tested. Firstly, a t-test can be conducted to test for differences in mean between crisis and non-crisis years. Consequently, control variables can be added and the following regression can be run to test the hypothesis:

(3) 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵_𝑖𝑖 = 𝛼𝛼_𝑖𝑖 + 𝛽𝛽₁𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶_𝑖𝑖 + 𝛽𝛽₂𝐵𝐵𝐶𝐶𝐶𝐶𝑤𝑤𝑡𝑡𝐶𝐶_𝑖𝑖 + 𝛽𝛽₃𝐼𝐼𝑛𝑛𝑡𝑡𝐸𝐸𝐶𝐶𝑡𝑡𝐶𝐶𝐸𝐸_𝑖𝑖 + 𝛽𝛽₄𝑃𝑃𝑃𝑃𝐶𝐶𝑡𝑡𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶_𝑖𝑖 + 𝛽𝛽5𝐵𝐵𝑃𝑃𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖+ 𝛽𝛽6𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 + 𝜀𝜀𝑖𝑖

𝐵𝐵0: 𝛽𝛽1 = 0 𝐵𝐵0: 𝛽𝛽1 ≠ 0

Hypothesis: The three-year Buy-and-Hold Abnormal Returns for firms that went public during the crisis years 2008 and 2009 will be different from firms that went public outside crisis years.

𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝑖𝑖 denotes the three-year buy-and-hold abnormal return for IPO firm i. The variable of interest is 𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶_𝑖𝑖, which is a dummy variable that denotes whether the IPO of the firm took place during crisis years. Moreover, two firm characteristics Assets and Industry are added as control variables. 𝐵𝐵𝐶𝐶𝐶𝐶𝑤𝑤𝑡𝑡𝐶𝐶_𝑖𝑖 is a variable that states for the firm’s post-deal totals assets in

(22)

the firm. As stated in Section 2, these two variables are carried out by existing literature as to have an effect on BHAR. Furthermore, to check whether the crisis had an effect on the years after the crisis, and thus caused post-crisis abnormal returns to behave different from pre-crisis abnormal returns, we look at the dummy variable 𝑃𝑃𝑃𝑃𝐶𝐶𝑡𝑡𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖, which shows whether or not a firm’s IPO was in the years following the crisis. Finally, the dummy variables 𝐵𝐵𝑃𝑃𝑡𝑡𝐻𝐻𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 and 𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 denote whether or not a firm went public in a hot or cold period. 𝛼𝛼_𝑖𝑖 is the constant, while 𝜀𝜀_𝑖𝑖 is the residual. The student t-statistics is being used to either accept or reject the hypothesis.

Since there is just one observation per company, there is no time indicator included in the equation. Also, the standard errors of the regressions are estimated by the use of robust standard errors, to correct for heteroscedasticity and the lack of normality in the sample. The use of robust standard errors is in line with Huber (1967).

Robustness is a key condition for causal interpretations to be valid. In the regression, I already checked for robustness by adding control variables. To further analyze whether the results of this regression also hold when subjecting BHAR to more profound robustness checks, additional regressions with exactly the same variables are being run.

The first additional regression is done without including BHAR outliers. Outliers are points that are distinctly different from other observations. The outliers are determined by calculating 1.5 interquartile ranges below the first quartile and 1.5 quartile above the third quartile of the set of BHARs. The data points below and above these points are being marked as outliers. Secondly the regression has been run with a smaller time frame. Instead of taking time set 2003-2013, the time frame is set as crisis years 2008 and 2009, in addition to 2 pre-crisis and 2 post-pre-crisis years (2006-2011).

(23)

5. Research Results

This section analyzes the results of the regressions which are estimated with models explained in Chapter 4.

The result of the t-test conducted to test for differences in mean between the crisis group and the non-crisis group is provided in Table 4. This t-test tests our hypothesis, which states that there is no difference between BHARs in crisis and non-crisis years. The results imply that the null hypothesis should not be rejected, because the p-value is 0.1075.

Table 4: Equality by mean, two-sample t-test

Group N Mean Std. Error Std. Dev. [95%

conf. interval] 0 - Non-Crisis 751 -4.002 2.135 58.465 -8.190 0.186 1 - Crisis 55 9.080 7.147 53.001 -5.248 23.409 Combined 806 -3.109 2.049 58.171 -7.131 0.913 Difference -13.082 8.118 -29.017 2.852 Diff = Mean(0)– Mean(1) t = -1.612 H0: Diff = 0 df = 804 H1: Diff ≠ 0 p-value = 0.1075

Additionally, to control for other factors carried out by existing literature as to have an effect on BHAR, control variables are added to the model and a multiple regression is being run. The results of the regression are presented in Table 5. The R-squared of the model is 3%, which implies that 3% of the variation of BHAR is explained by the model. Also, according to these results, the coefficient for the crisis dummy variable is 0.74 and is statistically not significantly different from zero.

𝐵𝐵0: 𝛽𝛽1 = 0 𝑐𝑐𝐸𝐸𝑛𝑛𝑛𝑛𝑃𝑃𝑡𝑡 𝑏𝑏𝑤𝑤 𝐶𝐶𝑤𝑤𝑟𝑟𝑤𝑤𝑐𝑐𝑡𝑡𝑤𝑤𝑡𝑡 𝐸𝐸𝑡𝑡 𝐸𝐸𝑛𝑛𝐸𝐸 𝐶𝐶𝑤𝑤𝐸𝐸𝐶𝐶𝑃𝑃𝑛𝑛𝐸𝐸𝑏𝑏𝐸𝐸𝑤𝑤 𝐸𝐸𝑤𝑤𝑙𝑙𝑤𝑤𝐸𝐸, 𝐶𝐶𝑖𝑖𝑛𝑛𝑐𝑐𝑤𝑤 𝑡𝑡ℎ𝑤𝑤 𝐻𝐻 − 𝑙𝑙𝐸𝐸𝐸𝐸𝐸𝐸𝑤𝑤 𝑤𝑤𝑖𝑖𝑙𝑙𝑤𝑤𝐶𝐶 0.928.

(24)

Statistically, this implies that there is no significant difference between the three-year buy-and-hold abnormal returns of firms that went public during crisis years and non-crisis years. In Section 3, the raw results showed us that the average return of the crisis years is 7.92%, while the average of the non-crisis years is -3.77%. However, our regression concludes that this difference is not significant.

Table 5: BHAR Regression

(1) VARIABLES BHAR Crisis 0.738 (0.928) Assets -1.24e-07* (0.051) Industry 0.254 (0.761) PostCrisis -13.73*** (0.009) HotPeriod -21.10*** (0.000) ColdPeriod -10.74 (0.190) Constant 9.142 (0.167) Observations 806 R-squared 0.030

Notes: p-values in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Furthermore, the control variables 𝐵𝐵𝐶𝐶𝐶𝐶𝑤𝑤𝑡𝑡𝐶𝐶_𝑖𝑖, 𝐼𝐼𝑛𝑛𝑡𝑡𝐸𝐸𝐶𝐶𝑡𝑡𝐶𝐶𝐸𝐸_𝑖𝑖, 𝑃𝑃𝑃𝑃𝐶𝐶𝑡𝑡𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶_𝑖𝑖, 𝐵𝐵𝑃𝑃𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡_𝑖𝑖 and 𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 are being analyzed.

By looking at the variables 𝐵𝐵𝐶𝐶𝐶𝐶𝑤𝑤𝑡𝑡𝐶𝐶𝑖𝑖 and 𝐼𝐼𝑛𝑛𝑡𝑡𝐸𝐸𝐶𝐶𝑡𝑡𝐶𝐶𝐸𝐸𝑖𝑖. The regression reports that the coefficient of total company assets is low; for every additional Asset, the BHAR only decreases with 1.24 ∙ 10−7. With a p-value of 0.051 the variable is statistically significantly different from zero at a level of 10%. However, at a stronger significance level of 5% the variable is statistically not different from zero.

The results of the variable 𝐼𝐼𝑛𝑛𝑡𝑡𝐸𝐸𝐶𝐶𝑡𝑡𝐶𝐶𝐸𝐸_𝑖𝑖 show us a positive coefficient of 0.25. Also, we can conclude that there is no significant difference between the buy-and-hold abnormal returns of different industries, since the p-value of this variable is 0.76.

(25)

To see whether the crisis had an effect on the years after the crisis, and thus caused post-crisis abnormal returns to behave different from pre-crisis abnormal returns, we look at the dummy variable 𝑃𝑃𝑃𝑃𝐶𝐶𝑡𝑡𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖. According to the results of the regression, this dummy variable has a coefficient of -13.73 and is significantly different from zero at a strong significance level of 1%. Economically, this implies that companies that had an IPO in the years after the crisis on average underperformed companies that went public before the crisis.

Additionaly, we can look at the variables 𝐵𝐵𝑃𝑃𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 and 𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖 to verify whether going public during a hot or cold period has an effect on the abnormal returns of a company. The results show that the variable 𝐵𝐵𝑃𝑃𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡_𝑖𝑖 has a coefficient of -21.10 and a significant p-value of 0.00, implying that going public during on average leads to lower BHARs in comparison to firms that go public in other periods. In section 3 we analyzed that there were no Hot periods during the crisis years. Finally, we look at the control variable 𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡𝑖𝑖, which is not significant with a p-value of 0.190 and a coefficient of 0.190, implying that there is no significant difference between the BHARs of firms going public between cold periods and other periods.

Since not all of the control variables are significant, there are relatively higher standard errors within the model and therefore we can state that the estimation of the corresponding coefficients is improper.

Table 6 provides us the correlation between variables. The correlation matrix helps us to check for multicollinearity. Keller and Warrack (2003) state that there is a multicollinearity problem within a model when the absolute correlation value between two variables is higher than 0.7 In our correlation matrix, no values above 0.7 are shown. Additionally, in line with the theory of Ibbotson et al. (1994) explained in Section 2, Table 6 shows a positive correlation between BHAR and Crisis. With a value of 0.567, this correlation is moderate. Furthermore, 𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝑖𝑖𝐶𝐶 is negatively correlated to all other variables, except for 𝐶𝐶𝑃𝑃𝐸𝐸𝑡𝑡𝑃𝑃𝑤𝑤𝐶𝐶𝑖𝑖𝑃𝑃𝑡𝑡 , which is also in line with the theory explained in Section 2.

(26)

Table 6: Correlation Matrix

To analyze whether the results of this regression, stating that the long-run IPO performance of firms going public during crises are not significantly different from IPOs issued outside crisis periods, also hold when we subject BHAR to further robustness checks, a second and third regression are run. The results of these regressions can be found in Table 7.

The second regression in the table is run by shortening the time frame to 2006-2011. The R-squared of this new model is 0.021, meaning that 2.1% of the variation of BHAR is explained by the model. Furthermore, our variable of interest has a higher coefficient of 7.68 a lower p-value compared to the first regression. However, this p-value of 0.340 is still not significant, economically implying that there is no significant difference between crisis and not-crisis periods. When looking at the third regression we see that the R-squared is raised to 0.047 by eliminating our outliers. However, our variable of interest is still not significant because it shows a p-value of 0.591. In conclusion, adding regressions to subject BHAR to further robustness checks did not change the main statistical conclusions of the first regression.

BHAR Crisis Assets Industry Post crisis HotPeriod ColdPeriod

BHAR 1.000 Crisis 0.567 1.000 Assets -0.0282 -0.0031 1.000 Industry -0.0027 -0.0034 0.0234 1.000 PostCrisis -0.0425 -0.1581 0.0237 0.0305 1.000 HotPeriod -0.1345 -0.2487 -0.0907 0.0506 -0.3480 1.000 ColdPeriod -0.0117 0.2035 0.1727 0.0103 -0.0392 -0.1987 1.000

(27)

Table 7: BHAR regressions

(1) (2) (3)

(2003-2013) (2006-2011) (2003-2013, outliers eliminated)

VARIABLES BHAR BHAR BHAR

Crisis 0.738 7.684 3.634

(0.928) (0.340) (0.591)

Assets -1.24e-07* -1.79e-07*** -8.79e-08

(0.051) (0.000) (0.166) Industry 0.254 0.815 0.597 (0.761) (0.364) (0.295) PostCrisis -13.73*** -4.079 -9.884** (0.009) (0.580) (0.011) HotPeriod -21.10*** -11.19** -18.18*** (0.000) (0.024) (0.000) ColdPeriod -10.74 -1.134 -5.715 (0.190) (0.908) (0.453) Constant 9.142 -3.048 -3.847 (0.167) (0.631) (0.374) Observations 806 524 763 R-squared 0.030 0.021 0.047

Notes: p-values in parentheses; *** p<0.01, ** p<0.05, * p<0.1

(28)

6. Summary and Conclusion

The aim of this study was to investigate the relationship between the long-run performance of companies that went public during the crisis and companies that went public in non-crisis years within the European Union. This thesis builds on previous literature regarding IPO performance, and contributes by researching whether a new anomaly exists. Moreover, the time period including stock prices until 2016 is very recent and thus contributes to the existing research. The main research question of this paper has been: ‘Is the long run performance of IPOs issued during periods of financial crises different from IPOs issued outside periods of financial crises?’.

According to all our tests the dummy-variable Crisis is statistically not significantly different from zero, implying that within the European Union there is no prove of a difference in three-year BHARs between firms that went public during the crisis period 2008-2009 and firms that went public outside this period. This is not in line with the hypothesis stated in Section 2, which was based on theories of Ibbotson et al. (1994) and Helwege and Liang (2004). Based on the first theory I expected crisis firms to perform better than non-crisis firms, while the second theory expected the opposite. Since our statistical results show neither an over- or underperformance of firms that went public during crisis years a possible explanation can be that these theories balanced each other out. However, findings stating non-significance cannot be concluded as “no effect exists”. Also, my raw results showed a difference between the groups, since the average returns were 11.69% higher in crisis years compared to non-crisis years. This raises new questions new questions that further researchers can explore.

More robust results could be found by using a longer time set: Loughran and Ritter (1995) made use of a period of 20 years in order to let the results be generalizable. This would also be a solution for the industry clustering problem because this thesis’ time horizon of 11 years is limited to solve it. Also, a shortcoming of this paper is the lack of significance of the control variables in the main regression. This caused a relatively higher standard error and inaccurate coefficients. It is recommended to identify new IPO characteristics that are capable to predict the long run performance of IPOs. Furthermore, the used BHAR method has its shortcomings. Since the method makes use of compounded returns, single years might over-influence the results. Also, using a benchmark of matching firms instead of the local market index could be considered in future research. Besides, a shortcoming of the used equally

(29)

for small firms. Future research may consider using the value weighted-BHAR. Additionally, since we state that the optimism of investors influences the long run performance, further research may study which effect the crisis had in other markets to see whether cultural differences affect the IPO performance. Finally, it may be interesting to perform this research again in a few years in order to be able to make use of more post-crisis years.

(30)

References

Acharya, V. V., Cooley, T. F., Richardson, M., & Walter, I. (2010). Regulating Wall Street.

Aggarwal, R., & Rivoli, P. (1990). Fads in the Initial Public Offering Market? Financial

Management, 19(4), 45-57.

Ang, J. S., & Zhang, S. (2014). Evaluating Long-Horizon Event Study Methodology.

Handbook of Financial Econometrics and Statistics, 383-411.

Baker, M., & Wurgler, J. (2000). The Equity Share in New Issues and Aggregate Stock Returns. The Journal of Finance, 55(5), 2219-2257.

Bancel, F., & Mittoo, U. R. (2009). Why Do European Firms Go Public? European Financial

Management, 15(4), 844-884.

Benveniste, L. M., Busaba, W. Y., & Wilhelm, W. J. (2002). Information Externalities and the Role of Underwriters in Primary Equity Markets. Journal of Financial Intermediation,

11(1), 61-86.

Benveniste, L. M., & Spindt, P. A. (1989). How investment bankers determine the offer price and allocation of new issues. Journal of Financial Economics, 24(2), 343-361.

Bessler, W., & Seim, M. (2012). The performance of venture-backed IPOs in Europe. Venture

Capital, 14(4), 215-239.

Block, F. (2010). Crisis and renewal: the outlines of a twenty-first century new deal.

Socio-Economic Review, 9(1), 31-57.

Brau, J. C., & Fawcett, S. E. (2006). Initial Public Offerings: An Analysis of Theory and Practice. The Journal of Finance, 61(1), 399-436.

(31)

Brav, A., & Gompers, P. A. (1997). Myth or Reality? The Long-Run Underperformance of Initial Public Offerings: Evidence from Venture and Nonventure Capital-Backed

Companies. The Journal of Finance, 52(5), 1791-1821.

Brennan, M., & Franks, J. (1997). Underpricing, ownership and control in initial public offerings of equity securities in the UK. Journal of Financial Economics, 45(3), 391-413.

Cao, J., & Lerner, J. (2009). The performance of reverse leveraged buyouts. Journal of

Financial Economics, 91(2), 139-157.

Chahine, S. (2004). Long-run abnormal return after IPOs and optimistic analysts' forecasts.

International Review of Financial Analysis, 13(1), 83-103.

Chemmanur, T. J. (1993). The Pricing of Initial Public Offerings: A Dynamic Model with Information Production. The Journal of Finance, 48(1), 285.

Choe, H., Masulis, R. W., & Nanda, V. (1993). Common stock offerings across the business cycle. Journal of Empirical Finance, 1(1), 3-31.

Cornelli, F., Goldreich, D., & Ljunqvist, A. (2006). Investor Sentiment and Pre-IPO Markets.

The Journal of Finance, 61(3), 1187-1216.

Corwin, S. A., & Schultz, P. (2005). The Role of IPO Underwriting Syndicates: Pricing, Information Production, and Underwriter Competition. The Journal of Finance, 60(1), 443-486.

Fama, E. F. (1997). Market Efficiency, Long-Term Returns, and Behavioral Finance. SSRN

Electronic Journal, 49(3), 283-306.

Helwege, J., & Liang, N. (2004). Initial Public Offerings in Hot and Cold Markets. Journal of

(32)

Huber, P. J. (1967). The behavior of maximum likelihood estimates under nonstandard conditions. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and

Probability. Berkeley, CA: University of California Press, vol. 1, 221–233.

Ibbotson, R. G. (1975). Price performance of common stock new issues. Journal of Financial

Economics, 2(3), 235-272.

Ibbotson, R. G., & Jaffe, J. F. (1975). "Hot Issue" Markets. The Journal of Finance, 30(4), 1027.

Ibbotson, R. G., Sindelar, J. L., & Ritter, J. R. (1994). The market's problems wit the pricing of Initial Public Offerings. Journal of Applied Corporate Finance, 7(1), 66-74.

Jaskiewicz, P., González, V. M., Menéndez, S., & Schiereck, D. (2005). Long-Run IPO Performance Analysis of German and Spanish Family-Owned Businesses. Family Business

Review, 18(3), 179-202.

Jenkinson, T., & Ljungqvist, A. (2001). Going public: The Theory and evidence on How Companies Raise Equity Finance (2nd edition). Oxford University Press.

Levis, M. (1993). The Long-Run Performance of Initial Public Offerings: The UK Experience 1980-1988. Financial Management, 22(1), 28.

Levis, M. (2011). The Performance of Private Equity-Backed IPOs. Financial Management,

40(1), 253-277.

Loughran, T., & Ritter, J. (2004). Why Has IPO Underpricing Changed Over Time?

Financial Management, 5-37.

Loughran, T., & Ritter, J. R. (1995). The New Issues Puzzle. The Journal of Finance, 50(1), 23.

Miller, E. M. (1977). Risk, Uncertainty, and Divergence of Opinion. The Journal of Finance,

(33)

Obstfeld, M., & Rogoff, K. (2009). Global Imbalances and the Financial Crisis: Products of Common Causes. Federal Reserve Bank of San Francisco Asia Economic Policy Conference,

Santa Barbara, CA, October 18-20.

Pagano, M., Panetta, F., & Zingales, L. (1998). Why Do Companies Go Public? An Empirical Analysis. The Journal of Finance, 53(1), 27-64.

Ritter, J. R. (1991). The Long-Run Performance of Initial Public Offerings. The Journal of

Finance, 46(1), 3-27.

Rock, K. (1986). Why new issues are underpriced. Journal of Financial Economics, 15(1-2), 187-212.

Shiller, R. J. (1988). Initial Public Offerings: Investor Behavior and Underpricing. Working Paper No 2806, National Bureau of Economic Research, Cambridge.

Schöber, T. (2008), Buyout-Backed Initial Public Offerings, Dissertation of the University of

St. Gallen, no. 3479

Stehle, R., Ehrhardt, O., Przyborowsky, R. (20000. Long-run stock performance of German initial public offerings and seasoned equity issues. European Financial Management 6, 173-196

Viviani, D., Giorgino, M., & Steri, R. (2008). Private Equity-Backed IPOs and Long-Run Market Performance Analysis of Italian Firms. The Journal of Private Equity, 11(3), 50-60.

Zingales, L. (1995). Insider Ownership and the Decision to Go Public. The Review of