The transfer of football players and its subsequent effect on the share price of listed football clubs

Bachelor Thesis

The transfer of football players and its subsequent effect on the

share price of listed football clubs

Student: Friso Hoving Student number: 10381813

University: University of Amsterdam Specialization: Economics & Finance Supervisor: dhr. Aaron Kamm

1

1.STATEMENT BY THE AUTHOR

I hereby declare that this submission is my own work and to the best of my knowledge, it contains no material previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at any educational

2

2. Abstract

The transferring of players plays a prominent role in modern-day football. The takeovers of a number of major European football clubs by wealthy private investors brought in loads of new capital into the football industry. However, many football clubs across Europe are listed on the stock exchange and hereby publicly owned. This research investigates how the news of player transfers affects the share price of listed football clubs in terms of abnormal return. Data on 89 incoming and outgoing transfers subdivided over four European football clubs from different countries is examined. All transfers took place between the summer transfer windows of 2010 and 2014. Besides, all players were transferred for an amount of at least 5 million Euro. The results on all incoming transfers did not show an average abnormal return significantly different than zero, as expected. The average abnormal return on all outgoing transfers did show a positive abnormal return significantly different than zero though. The average abnormal return of all transfers

combined together was insignificantly different than zero again. This research consists of somewhat mixed results so further research is needed to determine the real effects of player transfers on the share prices of listed football clubs.

3

3. Table of contents

4

4.Introduction

The trading of players between clubs has always been common in football, like in any team sport (Frick, 2007). However it is hard for football clubs to estimate the expected future benefits of a player, because in the end a club only wants to acquire a player if he can improve the sportive performance of the team. Another aspect of a player transfer is the influence on the expected future financial performance. This can be expressed in the share price of a football club, provided that the club is listed on the stock exchange. The main focus of this research will be on this latter effect. More specifically the research question is the following:

Do the transfers of football players cause an abnormal return in the share price of listed football clubs?

For a firm can an initial public offering (IPO), the first sale of stock by a private company to the public, be a convenient way to raise large amounts of capital, which can be used for several reasons. For football clubs these raised capital is mainly invested in stadium renovations, player staff and the IPO increases the fan base of the club (Schaffer, 2005). Normally the motives of investors in

publicly traded firms are purely economic with the aim to make a profit on their investment. However, the benefits for a fan-investor are usually psychic rather than tangible. For those who purchase stock in a professional sports team, the main reason is to feel more connected to their favorite professional sport team and they often do not realize or even expect to make any return on their investment. Thus, from a fan’s perspective is the purchase of stock another way to support their club rather than a financial investment (Schaffer, 2005).

Nowadays there are over 30 European football clubs from different competitions listed on the stock exchange. The first European club that went public was Tottenham Hotspur F.C. in 1983 (Renneboog & Vanbrabant, 2000). In the years following more English football clubs performed an IPO as well. Some other major clubs from different countries that are listed on the stock exchange are Juventus F.C., AFC Ajax, Olympique Lyonnais and FC Porto. The financial state of these football clubs can be measured in its market value. Whenever a football club goes public the club will be divided into shares with a matching share price. These share prices will alter continuously under the influence of all kinds of developments in the market or specifically. Several

firm-5

specific developments can have an effect on the share price of a football club, such as the performance on the pitch or sponsor deals and the auction of television broadcasting rights.

Earlier research concludes that the stock market responds positive to victories and negative to defeats. The response to defeat is hereby stronger than the response to victory (Scholtens & Peenstra, 2009). However, to achieve sporting successes as a team, the quality of the players is decisive and in order to attract top players money is needed. Transfers with an high transfer amount could affect not only equity, debt and revenues for both football clubs involved in the transfer but investors on the stock exchange as well.

In order to answer the research question, existing literature on the pricing and trading of assets in general is examined. Also other variables that may have an effect on the share price of a football club will be discussed. In this thesis the data of multiple transfers with a significant transfer amount, at least 5 million Euros, of four different European football clubs will be collected and the exact effects on the share price will be measured by calculating whether there was an abnormal return during the period of these transfers.

Section 4 provides an overview of the existing literature on all relevant information regarding the research topic. In section 5 the methodology and research method will be explained and an the dataset will be clarified. In section 6 the results of this research will be shown and these results will be discussed in section 7. The final section 8 consists of the conclusion and the research question is attempted to be answered. In this section there will be recommendations for follow-up research as well.

6

5. Literature review

In this chapter the existing literature on initial public offerings (IPO) as well as special aspects of football clubs going public are being discussed. Then the importance of transfers in modern day football is explained and existing theories on how stock prices will react on announcements on the trading of assets is discussed.

First of all it is important to investigate why firms in general choose to go public. In most cases, the primary answer is the aim to raise equity capital for the firm and creating a public market in which the shareholders can convert their wealth into cash at a future date (Ritter & Welch, 2002). Thus, the main reason is to bring in more capital and hereby expand to a larger market. This holds for (major) football clubs as well since the sport of football is practiced worldwide and prices of some of the main assets from a club, transfers for example, have risen extremely over the years. For an overview of the ten international football transfers with the highest transfer amount of all time see the following table:

Player Season Transfer

amount (€)

Departing club Receiving club

Cristiano Ronaldo (Por) 09/10 102,2 Manchester United Real Madrid F.C. Zinedine Zidane (Fra) 01/02 101,92 Juventus F.C. Real Madrid F.C.

Bale (Wal) 13/14 100 Tottenham Hotspur FC Barcelona

Luís Figo (Por) 00/01 86,54 FC Barcelona Real Madrid F.C.

Zlatan Ibrahimovic (Swe) 09/10 75,02 Internazionale FC Barcelona

Hernán Crespo (Arg) 00/01 73,37 Parma F.C. S.S. Lazio

Kaká (Bra) 09/10 69,61 AC Milan Real Madrid F.C.

Edinson Cavani (Uru) 13/14 64 SSC Napoli Paris Saint Germain

Carlos Tévez 09/10 62,76 Manchester United Manchester City F.C.

Fernando Torres (Spa) 11/12 61,78 Liverpool F.C. Chelsea F.C. Table 1: top ten most expensive football transfers in history of football (www.football-times.net, 2015).

7

These amounts are adjusted for inflation using the annual consumer price index (CPI) from the country of the buying club. However, this table consist only of transfers until the summer transfer window of 2013. In the past two years there have been a couple of multi-million transfers that would probably have made the list as well. Most notable are the transfers of players like Ángel di Maria (Arg), Neymar jr. (Bra) and Luís Suarez (Uru), who all were transferred for an amount exceeding the 70 million Euros. Besides, some of the transfer amounts listed in the table are open to debate since clubs will not always disclose the exact amount that is payed due to certain financial constructions. However, it will give an insight in the amounts circulating in modern-day football.

One of the reasons why transfer amounts have risen the past decades is the involvement of wealthy investors, mostly from Eastern Europe and the Middle East, whose aim it is to gain status by taking over an European football club and invest in top players in order to win important prices, with the UEFA Champions League being the most prestigious one. In recent years European football have been dominated by the news regarding takeovers of big football clubs such as Manchester City F.C. from England and Paris Saint Germain from France. In both cases the new owners were extremely wealthy oil sheiks who brought in a lot of private capital, mainly used to invest in players for their first team. Although the origins and the amount of private capital of these investors were off a different order, the takeovers weren’t new to the game of football. In the English Premier League for example the most important source of funding is the capital injected by new owners, with an total amount of more than five times the total amount generated through stock exchange listings (Franck, 2010).

Nevertheless, many clubs throughout Europe are still publicly owned. Most football clubs which have carried out an IPO have done so to generate the funds required to improve their existing stadiums or to build new facilities (Cheffins, 1998). Cheffins (1998) also states that an additional type of expenditure that the public offerings of shares have financed is the purchase of playing staff since cash is traditionally the primary means of exchange in paying suitable transfer fees for

players. Renneboog and Vanbrabant (2000) acknowledge this by claiming that the main reason of an IPO is the need for additional funding needed to attract top players. Besides they state that clubs in lower divisions hope that this additional funding will create enough leverage to make the

promotion to the highest division in order to get access to even larger amounts of funding through promotion and the sale of television broadcasting rights.

An interesting feature of IPO’s is the fact that most of the time they are underpriced.

Abundant empirical evidence indicates that initial public offerings of common stock generate short-run returns, on average, for investors fortunate enough to purchase the stock for the offer price (Barry & Jennings, 1993). Many different theories have been suggested for this underpricing of

8

IPO’s. Barry and Jennings (1993) bring some of them forward, varying from: asymmetric information between the underwriters and the issuers, the underpricing being a mechanism to induce investors to produce and reveal private information and that high-quality issuers purposely underprice IPO’s to pave the way toward a more successful seasoned offering in the future. Another theory suggests that increased investment in research and development (R&D) is correlated with increased information asymmetries concerning the value of the firm and hereby with a higher level of underpricing (Heeley, Matusik & Jain, 2007).

Once a football club is listed on the local exchange market there are a lot of factors that will influence the share price of this club. On the one hand the results on the pitch will reflect in the share price since this tells investors worldwide something about the performance of the club. However, while match results affect the share price, these effects are modest compared with the changes in club stock prices caused by other variables (Bell, Brooks, Matthews & Sutcliffe, 2012). Another influence on the share price are major shifts in capital and assets, such as the transferring of players. The existence of the entire transfer-market in modern day football is important, because without the ability to sell their players on the transfer market, some clubs would inevitably end up with their operating position too weak to survive without transfer revenue (Magee, 2002). When a mediocre football club possesses a player who performs outstanding, it is only a matter of time before more wealthy clubs, like Real Madrid C.F. or Manchester United F.C., take an interest in this player and these clubs are able to offer him better terms of condition including a higher wage. To transfer the player both clubs needs to reach an agreement regarding the transfer fee, which often is the only financial compensation to the selling club for the loss of its player. Magee (2002) further states that many clubs were reliant on survival through transfer dealings and some lesser clubs would likely to be forced out of business without the safety net of transfer revenues.

As long as a player is under contract with a football club he is obligated to serve it out unless an appropriate transfer fee is payed, this is a characteristic of any contract after all. Football clubs were relatively in control of their players since clubs were able to negotiate about a transfer fee even though the players contract had ended. This changed in December 1995 when the European Court of Justice ruled that the commonly used transfer system for professional football players violated Article 39 of the Treaty of Rome in that it hampered the mobility of professionals (Feess & Muehlheusser, 2003). This verdict is known as the Bosman ruling, named after Jean-Marc Bosman. Bosman was offered a new contract with the Belgian club, RFC Liège, on inferior terms to his previous contract which had expired. Bosman’s club refused permission for him to join a French club, US Dunkerque. The player sued RFC Liège, citing restraint of trade (Simmons, 1997).

9

income to smaller clubs, a more unequal distribution of players’ salaries, lower transfer fees in the long-term and longer, more complex, contracts. Moreover, the percentage of player moves

involving payment of a transfer fee has declined from more than 95% in the 1980s to less than 40% in the more recent past. Clearly, this development is to be attributed to the Bosman ruling as well (Frick, 2007).

This ruling maybe is not directly of importance for this research since only data on transfers years after the ruling is examined, however it did cause more frequent transfers with lower transfer fees and it also increased the duration of contracts. Hereby it became more orderly to predict when a transfer is about to take place because, unless a new contract is signed, one year before expiration of a player’s contract, it is almost certain that the player in question is going to be transferred and this might have an influence on the share price due to speculations about the transfer in the media causing a potential abnormal return. So it is a relevant moment in the football industry since it transformed the transfer market to the way it is now.

Besides this development there are other variables that explain the variation in transfer fees. Frick (2007) states that player age, career games played and international caps all have a positive yet decreasing influence on the amount of money paid for the services of a player. Moreover, characteristics of the buying as well as the selling club have also been shown to influence transfer fees. The more successful the buying and/or the selling club is (either in economic or in sporting terms), the higher the transfer fee that the two clubs agree upon.

With football players being important assets for football clubs nowadays, there are different existing theories on trading of these assets. It is hard to value a football player though since every transfer is different and they do not occur frequently. Therefore, it is complicated to compare football players because they do not have a price in accordance with the market like shares have. This makes it hard to determine whether the amount of a transfer is undervalued or overvalued. The liquidity of the market is of importance on the pricing of an asset as well. When assets can be sold quickly without a discount, the market is considered to be liquid. If the market has many

transactions taking place, it means there are enough buyers and sellers present, hence a potential seller can find buyers without having to discount the price of an asset as much as it would have in a less liquid market (Schlingemann, Stulz & Walkling, 2001). This is the case with the transfer market, where during the transfer window many transactions between buyers and sellers take place. However, this period is limited to just three months in every season and not every player can be sold quickly without discounting its fundamental value. Only for top players are many potential buyers active and often they can be sold quickly with a profit since they already proved themselves

10

to be a valuable addition to a football club. All in all it is hard to determine to what extend the transfer market is liquid.

Another theory on the trading of assets and the subsequent effects on share prices is the efficient market hypothesis (EMH). This hypothesis is associated with the view that stock market price movements are approximately random. If new information develops randomly, then so will market prices, making the stock unpredictable (Malkiel, 2005). It is a worldwide acknowledged theory and in most researches markets are assumed to be efficient. However, Malkiel (2005) states that in recent years some financial economists and analysts were convinced that this hypothesis should be rejected since in several instances the market prices failed to reflect all information. The efficient market theory consist of three forms. Whenever the set of variables in the dataset only contains past and current prices, the weak form of the EMH is tested. The semi-strong form of EMH includes all publicly available information. At last, if all public and private information is included the strong form of EMH is being tested (Timmermann & Granger, 2004).

However, in most studies markets are assumed to be either weakly efficient or semi-strong efficient since private information often is hard to measure and hard to acquire. In efficient markets prices of securities reflect all available information to investors. So there are no deceptive investors and thus all investments in efficient markets are fairly priced. The price of the security reflects the present value of the expected future cash flows, which is influenced by multiple factors including volatility and liquidity (Clarke, Jandik & Mandelker, 2001). For simplicity is in this thesis the market assumed to be efficient and so there is no reason to expect that prices are too high or too low. Therefore, prices adjust before an investor has the time to trade on new information and subsequently make a profit (Clarke et al., 2001). This means that investors will not be able to make a return different than the expected return calculated by a normal performance model. So the hypothesis regarding the research question is the following:

H1: The transfer of football players do not cause an average abnormal return in the share price of a listed football club.

11

6. Methodology

Every season consists of two transfer windows, one during the summer from the first of July until 31 August and the other one in the winter from the first of January until 31 January. The time period used in this research is the period between the two summer transfer windows of respectively 2010 and 2014. This period is chosen because transfer amounts have risen during last couple of years (see table 1), so the more recent transfer window the more transfers with a significant transfer amount to examine there are. For each transfer to have a significant transfer amount in this research it must at least have a transfer amount of 5 million Euros. This amount is chosen random, but it is safe to assume that transfers of this magnitude are discussed extensively in the media in advance of the actual transfer, which might influence expectations of investors and hereby the potential abnormal returns. Furthermore, the chosen time period in this thesis is an interesting period of time in European football since the Union of European Football Associations (UEFA), the administrative body for association football in Europe, in 2010 introduced a system of financial regulation which they called Financial Fair Play (FFP). This is the reason why the time period investigated in this research is limited to 2010, because otherwise transfers that took place before this year would be subject to different regulations concerning the transferring of players. These set of regulations set certain criteria for any club to be met in order to take part in UEFA’s two main competitions, the Champions League and the Europa League. The two main criteria are the following:

1. No overdue payables. This means that a club must be fully up to date with payments to creditors. 2. Break-even. This means that a club must limit team spending on player wages and transfers to their revenues obtained from ‘football related activities’. For these purposes the balance of income and expenditure are calculated over a three-year period, and the balance is subject to an acceptable deviation of €5 million (Peeters & Szymanski, 2014).

Contrary to the increasing transfer amounts in recent years, the break-even rule in particularly should have caused a reduction of the number of transfers and of the transfer amounts (Peeters & Szymanksi, 2014). This means there are less multi-million transfers to investigate in these recent years then there would have been without the this set of regulation. However, one of the reasons

12

why this effect might not yet had an considerable effect is because the regulations came fully into force only recently at the end of the 2013/2014 season.

Incoming and outgoing transfers of three European football clubs are examined including Juventus F.C., F.C. Porto and BVB Dortmund. The selection of these football clubs from different European competitions is deliberate since it will improve the statistical power of the research. These specific competitions are chosen because they are all part of major European football competitions. However, the ‘big five’ European football competitions are the English Barclays Premier League, the Spanish Primera División, the German Bundesliga, the French Ligue 1 and the Serie A from Italy. F.C. Porto isn’t in one of these competitions but in the Portugese Primeira Liga, which is considered to be a very strong competition as well and besides this club has transferred many players for considerable amounts in recent years, which made it suitable for this research. Besides these three football clubs the outgoing transfers of A.F.C Ajax Amsterdam are examined as well. This club didn’t buy many players for a significant transfer amount in the period between the transfer windows of 2010 and 2014, but they did sell enough players for a significant transfer amount in this period of time to include in this research. Details of all transfers are retrieved from www.transfermarkt.nl as well as the annual reports of the football clubs. The following table shows the details of all transfers from the football clubs investigated:

Juventus F.C. F.C. Porto BVB Dortmund AFC Ajax Total Transfers (in) Transfers (out) Total 22 11 33 15 14 29 12 5 17 - 10 10 49 40 89 Amount (in) Amount (out) Total amount 250.150.000 95.780.000 345.930.000 126.950.000 301.000.000 427.950.000 138.100.000 79.500.000 217.600.000 - 113.530.000 113.530.000 515.200.000 589.810.000 1.105.010.000

Table 2: Overview of transfer statistics

This research does not contain clubs from all of this this five major competitions. There are quite some English football clubs listed on the stock exchange including big clubs from the Premier League: Arsenal F.C., Manchester United F.C and Tottenham Hotspur F.C.. However, these clubs were not suitable for this research since the shares of Arsenal F.C. are relatively infrequently traded on a specialist market, the ICAP securities and derivatives exchange, and Manchester United F.C. changed in 2012 to the New York Stock Exchange. Shares from Tottenham Hotspur F.C. were traded on the AIM Index only until 2012. Then the club went into private ownership of chairman Daniel Levy.

13

Football clubs from the Spanish Primera División on the other hand are missing since most football clubs in Spain, including the two biggest clubs Real Madrid F.C. and FC Barcelona, are so-called registered associations. This means that supporters are not able to buy shares of the football club, but instead they buy a membership of the club. In addition, the members elect all important organs of the club, such as the management board and the president, and they oversee the fulfillment of their respective tasks (Wilkesmann & Blutner, 2002). Other research shows that revenues from membership fees are even found the be the most important revenue source of sport clubs across Western countries (Wicker, Breuer & Hennigs, 2012). This set-up allows supporters to become more engaged with their favorite football club but it also has its downsides. Wilkesmann and Blutner (2002) further state that clubs as a registered association often face the classical dilemma of having to represent the will of the members and follow the aims of the organization and individual interests usually conflict with the more fundamental strategic decisions of the club and hence the more members a club has the more likely it is that decisions will block the efficient accomplishment of broader policy goals. It is unfortunate that these two football clubs cannot be included in the research since these clubs always claim a prominent role in the transfer market with big signings in recent years. More than that, in the five biggest international transfers of all time one of these clubs was involved including the transfer of Gareth Bale for an amount of 100 million Euros two seasons ago, although this is never confirmed officially. See table 1 for an overview of the ten international football transfers with the highest transfer amount of all time.

This research can be considered as an event study. In economics, and in this research using data from financial markets, an event study can be used to measure the impact of a specific event on the firms’ value (MacKinlay, 1997). In this research the impact of a transfer on the firms’ value, expressed in the share price, is measured. MacKinlay (1997) further states that such a study is useful since the economic impact of the event can be derived using prices observed over a relatively short period of time as long as there is a certain rationality in the marketplace so that the effects of an event will be reflected in security prices immediately. This is the case in this research since the market is assumed to be efficient. This is convenient since in this research only share prices of a short period of time before, during and after the transfer will be examined to find out whether there is an abnormal return. MacKinlay (1997) states that besides the day of the event and the day after the event, where potential price effects of the transfer will occur, the period prior to the event is of importance as well since the market may acquire information in the media about a transfer before the official statement concerning this transfer is made. More specifically the event window in this research is limited to only one day before the day of the transfer until one day after the day of the transfer. Often transfers of at least 5 million Euro are discussed in the media for a vast period of

14

time in advance, however the event window in this research is limited to only one day prior to the actual transfer since most of the time this is the moment it is absolute sure that the transfer will take place. The advantage of such a short period of time is that most likely the transfer in question is the only firm-specific shock that can cause a potential abnormal return in the share price.

Thus, abnormal returns must be calculated. Before deriving the abnormal returns, first the expected (normal) return must be determined. This is the expected return when the event would not have occurred. The abnormal return can now be calculated be taking the difference between the actual and the normal return (Scholtens & Peenstra, 2009). At first the expected return is calculated, which can be done by different models including the Capital Asset Pricing Model (CAPM), the constant mean return model or the market model. For this research the market model is used since it is well-specified and powerful under a wide variety of conditions and there is no evidence that more complicated models will be more beneficial (Brown & Warner, 1980). Besides, in the market model is the variance of the abnormal return reduced by removal of the portion of the return that is related to variation in the market’s return. This may increase the potential to detect event effects

(MacKinlay, 1997). The market model is defined as:

R*jt = αj + βj * Rmt + εjt

Where R*jt is the expected return of firm ‘j’ in period t,αj is the intercept term, βj represents the systematic risk of firm ‘j’,Rmt is the actual return of the market in period t andεit is the error term with an expected value of zero and a variance of σ2. The model is based on a stable linear relation between the return on the market and the return on the security ‘j’. The data from all clubs

regarding return in share prices as well as the return on the market indices are retrieved from DataStream. In order to calculate the expected return, the following market indices were used: the Financial Times Stock Exchange Milano Indice Borsa (FTSE-MIB), the Portuguese Stock Index 20 (POPSI20), the Deutsche Aktienindex (CDAXGEN) and the Amsterdam Exchange Index (AEX). Hence, βj can be estimated using an ordinary-least-square regression (OLS) of the return on the market and the return of the specific football club during a certain estimation period. This

estimation window must be defined. For this research an estimation window of 100 days before the event window of a transfer is chosen. It is common to use the period before the event window and it is important that the event period itself is not included in the estimation period otherwise the

parameter estimations in the market model could be influenced by the event (MacKinlay, 1997). The OLS-regressions are performed with a confidence level of 95%. These model must be

15

calculated manually for every transfer independently since every transfer is a separate event with a corresponding estimation period. After the expected returns are established the abnormal return can be calculated as:

ARjt = Rjt - R*jt

Where ARjt is the abnormal return of firm ‘j’ in period t, Rjt is the actual return of firm ‘j’ in period t and R*jt is the expected return of firm ‘j’ in period t. When all abnormal returns are calculated the average abnormal returns (AAR) can be determined for the incoming transfers, outgoing transfers and all transfers together of every football club separately and in total. Afterwards, the significance of these results must be determined. This is done using multiple one sample t-test. For this purpose the following formula is used:

t =(𝑥𝑥 −  0)

16

7. Results

The following table shows the results of the effect for all incoming transfers, all outgoing transfers and all transfers together respectively on the share price in terms of abnormal return. These results are tested against an two-tailed significance level (α) of 5%.

All incoming transfers All outgoing transfers All transfers

Average Abn. Return -0,6624% 1,5452% _0,3298%

Standard Deviation 4,098% 2,924% _3,764%

Sample size 49 40 ₈₉

t-statistic -1,1316 3,3418 _0,8265

Critical t-value 2,01 2,02 _1,99

P-value 0,2634 0,0018 _0,4108

Table 3: Overview of results one sample t-test

For an incoming transfer of the football clubs investigated in this research, the effects in terms of abnormal return are -0,66% on average. With a p-value of 0,26 this result is not significantly different than 0 at neither a 5% or a 10% significance level. So the null hypothesis that a transfer will not cause an abnormal return on the share of a listed football club holds regarding incoming transfers.

In the case of an outgoing transfer, the abnormal return of a football club is 1,55% on average. With a p-value of 0,0018 this result is significantly different than 0 at a significance level of 5% and even for a significance level of 1%. This is not in line with the null hypothesis that the transfer of a football player will not cause an abnormal return on the share of a listed football club.

When all transfers together are being examined however, the abnormal return is 0,33% on average. This result is again not significantly different than 0 at all reasonable significance levels of 1%, 5% or even 10%. Thus, the null hypothesis will not be rejected in this case; all transfers

17

8. Discussion

As forecasted, do all transfers together not cause a significant abnormal return. The results on the incoming and outgoing transfers separately are somewhat remarkable though. The average

abnormal return on incoming transfers is negative, which means that the actual returns in the event windows of all these transfers was smaller on average than the expected returns based on the assumptions that these transfers would not occur. This result is contrary to the average abnormal return on outgoing transfers, which is positive. This means that the actual returns in the event windows of all outgoing transfers on average are bigger than the expected return on the share of the football club. Investors apparently attach more value to the money a selling football club receives for the transfer of a player than to the signing of a new player. This implies either investors do not fully believe in the added value of an incoming transfer or the transfers investigated are overpriced.

It is hard to appoint the exact reason why incoming transfers do not cause a significant abnormal return different than zero while outgoing transfers do cause a certain abnormal return. Many incoming and outgoing transfers took place at the same date, which could have had opposite effects on expectations and following investments. In the case of all transfers these effects are incorporated and together with an increased sample size these results have a higher statistical power. The market does not react to the news of the transfer, incoming or outgoing, of a player. This should mean the assets actually are priced correctly.

However, there are some shortcomings to the research method which might have caused the somewhat opposing results in incoming and outgoing results. There could have been other firm specific effects that had an influence on the share price of the football club in the event window of the transfers. These potential effects are limited though due to the small event window of three days, but in the case of some transfers the event window and estimation period for the market model did overlap one or two days because these transfers took place in the weekend, when the financial markets are closed. The event window had to be extended with one or two days in order to measure returns on the market and on the share of the football club. Hereby are the estimated parameters of the market model slightly biased. Besides, some transfers are made up of multiple interest parties sharing the payment of the transfer amount for a player and together with other clauses in the contract this causes other clubs to remain being involved in the transfer rights of this player. This might influence the expected added value of a transfer for investors and hereby the investment and

18

following return on the share of the football clubs involved. These above mentioned influences are equally likely to occur with incoming as outgoing transfers though.

Another notable observation is that many transfers took place on the same date during the investigated period of time. For example, multiple players are transferred on the first of July since this is the first day of the transfer window and deals on transfers are often officially executed on this day. Earlier research showed that event studies are prone to so-called cross-sectional correlation among abnormal returns when the day of an event is the same for different firm samples (Kolari &

Pynnönen, 2010). As a result are test statistics unable to assume complete independence of abnormal returns. Even when the cross-correlation is relatively low, the clustering of event dates can cause a serious over rejection of the null hypothesis of zero abnormal returns on average. This might be the case in the rejection of the null hypothesis regarding the average abnormal returns of all outgoing transfers. Other research suggests that tests for mean abnormal returns should therefore preferably use standard errors that are robust to cross-sectional variation in true abnormal returns (Harrington & Shrider, 2007). To prevent the problem of cross-sectional correlation, scaled

abnormal returns should be used instead. Scaled abnormal returns are the abnormal returns divided by the standard deviation of the residuals in the normal performance model, used to calculate the estimation period, corrected by the prediction error. Tests on scaled abnormal returns are proven to be superior in terms of power over tests based on non-scaled returns (Kolari & Pynnönen, 2010).

19

9.Conclusion

The focus of this research is to investigate the effect of the transfer of football players on the share price of the clubs involved in terms of abnormal return. With Tottenham Hotspur F.C. being the first football club that went public in 1983, many European clubs followed mainly to raise additional capital to expand their stadium and to invest in new players. There are several other factors influencing the share price of a football club such as match results for example. However, with the involvement of wealthy investors from all over the world in European football in recent years, most of this capital is spend on the transfer of top players, which increased the amount payed for these players.

The efficient market hypothesis states that prices of securities reflect all available, public and private, information to investors. These prices are based on the present value of the expected future cash flow, which results in all securities being fairly priced. Prices adjust before investors have time to act on a new piece of information and they are unable to make a return different than the expected return calculated by a normal performance model. This means that the event of a transfer will not cause an abnormal return on the share of the football club involved. The results regarding incoming transfers show a negative average abnormal return. However, these result are not significantly different than zero, which is in line with the expectation stated above.

When investigating outgoing transfers there is an opposite outcome though. These results show a positive abnormal return significantly different than zero even at a 1% significance level. Although several different factors can be the cause of this opposite result, one factor is the fact that multiple event days in this research are the same among the sample firms. This causes cross-sectional correlation in the abnormal returns, which is followed by a serious over rejection of the null hypothesis stating that the abnormal returns on average should be zero. In the case of all incoming and outgoing transfers added together the average abnormal return was positive. The return is not significantly different than zero though. This is in line with the expectation based on earlier research again.

This research is based on 89 observations of only four sample clubs. In order to investigate the real effects of transfers on the share price, additional research is needed. Although the presence of listed European football clubs is relatively low, more observations from different football clubs across different countries should increase the statistical power of the results. It could also be useful

20

to make a distinction between transfers with a high transfer amount and transfers with a lower transfer amount. Including extra transfers will increase the number observations and comparing the results of both groups could possibly be effective as well. To avoid the problem of cross-sectional correlation, additional tests using scaled abnormal returns can be done to test whether the abnormal returns are on average significantly different than zero. Commonly used scaled tests are the t-statistics of Patell (1976) or Boehmer, Musumeci and Poulsen (1991).

21

10.References

Barry, C. B., & Jennings, R. H. (1993). The opening price performance of initial public offerings of common stock. Financial Management, 54-63.

Bell, A. R., Brooks, C., Matthews, D., & Sutcliffe, C. (2012). Over the moon or sick as a parrot? The effects of football results on a club's share price. Applied Economics, 44(26), 3435-3452. J Magee - Power games: A critical sociology of sport, 2002

shifting balances of power in the new football economy

Boehmer, E., Masumeci, J., & Poulsen, A. B. (1991). Event-study methodology under conditions of event-induced variance. Journal of Financial Economics, 30(2), 253-272.

Brown, S. J., & Warner, J. B. (1980). Measuring security price performance.Journal of financial

economics, 8(3), 205-258.

Cheffins, B. R. (1998). Playing the Stock Market: Going Public and Professional Team Sports. J.

Corp. L., 24, 641.

Clarke, J., Jandik, T., & Mandelker, G. (2001). The efficient markets hypothesis. Expert Financial

Planning: Advice from Industry Leaders, 126-141.

Feess, E., & Muehlheusser, G. (2003). The Impact of Transfer Fees on Professional Sports: An Analysis of the New Transfer System for European Football*. The Scandinavian Journal of

Economics, 105(1), 139-154.

Franck, E. (2010). Private firm, public corporation or member's association governance structures in European football. International Journal of Sport Finance, 5(2), 108-128.

Frick, B. (2007). The football players’ labor market: Empirical evidence from the major European leagues. Scottish Journal of Political Economy, 54(3), 422-446.

22

Harrington, S. E., & Shrider, D. G. (2007). All events induce variance: Analyzing abnormal returns when effects vary across firms. Journal of Financial and Quantitative Analysis, 42(01), 229-256.

Heeley, M. B., Matusik, S. F., & Jain, N. (2007). Innovation, appropriability, and the underpricing of initial public offerings. Academy of Management Journal, 50(1), 209-225.

Kolari, J. W., & Pynnönen, S. (2010). Event study testing with cross-sectional correlation of abnormal returns. Review of Financial Studies, hhq072.

MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of economic literature, 13-39.

Malkiel, B. G. (2005). Reflections on the efficient market hypothesis: 30 years later. Financial

Review, 40(1), 1-9.

Patell, J. M. (1976). Corporate forecasts of earnings per share and stock price behavior: Empirical test. Journal of accounting research, 246-276.

Peeters, T., & Szymanski, S. (2014). Financial fair play in European football.Economic

Policy, 29(78), 343-390.

Renneboog, L., & Vanbrabant, P. (2000). Share price reactions to sporty performances of soccer

clubs listed on the London Stock Exchange and the AIM. Tilburg University.

Ritter, J., & Welch, I. (2002). A review of IPO activity, pricing, and allocations (No. w8805). national bureau of economic research.

Schaffer, R. (2005). Piece of the Rock (or the Rockets): The Viability of Widespread Public Offerings of Professional Sports Franchises, A. Va. Sports & Ent. LJ, 5, 201.

Scholtens, B., & Peenstra, W. (2009). Scoring on the stock exchange? The effect of football matches on stock market returns: an event study. Applied Economics, 41(25), 3231-3237.

23

Simmons, R. (1997). Implications of the Bosman ruling for football transfer markets. Economic

Affairs, 17(3), 13-18.

Timmermann, A., & Granger, C. W. (2004). Efficient market hypothesis and forecasting. International Journal of Forecasting, 20(1), 15-27.

Wicker, P., Breuer, C., & Hennigs, B. (2012). Understanding the interactions among revenue

categories using elasticity measures—Evidence from a longitudinal sample of non-profit sport clubs in Germany. Sport Management Review, 15(3), 318-329.

Wilkesmann, U., & Blutner, D. (2002). Going public: The organizational restructuring of German football clubs. Soccer and Society, 3(2), 19-37.

24

11. Appendix

Juventus F.C.

Event Player Amount

(mill.€) Ab-normal Return % 12-08-10 Poulsen 5,48 0.1136071 27-08-10 Diego 15,5 -2.164520 25-07-11 Sissoko 8 -0.629729 10-07-12 Elia 5,5 -1.043921 06-08-12 Krasic 7 2.6110775 09-07-13 Gabbiadini 11,5 0.4131663 10-07-13 Zaza 10 0.9517117 16-07-13 Giaccherini 7,5 0.1223087 30-08-13 Matri 11 -0.832201 01-07-14 Immobile 8 2.2646277 04-07-14 Vucinic 6,3 0.5079355 Total: Average: 95,78 8,71 2.3140624 0.2103693

Overview outgoing transfers

Overview incoming transfers

Event Player Amount

(mill.€) Ab-normal Return % 01-07-10 Bonucci 15,5 0.4022005 01-07-10 J. Martinez 12 0.4022005 24-08-10 Krasic 15 -0.047449 01-07-11 Matri 15,5 0.3934655 01-07-11 Quagliarella 10,5 0.3934655 01-07-11 Lichtsteiner 10 0.3934655 01-07-11 S. Pepe 7,5 0.3934655 21-07-11 Vidal 12,5 -0.719091 01-08-11 Vucinic 15 0.9839740 25-08-11 Giaccherini 7,25 -1.066526 31-08-11 Elia 9 0.2304264 01-01-12 Padoin 5 1.3088006 01-07-12 Cáceres 8 -0.601645 01-07-12 Giovinco 11 -0.601645 02-07-12 Asamoah 18 0.2055693 02-07-12 Isla 13,9 0.2055693 24-08-12 Gabbiadini 11 0.9503952 01-07-13 Tevez 9 -0.237140 11-07-13 Ogbonna 13 -2.496158 01-07-14 Sturaro 6,5 2.2646277 03-07-14 Marrone 5 -0.763514 19-07-14 Morata 20 1.1566009 Total: Average: 250,15 11,37 3.1510552 0.1432297

25

FC Porto

Event Player Amount

(mill.€) Ab-normal Return % 03-08-10 Bruno Alves 22 2.0645615 29-08-10 Raul Meireles 13 1.1202309 01-07-11 Rúben Micael 5 2.4668259 18-08-11 Falcao 40 -3.218477 01-07-12 Guarín 11 0.4260896 07-08-12 Pereira 12 -0.180627 03-09-12 Hulk 40 -0.229017 01-07-13 J. Rodríguez 45 5.8050920 01-07-13 Moutinho 25 5.8050920 01-07-14 Fernando 15 6.1761367 01-07-14 Iturbe 15 6.1761367 01-07-14 Otamendi 12 6.1761367 11-08-14 Mangala 40 5.5268701 13-08-14 Defour 6 -2.658385 Total: Average: 301 21,5 35.45666 2.532618

Overview outgoing transfers

Overview incoming transfers

Event Player Amount

(mill.€) Ab-normal Return % 03-07-10 Moutinho 11 0.0704160 06-07-10 J.Rodríguez 7,35 -4.340972 31-08-10 Otamendi 8 -3.520551 01-01-11 Danilo 13 -6.540334 01-01-11 Alex Sandro 9,6 -6.540334 01-08-11 Defour 6 -12.69113 16-08-11 Mangala 6,5 -14.02342 07-07-12 J. Martínez 8,8 -7.097256 01-07-13 Herrera 8 5.8050920 01-07-13 Diego Reyes 7 5.8050920 13-07-13 Quintero 9,5 -8.881795 12-07-14 Adrián 11 -1.042617 15-07-14 Martins Indi 7,7 -4.172855 16-07-14 Brahimi 6,5 -2.458993 01-09-14 Otávio 7 -0.386653 Total: Average: 126,95 8,46 -60.01632 -4.001088

26

BVB Dortmund

Event Player Amount

(mill.€) Ab-normal Return % 01-07-11 Nuri Sahin 10 4.5656496 01-07-12 Kagawa 16 3.8870813 01-07-12 Lucas Barrios 8,5 3.8870813 06-01-13 Perisic 8 2.1094037 01-07-13 Gotze 37 -1.6007970 Total: Average: 79,5 15,9 12.848418 2.5696837

Overview outgoing transfers

Overview incoming transfers

Event Player Amount

(mill.€) Ab-normal Return % 01-07-11 Gundogan 5,5 4.565649 01-07-11 Perisic 5,5 4.565649 01-07-12 Reus 17,1 3.887081 01-07-12 Schieber 5,5 3.887081 01-07-13 Sokratis 9,9 -1.600797 04-07-13 Aubameyan g 13 2.512660 09-07-13 Mkhitaryan 27,5 -0.793821 01-07-14 Adrián Ramos 9,7 3.822951 01-07-14 Nuri Sahin 7 3.822951 02-07-14 Immobile 19,4 0.395099 17-07-14 Ginter 10 -0.687081 31-08-14 Kagawa 8 0.028382 Total: Average: 138,1 11,51 24.40580 2.033817

27

AFC Ajax Amsterdam

Overview outgoing transfers

One-sample t test

---

Variable | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval] ---+---

incomi~r | 49 -.0066244 .005854 .0409782 -.0183947 .0051459

--- mean = mean(incomingar) t = -1.1316 Ho: mean = 0 degrees of freedom = 48 Ha: mean < 0 Ha: mean != 0 Ha: mean > 0 Pr(T < t) = 0.1317 Pr(|T| > |t|) = 0.2634 Pr(T > t) = 0.8683

Results t-test all incoming transfers

Event Player Amount

(mill.€) Ab-normal Return % 31-01-11 Suárez 26,5 2.64150434 01-07-11 De Zeeuw 6 -0.6438069 01-08-11 Stekelenburg 7,33 -0.0570166 12-07-12 Verthongen 12,5 -3.4648957 16-08-12 Anita 8,5 0.04243549

03-09-12 Van der Wiel 6 4.93528271 30-08-13 Eriksen 13,5 -0.7757802 02-09-13 Alderweireld 7 0.41371354 01-07-14 S. De Jong 8,7 0.06146396 01-09-14 Blind 17,5 8.03666498 Total: Average: 113,53 11,35 11.1895653 1.11895653

28 One-sample t test

---

Variable | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval] ---+---

Out | 40 .0154522 .0046239 .0292443 .0060994 .0248049

--- mean = mean(Out) t = 3.3418 Ho: mean = 0 degrees of freedom = 39

Ha: mean < 0 Ha: mean != 0 Ha: mean > 0

Pr(T < t) = 0.9991 Pr(|T| > |t|) = 0.0018 Pr(T > t) = 0.0009

Results t-test all outgoing transfers

One-sample t test

---

Variable | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval] ---+---

alltra~r | 89 .0032977 .00399 .037642 -.0046317 .011227 ---

mean = mean(alltransfersar) t = 0.8265

Ho: mean = 0 degrees of freedom = 88

Ha: mean < 0 Ha: mean != 0 Ha: mean > 0

Pr(T < t) = 0.7946 Pr(|T| > |t|) = 0.4108 Pr(T > t) = 0.2054