Print Reference This Reddit This

Study On Abnormal Returns Based On Different Things Finance Essay

Introduction

Today, financial markets are counted among the highest regulated markets of our economy. Market integrity plays a crucial role in respect to the efficiency and the integration of financial markets. The public confidence in financial markets, especially after the undergone financial crisis, can only be obtained with a high degree of transparency. Market abuse therefore strongly harms the integrity of financial markets and public confidence in securities and other financial instruments. Insider trading is often considered as market abuse on the part of corporate insiders as they are in possession of qualitatively better and reliable information about their companies’ future prospects. This perception is not true per se and a firm distinction between legal and illegal insider trading is indispensable. Both terms are defined by insider trading regulations. One of the very first insider trading laws was enacted by the American Congress in 1934. According to Bhattacharya and Daouk (2002), the existence of insider trading laws and their enforcement, except for the cases of the United States (US), United Kingdom (UK), Canada and France, can essentially be regarded as a phenomenon of the 1990s and today nearly all developed countries have specific regulation frameworks for insider trading. However, Switzerland represents a special case as the scope of offences of illegal insider trading was surprisingly narrow until 2008. Although Switzerland has done a big step closer towards the EU’s directive on market misuse and is following the Anglo-Saxon perception that disreputable behaviour of market players should be stronger penalized, the regulation on insider trading in other developed countries is still more stringent than in Switzerland. This paper gives a detailed account of the current jurisdiction of insider trading in Switzerland compared to the European Union (EU) and the US. There are remarkable differences between the three sets of regulations, especially with regards to (i) the definition of insider and legal insider trading, (ii) the length of the reporting period of insider trades, (iii) the application of trading bans and (iv) the imposed penalties in case of violation of existing law.

The main purpose of this paper is to examine whether corporate insiders were able to earn abnormal returns on the Swiss, German and Austrian stock market during the investigation period from July 2005 to December 2009. Special attention is devoted to the research question whether there are striking differences concerning the results when periods of economical up- and downturn are investigated independently. Furthermore it is investigated whether the firm size, the transaction volume or the fact that several management transactions of the same company took place on a particular day have a significant influence on the abnormal returns. Finally, it is questioned whether outsiders are able to earn abnormal returns by mimicking published management transactions. This particular examination will help to define the degree of market efficiency of the three stock markets during the investigated time frame. According to the Efficient Market Hypothesis stated by Fama (1970), the strong form of market efficiency can only be reached if asset prices fully reflect all public and private information available. In a nutshell, there is no possibility to earn abnormal returns neither by taking advantage of insider knowledge nor by mimicking insider positions based on public information. Furthermore, Fama’s semi-strong form of market efficiency suggests that only information that is not publicly available can benefit investor’s seeking to generate abnormal returns on investment. Hence, only corporate insiders might earn abnormal returns by exploiting their informational edge. Consequently, the strategy of mimicking management transactions does not yield to significant abnormal returns.

The remainder of this paper is organized as follows. The next section reviews the existing literature on insider trading. Section 3 summarizes the Swiss jurisdiction on insider trading compared to the EU and the US. Section 4 provides a description and initial characterization of the data sample. Section 5 discusses the methodology of the event study. Section 6 analyses the empirical results and section 7 concludes.

Literature Review

The need for insider trading regulation and law has been widely debated in the literature. On the one hand, opponents of insider trading regulation bring forward the argument that restrictions are inefficient because insider trading contributes to more informative stock prices in that prices reflect firm’s intrinsic value in a more accurate way (Carlton and Fischel, 1983; Dye, 1984). Meulbroek (1992) investigated illegal insider trades detected by the Securities and Exchange Commission (SEC) and came to the conclusion that insider trading is discovered by the stock market and a substantial proportion of the information is integrated in the stock price before it becomes public. She further states that regulation impeding insider trading may result in less informative prices. Amman and Kessler (2004) find that in general the trading volume before publications of earning figures dries-out what is ascribable to investors waiting for new information before trading. They could thus not determine systematic insider trading on the Swiss stock market before the publication of financial results. The mentioned study investigated the time period of 1997 to 2003, when management transactions in the run-up of price-sensitive corporate news did not yet fall in the category of illegal insider trading. Hence, it can be assumed that the legal imposition of blackout periods would not substantially improve the situation of outsiders during pre-announcement periods in Switzerland. On the other hand, an opposite view theorizes that insider trading deters outside investors from information collection as the gains available to them are limited by insider trading (Fishman and Hagerty, 1992). If such a crowding-out effect dominates, information asymmetry is increased what leads to a diminished trading willingness of uninformed investors (Ausubel, 1990). Leland (1992) further argues that a discouragement of outsiders can cause decreasing stock market participation with a lower degree of liquidity and efficiency. More recent work by Bettis, Coles and Lemmon (2000) was focused on corporate policies concerning the regulation of insider trading in the US. They found evidence that blackout periods imposed by corporate policy successfully restrain trading by insiders and furthermore lead to narrower bid-ask spreads by about two basis points. Proponents therefore allege that insider trading regulation positively influences public confidence and participation in the stock market.

Today, corporate insiders in most developed countries are legally obligated to disclose their undertaken transactions in securities of their own company. Since this development can be characterized as a phenomenon of the 1990s (Bhattacharya and Daouk, 2002), previous research is basically focused on data from the US and the UK. Only in the last ten to fifteen years has academic research widely started to investigate insider trading in nations different to the Anglo-Saxon’s.

Abnormal Returns of Corporate Insiders

One of the very first studies analyzing whether insider trading yields abnormal returns was done by Lorie and Niederhoffer (1968). The authors reach the conclusion that shares strongly in demand by corporate insiders outperform the Dow Jones Industrial Index during the following six months, however, the degree of abnormal return is not specified. Applying more powerful techniques and data sets, Jaffe (1974) and Finnerty (1976) find that insiders are able to anticipate and take advantage of residual return in the near future. They identify corporate insiders as contrarian investors (buying past losers, selling past winners), but better at timing the market than simple contrarian strategists. Furthermore, they state that corporate insiders obtain significant abnormal returns with purchase and sale transactions. These findings are supported by later studies of Seyhun (1986, 1988), Rozeff and Zaman (1988), Lin and Howe (1990), Gregory, Matatko, Tonks and Purkis (1994), Hillier and Marshall (2002), Piotroski and Roulstone (2005) and Betzer and Theissen (2009). Other studies state that purchase transactions are followed on average by larger abnormal returns than sale transactions (e.g. Friederich, Gregory, Matatko and Tonks, 2002; Fidrmuc, Goergen and Renneboog, 2006; Bajo and Petracci, 2006; Dymke and Walter, 2008). This result is explained by the fact that sale transactions are often due to liquidity needs and portfolio rebalancing objectives, whereas purchase transactions are assumed to incorporate favourable information. On the other hand, several studies cannot find significant abnormal returns for insider sale transactions (e.g. Lakonishok and Lee, 2001; Jeng, Metrick and Zeckhauser, 2003). Eckbo and Smith (1998) state that firms listed on the Oslo Stock Exchange do not earn abnormal profits at all, regardless of purchase or sale transactions.

Abnormal Returns based on Firm Size

Literature supplies strong evidence that the Small Firm Effect of Fama and French (1992) is also attributable to insider trading. Seyhun (1986) was among the first to detect significant higher abnormal returns for insiders of small caps than for those of large firms. Hence, he identifies a negative relation between abnormal returns of insiders and the correspondent firm size. These findings were later confirmed by Wong, Cheung and Wu (2000), Lakonishok and Lee (2001) and Zingg, Lang and Wyttenbach (2007). Gregory, Matatko, Tonks and Purkis (1994) underline these findings with the explanation that directors of small caps are probably more acquainted with the future prospects of their companies than directors of larger companies, as the latter generally include more non-executive directors. In addition, Lakonishok and Lee (2001) argue that smaller firms have more information asymmetries between managers and investors what can be assumed to yield larger profits for insiders. Hillier and Marshall (2002) and Piotroski and Roulstone (2005) confirm these findings by saying that small firms have a less active analyst coverage and thus a weaker information environment resulting in prices of small firms not fully reflecting information and thus insiders can more effectively use private information. The study of Jeng, Metrick and Zeckhauser (2003) contradicts the above stated findings and alleges that purchases in small firms do not yield significantly higher returns than do purchases in large firms.

Abnormal Returns based on Trading Volume

The academic literature does not provide a clear answer to the question whether there is a correlation between abnormal returns and trading volumes of insider transactions. According to Seyhun (1986), corporate insiders are able to identify the value of their inside information and therefore tend to trade greater volumes in order to exploit more valuable information. Evidence that high-volume trades by insiders generally yield higher abnormal returns, is also found by Jeng, Metrick and Zeckhauser (2003) and Fidrmuc, Goergen and Renneboog (2006). Hence, high- volume trades by insiders are associated with strong information content. Especially high-volume selling trades could signal a less prosperous future of the company in question. In contrast, the Stealth Trading Hypothesis of Barclay and Warner (1993) states that insiders tend to strategically disguise their high-volume trades by splitting them up into several medium-sized trades in order to avoid strong signal effects. Friederich, Gregory, Matatko and Tonks (2002) discover this hypothesis in their research and thus state that medium-sized trades are more informative for short-term returns than large ones. Contrary, Givoley and Palmon (1986) do not find any relation between the volume of insider trades and abnormal returns.

Abnormal Returns based on Clustered Transactions

According to Fidrmuc, Goergen and Renneboog (2006), clustered transactions are perceived as stronger and more credible by outside investors. Hence, when several corporate insiders purchase their company’s stocks on the same day, they communicate a positive signal about the future prospects. In the case of several sale transactions, the opposite signal is sent to uninformed investors. Friederich, Gregory, Matatko and Tonks (2002) confirm this hypothesis for the UK stock market and state that larger abnormal returns of clustered management transactions are found. Hillier and Marshall (2002) argue that corporate insiders tend to sell together after a large increase in price what surprisingly does not strongly affect the share price abnormal returns of the security after the trade, as one would may expect. Hence, they state that the strength of the market reaction to insider sales is weaker than for purchase transactions, as sale transactions are mostly driven by liquidity and diversifying reasons.

Abnormal Returns based on Mimicking Strategy

Another controversial discussed issue is the question whether the mimicry of published insider trades by outsiders is a successful investment strategy. Early research by Jaffe (1974) and Finnerty (1976) show that outsiders can earn abnormal profits by imitating insider trades. This finding has been re-examined by Seyhun (1986) who states that outsiders indeed earn a gross return of 1.4 per cent in the 100 days following the public release of the insider transaction, but this level of return is certainly unprofitable after the deduction of transaction costs. Rozeff and Zaman (1988), Lin and Howe (1990) and Friederich, Gregory, Matatko and Tonks (2002) confirm that potential abnormal returns by mimicking outsiders are whittled down to zero because of transaction costs. Del Brio, Miguel and Perote (2002) do not find any evidence at all for the profitability of mimicking strategies on the Spanish stock market. This evidence is consistent with the semi-strong form of Famas’ Efficient Market Hypothesis. In contrast, Bajo and Petracci (2006) state that mimicry is a profitable investment strategy on the Italian stock market. According to Stotz (2006) and Dymke and Walter (2008) the same holds true for the German stock market where outsiders nearly obtain the same abnormal returns as insiders do. Bettis, Vickrey and Vickrey (1997) and Lakonishok and Lee (2001) state that outsiders are indeed able to yield positive net abnormal returns, but only when mimicking large insider trades executed by the top management.

Comparative Law Analysis

US Jurisdiction on Management Transactions

After the Wall Street Crash of 1929, the US Congress reacted to this occurrence by enacting the first insider trading laws, namely the Securities Act in 1933 and the Securities Exchange Act (SEA) in 1934, respectively. The aim of these insider laws was to insure the maintenance of fair and honest markets. The core of the modern federal insider trading prohibition derives its statutory authority directly from section 16b and indirectly from section 10b rule 10b-5 of the SEA.

However, the term insider is not clearly defined by statute in the context of the US prohibitions against insider trading. Today, virtually anyone who possesses material non-public information is required to either disclose before trading or to abstain from trading.

Before the enforcement of the Sarbanes-Oxley Act (SOX) of 2002, section 16a of the SEA said that every person who is a director, an officer or directly or indirectly the beneficial owner of more than ten per cent of any class of the company’s equity securities (henceforward liable individual[s]), the issuer has to report the transactions by the tenth day of the calendar month after the trading month, leading to a maximum time lag of 40 trading days before the public is informed about the undertaken transaction. By implementing the provisions of the SOX section 403a, the Security and Exchange Commission (SEC) adopted new rules and amendments to Section 16 of the SEA. Today, shareholders meeting the definition requirements of a director, an officer or a principal stockholder for the first time, have to forward the Initial Statement of Beneficial Ownership of Securities (Form 3) within ten days of becoming such a liable individual. Transactions conducted by liable individuals on their company’s shares have to be reported to the SEC and to the corresponding national security exchange by sending Form 4 electronically before the end of the second business day following the day on which the transaction is executed. Furthermore, issuers maintaining a corporate website shall publish the transaction on that corporate website, not later than the end of the business day following the reporting to the SEC. The SEC on their part must publish the reporting online on its website not later than by the end of the business day following the receipt of the filing. Hence, transactions conducted by liable individuals become public within 3 trading days (Figure x).

According to Section 21A of the SEA, penalties up to the greater of one million US-Dollar or three times the profits made or losses avoided (civil case) or penalties up to one million US-Dollar and ten years imprisonment (criminal case) can be imposed in response to frequent wilful insider trading violations.

EU Jurisdiction on Management Transactions

In the case of the EU, the first community prohibition of illegal insider trading was enforced in 1989 and has been replaced and standardized Europe-wide by the Directive on Insider Dealing and Market Abuse in 2003, which is one of the main measures in the EU Financial Services Action Plan. This so-called 2003 Directive obliged all Member States to set up minimal regulatory standards on market abuse until October 12th 2004, however, Member States are free to issue their individual standards as far as they exceed the 2003 Directive.

According to Article 1 of the Commission Directive 2004/72/EC not only persons discharging responsibilities within an issuer like senior executives and members of the administrative, management or supervisory bodies are obliged to disclose transactions for their own account, but also third parties in close relationship with them, like husband and wife, children, civil registered partner, who have shared the same household as that person for at least one year on the date of the transaction concerned, but also any legal person whose managerial responsibility are discharged of a person mentioned above, fall into the category of liable individuals.

The substance of the reporting obligation consists of transactions in equity securities, unit trusts, money markets instruments, financial derivatives and any other instrument trading on a regulated market or for which a request to trade has been made. The respective notification shall be made within five working days of the transaction date via the Regulated Information Service to the competent authority of that Member State. Member States are allowed to decide whether such notifications are due only if the total amount of transaction of one single liable individual has reached EUR 5,000 within a given calendar year or whether such notifications are due independent of the amount of transaction.

In case of infringement of the Directives, the Member States have the discretion to decide on the amount of fines and the types of administrative measure applicable in the respective case. Since the Directive does not define appropriate sanctions, there is quite a divergence in the severity of imposed sanctions among Member States.

German Jurisdiction on Management Transactions

The first German insider trading law was implemented as late as 1994 (Bris, 2005). According to Article 14 of the German Securities Trading Act (Wertpapierhandelsgesetz, WpHG) it is prohibited to exploit and transmit insider information. The term insider information is defined as any information of concrete nature not publicly known and that, if published, would likely have a significant effect on the stock price of the respective company. The reporting requirement of management came into effect on July 1st 2002, as a result of the fourth Financial Markets Development Act (Finanzmarktförderungsgesetz). The German legislation on legal insider trading is congruent with the standards set up by the EU (see section 3.2).

Article 15a para. 1 of the WpHG obliges persons discharging managerial responsibilities within an issuer and third parties in close relationship to disclose transactions for their own account. In congruence with the European regulation, liable individuals are required to report the undertaken management transactions to the issuer and Federal Financial Supervisory Authority (Bundesanstalt für Finanzdienstleistungsaufsicht, BaFin) within five business days, as soon as they exceed an aggregate amount of EUR 5,000 within a given calendar year. After receipt of the liable individual’s notification, the issuer is required to publish the undertaken management transaction without delay on its website. Furthermore, the issuer has to forward proof of publication to BaFin. The publications must contain the name of the liable individual, the reason for disclosure, the name of the corresponding issuer, the description of the financial instrument, the ISIN, the nature of the transaction (purchase or sale), the execution date, the location of the transaction as well as the currency, the price and the volume of the transaction. Based on article 15a para. 5 WpHG in conjunction with article 13 Securities, Trading, Reporting and Insider List Regulation (Wertpapierhandelsanzeige- und Insiderverzeichnisverordnung, WpAIV), BaFin publishes all management transactions on its own website for the time period of one year. Consequently, management transactions exceeding the threshold value become public within five trading days (Figure x).

In the case of a wilful or grossly negligent breach of duty in regard to the notification (corporate insider) or publication (issuer) of management transactions, one may be punished with a fine up to EUR 100,000. According to article 39 para. 2(2d) and article 39 para. 2(5b) WpHG, a breach of duty has been committed if a notification (corporate insider) or a publication (issuer) has not been made correctly, in full, on time, as prescribed or has not been made at all.

Austrian Jurisdiction on Management Transactions

In the case of Austria, legal insider trading is regulated in the Austrian Stock Exchange Act (Boersegesetz, BoerseG) and the current regulation on disclosure on management transactions was enforced on May 2nd 2005. The Austrian legislation on legal insider trading follows, except for a few specific Austrian regulations, the minimal standards set up by the EU (see section 3.2).

In congruence with the European regulation, persons discharging managerial responsibilities within an issuer listed on the Austrian regulated market and third parties in close relationship to them are obliged to disclose transactions for their own account. The liable individual is required to report the undertaken management transactions to the Austrian Financial Market Authority (FMA) within five working days, from the moment when the threshold value of EUR 5,000 is collectively exceeded within a calendar year. The notification needs to include the name of the liable individual, the reason for disclosure, the name of the corresponding issuer, the description of the financial instrument, the nature of the transaction (purchase or sale), the execution date, the location of the transaction as well as the price and the volume of the transaction.Additionally, the liable individual is obliged to either agree to the online publication of the respective management transaction by the FMA and/or to make the transaction publicly available promptly through at least one of the electronic information dispersal systems like Reuters, Bloomberg or Dow Jones Newswire. According to the FMA, the public disclosure on their website is done within the day of receipt or on the next business day at the latest. Consequently, management transactions exceeding the threshold value become public within five trading days (Figure x).

Corporate insiders violating their obligations according to article 48 para. 4 of the BoerseG, commit a regulatory offence and may be sanctioned with a monetary fine of up to EUR 30,000, as far as the actions are not elements of a criminal offence of juridical competence.

In contrast, the UK applies even stricter rules than stipulated by the EU. According to the Model Code, directors’ dealings, as they are called in the UK, are prohibited for the time period of 60 days immediately preceding the announcement of annual and semi-annual results.

Swiss Jurisdiction on Management Transactions

In July 1988, article 161 was added to the Swiss Criminal Code (SCC) in regard to regulate insider trading in Switzerland. This provision prohibited knowledge exploitation of confidential facts that could have an influence on the market price. Paragraph 3 of the mentioned article defined only three situations which could have an influence on the market price, namely:

the imminent issuance of new participation rights

the merger of companies and

any similar fact of comparable importance

Under this loose wording, management transactions in the run-up of price-sensitive corporate news (e.g. financial reports, profit warnings) did not fall under the criminal offence of illegal insider trading. Consequently, convictions for illegal insider trading have been seldom and paragraph 3 of article 161 of the SCC was debilitating for the effective combat against illegal insider trading. Only with the repeal of paragraph 3 of article 161 of the SCC on October 1st 2008, an end was put to this strong limitation such that the ban on the exploitation of confidential facts is expanded to all stock price-relevant elements of a crime. Important to mention, not every misuse of confidential knowledge is criminal, it depends primarily on whether the confidential fact is capable of significantly influencing the price of the relevant security in a foreseeable manner. Hence, the penalty norm for insider trading was fundamentally extended in its scope of application.

The first disclosure on management transactions (DMT) was enforced on July 1st 2005, instigated by the Swiss Exchange (SIX). Since then, all companies listed on the SIX are obliged to disclose transactions conducted by the members of the board of directors and the senior management. Interestingly, according to the DMT only the top management is obliged to disclose transactions of their company’s shares, whereas the SCC defines a much broader range of subjects who could potentially exploit insider information. The substance of the reporting obligation consists of transactions in equity securities, conversion and share acquisition rights and financial instruments of any type whose price is materially dependent on the equity, conversion, purchase, subscription or sale rights. In contrast, corporate bonds do not fall in the category where a reporting is required. The liable individual must report his transaction to the issuer no later than on the second trading day following the execution of the transaction. Furthermore, the issuer is obligated to report all transactions where the threshold value of CHF 100,000 is collectively exceeded during a given calendar month by a liable individual. In this case, the fact must be reported via the electronic SIX Exchange Regulation Reporting Platform within two trading days, calculated from the date on which the issuer was informed about the exceeding of the threshold value. The SIX publishes the information on its website within the day of receipt, containing the function of the liable person (executive member of the board of directors / member of senior management or non-executive member of the board of directors), transaction date, type of transaction (purchase or sale), type of security, number of traded securities, price paid or received per security, total trade volume as well as the International Securities Identification Number (ISIN). Consequently, management transactions exceeding the threshold value become public within four trading days at the latest (Figure x). If the threshold value is not exceeded during a given calendar month, the issuer is obligated to make a so-called omnibus notification to the SIX, no later than four trading days following the end of the relevant month. This information is only published on a voluntarily basis.

Issuers, who fail to provide complete and contextually correct notifications of management transactions within the given time frame, may be sanctioned by the SIX Exchange Regulation. According to article 61 of the LR, sanctions may be imposed in the form of a reprimand, fine of up to one million Swiss francs in the case of negligence or 10 million in the case of wrongful intent, suspension of trading, delisting, exclusion from further listings or withdrawal of recognition. In contrast, corporate insiders violating article 161 SCC will be fined or punished with a prison sentence of up to three years.

Data

This empirical analysis covers reported insider transactions executed on the Swiss, German and Austrian stock market between the time period of July 1st 2005 and December 31st 2009. For the Swiss case, the published notifications on management transactions from the SIX and the Finanz und Wirtschaft website were taken as database. The German data set was entirely provided by BaFin and the Austrian was publicly accessible on the FMA website. In a first step, transactions are eliminated for companies which are no longer part of the Swiss Performance Index (for Switzerland), CDAX (for Germany) or Wiener Börse Index (for Austria). Furthermore, transactions where no daily stock prices are available for the estimation and the event window are excluded from the sample. In a last step, transactions of investment companies for whose stocks the market making is done by persons subject to reporting requirement as well as purchase transactions in connection with bonus payments or purchase rights are not taken into the data sample. After these modifications, the basic data sample for the investigation period is represented by a total of 4,468 management transactions for Switzerland, 10,269 transactions for Germany and 1,960 transactions for Austria, respectively (Panel A, C and E of Table X). The same table shows that neither of the three investigated markets has a balanced purchase/sale ratio. In the case of the Swiss stock market, 59 per cent are sale transactions, whereas the German and the Austrian stock market show a striking surplus of purchase transactions, namely 66 percent in the German and 81 per cent in the Austrian case (Panel A, C and E of Table X). A different picture is recognisable when it comes to transactions with stock options. Germany shows an almost equalized ratio, whereas the Swiss case is represented by 84 per cent sale transactions what is likely driven by the maturity of the options and not by insiders’ motives. In contrast, the Austrian case shows slightly more purchase transactions, but at the same the total number of transactions with options is marginal (Panel A, C and E of Table X). Hence, for the further investigation of legal insider trading, transactions in stock options are excluded and a new sub-sample, consisting of insider transactions in stocks only, is built. This new sub-sample is henceforward used as starting point for all subsequent analysis.

In order not to distort the cumulative returns, multiple transactions of one company on a particular day are regarded as one single transaction and accordingly integrated into the sample (Panel B, D and F of Table X).

The academic literature is not unison about the effects of insider purchase and sale transactions. In order to find the corresponding answers for the three stock markets, the first sub-sample aims to examine whether insiders were able to earn abnormal return with their executed purchase and sale transactions.

In order to investigate whether the Small Firm Effect of Fama and French (1992) is attributable to insider transactions on the three stock markets, the effect of the firm size is analyzed in a second sub-sample. The numbers of companies having reported insider transactions in each of the three stock markets are classified in three sub-samples, according to their market capitalisation. For this purpose, the SPI sub-indices, namely the SPI Large, the SPI Mid and the SPI Small, are used as reference for the classification. The SPI Large includes the 20 largest companies (9 per cent), the SPI Mid comprises the successive 80 companies (36 per cent), whereas the SPI Small captures the remaining 121 companies (55 per cent) listed on the SPI. The CDAX represents the benchmark index for the German stock market and the respective companies are classified into Small, Medium and Large by their market capitalisation, according to the calculated percentage rate of the SPI. For the Austrian stock market, the Wiener Börse Index is taken as benchmark for the categorisation by market capitalisation, even though the MSCI Austria represents the total return benchmark index for the successive event study calculations.

In a further step, the trading volume effect on abnormal returns of insider is examined. The total transaction values in Swiss francs are divided into five categories, respectively sub-samples (see Panel B of Table X (evt. Appendix) for exact threshold values). According to the Panel B’s (Appendix Table X, Y and Z), only seven per cent (Switzerland), four per cent (Germany) and six per cent (Austria) of all purchase transactions, respectively twelve per cent (Switzerland), 13 per cent (Germany) and 18 per cent (Austria) of all sale transactions, are above CHF 2 million. This could be an indication for the Stealth Trading Hypothesis of Barclay and Warner (1993), saying that insider tend to strategically disguise their high-volume trades by splitting them up into several smaller trades in order to avoid a strong signalling effect.

In order to investigate whether insiders earn abnormal returns by clustering their management transactions, two additional sub-samples are built. The first sub-sample comprises only transaction dates when the same or different insiders of a specific company execute multiple transactions on the same trading day. Consequently, the second sub-sample includes all remaining transaction dates.

Finally, to analyze whether the mimicry of management transactions can be regarded as a successful strategy for outsiders, the above mentioned sub-samples are tested for significance with a time lag of four (Switzerland) and five days(Germany and Austria). Since management transactions below CHF 100,000 are only published on a monthly basis, if at all, solely management transactions above the disclosure threshold are taken into the investigation of the Swiss stock market. In the cases of Germany and Austria, the threshold is set at EUR 5,000. For the same reason, all transactions below this amount are not part of the investigation of the mimicking strategy.

In a second part, the investigation is separated into two different time spans. The first investigation period ranges from July 1st 2005 to June 30th 2007 and represents the non-financial crisis time span. In contrast, the investigation period from July 1st 2007 to July 10th 2009 covers the financial crisis. The differentiation between periods of economical upturn and downturn is aimed to find out whether corporate insiders change their behaviour depending on the state of the economy.  or how does the market efficiency is influenced! The proceeding of the analysis remains exactly the same as described above. Evt. Noch etwas schreiben, dass Tables in appendix angeschaut werden können (die müssen wohl auch eingefügt werden)

Methodology

The standard event study, originally described by Fama, Fisher, Jensen and Roll (1969), is applied in order to determine whether insiders can generate abnormal returns and whether

outsiders are able to earn abnormal returns by mimicking the publicly available information about management transactions.

As the timeline of the event study depicted in Figure 2 shows, the event date is defined as

τ = 0 and represents the execution date of the management transaction. The event window ranges from τ = -30 to τ = 30 (τ = T1 + 1 to τ = T2) and thus comprises 61 trading days, whereas the 250 trading days before the event window are defined as the estimation window, reaching from τ = -280 to τ = -31 (τ = T0 +1 to τ = T1). T0 accordingly takes the value of -281.

In order to calculate the abnormal returns of the management transactions, the market model is employed. The daily abnormal return using the market model is defined as:

(1)

Where ARit represents the abnormal return for security i at time t, Rit the daily return on security i at time t and Rmt the daily return on the corresponding market index at time t. The parameters and are obtained by using an ordinary least squares (OLS) regression for the daily returns on security i on the market index returns over the estimation period. Daily return data are calculated as Rit = ln(RIit) – ln(RIi(t-1)) where RIit is the return index data point for the individual security and the market index as reported by Datastream.

The daily abnormal returns are then aggregated over time and across transactions yielding CARi(τ1,τ2), defined as the cumulative abnormal return from τ1 to τ2 where T1 < τ1 ≤ τ2 ≤ T2 :

(2)

The average abnormal return AAR(τ1,τ2) is defined as the sum over all abnormal returns ARi divided by the total number of transactions N :

(3)

In a final step, the cumulative average abnormal return CAARi(τ1,τ2) is defined as the sum of the average abnormal returns from τ1 to τ2 :

(4)

This paper applies two statistical tests in order to test the significance of the respective abnormal returns, namely the standardized t-test according to Mikkelson and Partch (1988) and the non-parametric rank test according to Corrado (1989). These tests in conjunction enable answering the question whether insiders are able to obtain abnormal returns and whether the mimicking strategy is economically worthwhile.

The standardized t-test respects the non-stationarity of daily stock variances. It is assumed that abnormal returns are multivariate and uncorrelated across securities. The corresponding null hypothesis states that the mean excess return from τ1 to τ2 (event window) is equal to zero, where T1 < τ1 ≤ τ2 ≤ T2 is tested with the following statistic J1 :

(5)

where N is the number of transactions. denominates the variance of the sum of the ARit series over the time period from τ1 to τ2 (event window). The variance is defined

as:

(6)

The factor represents the mean squared error of the market model regression for security i, T is the number of days in the interval and equal to τ2 - τ1 + 1 , A is the number of returns in the estimation period, is the mean market index return over the event period and is the mean market index return over the estimation period. This procedure demands the assumption that the variance of the daily stock returns during the event window is proportional to the variance over the estimation window and that the standardization factor is similar across securities (Zingg, Lang and Wyttenbach, 2007).

The non-parametric rank test is free of specific assumptions concerning the distribution of returns (Campbell, Lo and MacKinlay, 1997). The test statistic for the null hypothesis of zero mean abnormal return is defined as:

(7)

The standard deviation S(K) is computed by using all abnormal returns in the estimation and event window:

(8)

where

(9)

N is the number of transactions, Kit is the rank of k abnormal returns observed in the estimation and event period and represents the expected mean rank.

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: