Investing Like an Insider: An Event Study Exploring the Possibilities of Positive Return for Outside Investors Following an Insider's Behavior

(1)

DEPARTMENT OF ECONOMICS Uppsala University

Economics C: Thesis Work Author: Anders Carnland Supervisor: Lars Forsberg Spring 2019/VT2019

Investing Like an Insider

An Event Study Exploring the Possibilities of Positive Return for

Outside Investors Following an Insider’s Behavior

(2)

ABSTRACT

This study aims to investigate if an outside investor can gain positive return from investing in company stocks on the Swedish stock market following published announcements of insider stock purchases done through the Swedish financial regulatory authority Finansinspektionen’s public insider transaction registry. Studying a total of 5 966 announced stock purchases during the period 2014 – 2018, the study finds significant positive abnormal return over all studied time periods following the announcement date, regardless of differences in company size. Highest return was found in smaller companies, at the cost of accepting a higher degree of risk. Despite significant results showing informational value of the announced purchases, economic gain from following insider behavior could be inhibited by the cost of investment and would require the outside investor to pick the right stock, which could prove difficult.

Keywords: Insider trading, outside investor, event study, abnormal return, purchase, following insider behavior, Swedish stock market, announcement

Acknowledgements

I would like to give thanks to my supervisor Lars Forsberg for valuable insight and help in the process of writing my thesis this semester.

(3)

1. Introduction

Insider trading has long been an area of interest and of high contention. As a highly regulated area, security trading by an insider – defined as a person with managerial responsibilities (Finansinspektionen, 2019A) – continues to be a well-known subject to the general public as incidents of insider trading among publicly traded companies continue to be reported by media year after year. The attention in news media often revolves around financial crime, in many cases as a result of insiders’ failure to follow necessary regulation or insiders using non-public information for personal economic gain.

Insider trading is however not an illegal act. An important distinction to make is the difference between illegal and legal insider trading. The examples above show instances of illegal insider trading, where regulation has not been adequately followed by the insider, and information unavailable to the public is exploited by the insider to make preferred investment decisions. While insider trading is regulated and highly impacted by law, it is entire possible for an insider to act legally. Legal insider trading, as opposed to illegal, means that the insider complies with regulation, follows the law when investing in its own company, and does not take advantage of inside information that outside investors would not be able to access if it could not otherwise be obtained publicly.

Trading securities on the Swedish stock market requires that insider transactions are reported to the Swedish financial regulatory authority Finansinspektionen, who register and publish insider transaction information, making the information available to the public.

Outside investors are consequently given access to insider transaction information, contributing to stock market transparency and opening up the possibility for an outsider to use the information in their own investment decisions. An outside investor can therefore, in presence of proper regulation, base their own investment behavior on the behavior of insiders in publicly traded companies.

The purpose of this study is to provide a recent analysis of how insider stock transactions affect abnormal return – defined as the price development of the stock surrounding the transaction (the event window) minus the expected return from the stock if the event would not have taken place.

(5)

2

In other words, if stock prices following insider transactions are expected to diverge from the average return of the market, and if it therefore it possible for outside investors to gain abnormal returns by “following” an insider – mimicking insider behavior as an outsider.

The area is of particular interest as mutual funds exists today that invest based on insider activity among Swedish companies (Insiderfonder, 2019), and any option to overperform the market always is desirable for an investor.

Aiming to give as recent of a view of the subject as possible, the period of study will lead up to (and include) insider transactions performed for the year 2018. To limit the study, transactions will only include purchases, but not sales, of stocks for publicly listed companies on the Swedish stock market Nasdaq OMX Stockholm, starting at 2014.

To perform the analysis of how insider transactions relate to abnormal returns, a standard event study is performed, following MacKinlay’s (1997) framework for performing a short-term event study. The event study aims to test the hypothesis that insider stock purchases lead to positive abnormal return following public announcements of insider transactions. Abnormal return following insider purchase transactions on the Nasdaq OMX Stockholm stock exchange is measured after a number of days following the publication date in Finansinspektionen’s registry of insider transactions.

Findings of the study show highly significant abnormal returns following an insider’s public announcement of stock purchase through Finansinspektionen, ranging from the day after the announcement up to 20 days after. Return are however shown to be moderate, leaving the investor to have to take on risk in smaller sized companies to achieve higher returns – laying doubts on the real world practicality of following insider behavior, as finding the right stock could prove difficult, despite the announced purchases’ ability to act as a valueable signal for outside investors.

(6)

3

The remainder of this thesis will be structured as follows. In the second section (2), a brief background of insider trading and regulation will introduce the subject of insider trading, focusing on Swedish regulation. Following the introductory background, an overview of previous research and theory relevant to the subject of insider trading and returns will be given in the third section (3), leading into a definition of the hypothesis this thesis intends to test. The fourth section (4) presents the stock price and insider transaction data used to perform the event study of insider trading on the Swedish stock market. The method in which the event study is performed is then presented in more depth in the fifth section (5), from which the results are presented in the sixth section (6), summarized into a concluding section (7) where the findings of the study are reviewed.

2. Insider Trading Regulation

Found in Bhattacharya and Daouk’s (2002) survey of the worldwide presence of insider trading legislation, 87 out of 102 countries with a stock market had insider trading laws in 1998. Explained as a phenomenon of the 1990s, insider trading laws saw a sharp increase in implementation during the decade, with four of five emerging markets implementing laws and all of the 22 developed countries with stock markets at the time had present laws.

Enforcement of the laws also showed to take place in 82 percent of the developed countries and 25 percent of the emerging markets. Although prosecutions were notably few in the emerging markets, clear improvements in enforcement could be seen from only a decade earlier, when enforcement percentages were at low 23 for developed countries and 7 for emerging markets.

Perhaps more debated in the past, the presence and importance of insider trading laws can be seen in present-day developments of financial markets and regulation. Relevant to the Swedish context of insider trading are recent efforts by the European Union (EU) to expand market manipulation legislation and regulation. Introduced in 2014, regulation No.

596/2014, the Market Abuse Regulation (MAR) of the European Parliament (European Union, 2014), contains extensive restrictions against insider trading, market manipulation and unlawful use of insider information (Finansinspektionen, 2019C) for members of the union to follow.

(7)

4

Swedish regulation is therefore largely based on EU principle, who impose sanctions on unlawful behavior in member nations, which Finansinspektionen is responsible for in identifying on the Swedish market. In 2016, MAR was additionally implemented into Swedish law, expanding present rules and regulations (Finansinspektionen, 2019D).

The restriction against insider trading means that individuals operating on the Swedish stock market, who have access to insider information are forbidden from exploiting the information for personal or other individual’s economic gain by trading securities regarding the possessed information (Finansinspektionen, 2019C). To follow Swedish law, an insider who wishes to trade securities in their own public company stock has to report the transaction to Finansinspektionen at the latest 3 days after the date the transaction was made, if the transaction amount exceeds 5 000 EUR over the calendar year (Finansinspektionen, 2019D). Following a report, the information is automatically published immediately after Finansinspektionen has received notification of the report (Finansinspektionen, 2019E). The market, including outside investors, are then made aware of the transaction, including information of who made the transaction, the type of security, date of transaction, size and price, among other information. The public is thus offered valuable information that the insider otherwise would have as an informational advantage against outside investors, providing efficiency and stability to the financial market, two of Finansinspektionen’s main objectives as a financial supervisory authority.

3. Previous Research and Theoretical Overview

This section will begin with an introduction into the subject of insider transactions and stock price returns, by giving an overview of relevant previous research of the subject.

Following previous research, relevant theory to the area of study will be presented, leading into formulation of the hypothesis this study intends to test.

3.1 Review of Previous Research

Dating back to Smith (1940), early research on insider trading concluded that “…insiders did not make exceptional trading profits” when compared to the returns of indices. Lorie

& Niederhoffer (1968) question the results of the large majority of earlier research on the subject that did not show that insider trading could be particularity profitable.

(8)

5

Using more accurate information on prices and transaction dates, Lorie & Niederhoffer found that, contrary to earlier research, insiders were expected to outperform the market six months following purchase of shares. Additional studies by Jaffe (1974) and Finnerty (1976) concluded that insiders possessed special information, refuting the Efficient Market Hypothesis strong form of market efficiency (described in section 3.2), proving that insiders were able to outperform market returns. Seyhun (1986) finds similar results, showing that insiders could predict future abnormal stock price changes, buying its own company stock before an abnormal rise in price level, and selling before a decline.

Furthermore, Seyhun concluded that outside investors are not able to earn abnormal returns by using publicly available information about insiders’ transaction. In later study, Seyhun (2000) found that it is possible for outside investors to gain a profit by mimicking insiders, but by being prepared to hold the shares up to three months to compensate for transaction costs and accepting a rather high risk of losing money.

3.2 Market Efficiency

In an efficient market, prices should fully reflect all available information. Developed by Fama (1970), the Efficient Market Hypothesis (EMH) describes that security prices reflect all available information and adjust quickly to new information. An efficient market would according to EMH, make it impossible for an individual investor to outperform the market consistently, as the individual would not possess any information about a security that would not already be fully reflected in its price. The individual investor would then not be able to earn a higher return than the market by picking stocks or timing the market, two common examples of investor strategy. Fama divides the hypothesis into three level of market efficiency:

1. Strong form: Security prices reflect all publicly and privately available information.

2. Semi-strong form: Security prices reflect all publicly available information, such as annual reports, announcement of stock splits, historical price levels, etc.

3. Weak form: Security prices only reflect the information of historical price levels.

(9)

6

A market with a strong form of efficiency, where not only public information but private information is reflected upon security prices, the individual investor would not be able to consistently outperform the market as the security would perfectly reflect all available information. Although not likely in reality, no investor or group would have monopolistic access to information in a market with a strong form of efficiency. No insider would then possess information that the market does not, making it impossible for insiders or outside investors to consistently earn a higher return than the market.

The semi-strong form, where public, but not private information is reflected upon security prices, implies that some investors or groups could have monopolistic access to information. An insider could possess information about a company that is not publicly available, making it possible for the insider to earn higher return than the market by utilizing the information when making investment decisions.

In a market with a weak efficiency level where security prices only reflect historical price information, monopolistic access to information could be present, like the semi-strong form. Not only can an insider use private information to earn a higher return than the market by utilizing the information, an outside investor could also mimic insider behavior when making investment decisions as public information is not reflected in the prices. In the weak-efficiency market, public announcements of insider transactions would not be quickly reflected into the price of the security, creating an opportunity for outside investors to earn abnormal return by acting as an insider even after the information of an insider transaction has become public.

3.3 Signaling

First introduced by Michael Spence (1973) in his paper on job market signaling, the act of signaling is described as the action one party takes to send a signal to another party regarding the individuals characteristics. In the case of an individual applying for a job, the applicant could signal the employer with information about his or her attributes, in order to distinguish him- or herself from other applicants. When an informational asymmetry exists, the act of signaling could reduce the informational gap between two parties.

(10)

7

Grossman & Stiglitz (1980) further illustrate informational asymmetries by showing that an informed individual who has obtained information of how a security is expected to show high return is going to act on that knowledge, sending a signal to the uninformed individual. Givoly & Palmon (1985) further connect the area of signaling to insider trading, showing that the informational knowledge of insider transactions taking place produces significant abnormal return, and that outside investors value knowledge of the insider transactions taking place. However, no significant difference in abnormal return was found between insider transactions of large and small size. Another found aspect in their study was that abnormal return following insider transactions were independent of the publication of concurrent news. This supports the signaling effect, as outside investors are shown to value the information provided by insiders, leading outside investors to follow the signals and behavior of insiders. Insider purchase of shares could in this case be perceived as a signal of optimism from within the company. This signal could be valued by outside investors as it likely will be perceived by the outsider as information the public does not otherwise have access to in the same way the insider does.

In a more recent study, Bettis, Vickrey & Vickrey (1997) find results contrary to previous studies of Givoly & Palmon, showing that publicly available insider transactions regarding trades on the New York-, and American Stock Exchange (NYSE, AMEX) of larger size by high-ranking insiders lead to significant abnormal returns. In the study, transactions over a trade volume of 10 000 shares by high-ranking insiders were observed during the period 1995 – 1990. The authors further conclude the possibility for outside investors to gain abnormal returns by mimicking the action of the high-ranking insiders doing large volume trades, even after transaction costs. To gain abnormal returns however, an outsider would have to hold the shares for a considerable time period, showing significant abnormal returns after 26 weeks or more.

3.4 Hypothesis

Relating to previous research and theory, the notion that markets are fully efficient would seem unlikely, as insider transactions historically have shown to generate abnormal returns. If all publicly available information was reflected in security prices, the possibility of achieving abnormal returns by insider trading should not exist.

(11)

8

It could therefore be believed that markets are not fully on the semi-strong form of efficiency but show signs of a weak form. Signals sent by insiders to outside investors through trade of securities have also shown to be information of value to the outside investors. It therefore becomes interesting to further analyze how insider trading affects stock returns. The hypothesis this study aims to test is therefore formulated.

Hypothesis: Published announcements of insider stock purchases lead to positive abnormal returns.

4. Data Description

Data gathered in this study consists of historical stock price data and data on insider transactions for companies listed on the Swedish stock exchange Nasdaq OMX Stockholm (OMXS) for the five years 2014 - 2018. The following section describes the gathered data set and the limitations that have been made to exclude certain data points in order to shape the final data sample. A complete list of included and excluded OMXS companies can be further explored in Appendix 1.

4.1 Stock Price Data

Daily stock price data has been gathered from Thomson Reuters Eikon for all companies listed on Nasdaq OMXS as of April 2019 (Nasdaq, 2019A). This includes daily closing price data of all Large Cap, Mid Cap, and Small Cap listed companies over the time period 2 January 2013 – 29 March 2019. Larger focus has been set on the main market, not including Nasdaq First North or Aktietorget (now Spotlight Stock Market) listed companies, but still gathering price information for companies currently listed on OMXS with previous trading history on any non-OMXS exchange (e.g. Spotlight Stock Market or Nasdaq First North). The data includes daily closing prices dating back to January 2013 into March 2019, in order to adequately be able to perform the event study for the full five- year period of 2014 - 2018. Estimating expected return for certain insider transactions during 2014 and establishing event windows for insider transactions close to end-year 2018 requires some stock price data of 2013 and 2019, motivating the collection of this stock price data (further explained in section 3.2 and 4).

(12)

9

Daily market price data has also been gathered from Thomson Reuters Eikon, including closing price data for the Nasdaq OMX Stockholm PI index (OMXSPI) during the same time period 2 January 2013 – 29 March 2019. The index reflects the price information and development of all company stocks listed on Nasdaq OMXS (Nasdaq, 2019B). As the index includes all Large-, Mid-, and Small Cap listed companies on the Stockholm stock exchange, it should adequately fulfill the aim to closely resemble the companies from which price data has been gathered, which is desirable in the coming event study execution and expected return calculation (further explained in the thesis Method section 5).

4.2 Insider Transaction Data

Insider transaction data has been gathered from the Swedish financial supervisory authority Finansinspektionen PDMR transaction register (Swedish: Insynsregistret) (Finansinspektionen, 2019B). The register contains data on financial transactions performed by individuals with managerial responsibilities and individuals closely associated with them. Initial data collection includes all published insider transaction data for all Nasdaq OMXS listed companies over the five-year time period 2014 – 2018. This includes insider transaction data in all securities requiring reports to Finansinspektionen.

Recent insider transaction data has been gathered from the PDMR transaction register’s public database, available from July 2016 to the present date. Earlier transactions have been gathered from an extensive Excel spreadsheet given archival access to by Finansinspektionen directly through email- correspondence.

Initial insider transaction data gathered from Finansinspektionen consists of a total of 41 468 security transactions before any exclusion of data has been made.

(13)

10 4.3 Sample Selection

For the purpose of the study, only transactions of company public stock have been included. All transactions other than those strictly related to company public stock have been excluded. Exclusions include transactions in the following securities, among others:

1. Options and warrant exercises

2. Company stock- and options- program related transactions 3. Company stock gifts, inheritance and loans

4. Company stock purchase rights (Swedish: Teckningsrätter)

5. Contract payments of company stock related to a new share issue (Swedish: BTA – Betald Tecknad Aktie) or initial public offering

6. Company stock offered as performance awards

Insider transactions have been included for all company stock classes except preferred stock. This includes registered transactions of stock classes not available through trade on a public exchange or on an exchange outside of Sweden. All transactions have been included regardless of trading location, as long as the transaction has been registered through Finansinspektionen. In the case of company trading more than one class publicly on an exchange (e.g. both stock class A and B as for Atlas Copco A, Atlas Copco B) transactions have been divided between the relevant stock class and will be analyzed as separate stocks following the event study.

A clear limitation and choice of this study has been to only include insider purchases, but not insider sales. All insider transactions regarding company public stock sales have thus been excluded. The choice to only focus on purchases is made to put greater focus on how an outside investor can follow an insider’s behavior by buying company stock when information of the purchase is made public and if this action can lead to positive returns after holding on to the shares a number of days. As insider purchases could be viewed as a signal of optimism from within the company, sales could signal the opposite. In an insider sale situation, the outsider would also be required to already own the stock at the time of sale to benefit from insider behavior.

(14)

11

To focus on the optimistic view of trading stocks, and the potential for the outsider to earn a positive return, purchases is chosen as the sole focus of the analysis.

“Insiders have many reasons to sell shares but the main reason to buy shares is to make money” (Lakonishok & Lee, 2001)

A few further limitations have been made of the sample. Taking aid from Lakonishok and Lee’s (2001) sample selection methods in their study of insider activity informativeness, all insider purchase transactions below 100 shares were excluded from the sample, in addition to eliminating duplicate and amended (Swedish: Reviderade) transactions.

Additionally, all purchases with no price information were excluded, as these transactions were likely misregistered as purchases. Furthermore, a number of registered transactions did not pertain to its own company stock and were excluded, as they were likely misregistered by the individual making the registration to Finansinspektionen.

Transactions specified as stock purchases in subsidiaries were also excluded to only focus on the listed parent company stock.

Lastly, in the case that several insider purchases were published the same day in the same company but by different individuals, for example, the transactions were limited to only one per day and company, limiting the number of events to a maximum of one per day.

After excluding all non- stock related insider transactions, limiting the data to purchases, and eliminating the remaining rejected data points, the total number of transactions arrives at 11 425. Further limiting the final data set to one transaction per stock and publishing date brings the total number of insider transactions for the period 2014 – 2018 to 6 443.

Worth mentioning is a possible bias in sample selection that occur due to companies exiting the stock market. In cases where companies are delisted from OMXS, declared bankrupt, or the like, the company stock is not included in the sample. It is possible that insider trading took place in a number of such companies during the surveyed time period 2014 – 2018, which could create a degree of bias in the results as they would likely have performed worse than other companies who survived on the stock market. In this case the possible abnormal return would be slightly upwards biased, as a source of risk in investing is not observed.

(15)

12 4.4 Descriptive Statistics

Sample statistics consists of insider stock purchases in 102 Large Cap listed stocks, 135 Mid Cap stocks, and 94 Small Cap stocks. Presented below are the total number of published insider purchases in the sample, divided into the three categories of company size and year the transaction was published. Large Cap represents the companies with the highest market capitalization and Small Cap the lowest of the OMXS listed companies.

Table 4.1 Number of insider stock purchases reported through Finansinspektionen’s registry during the period 2014 – 2018, with per year statistics. Reported values are after all non-stock related transactions, clusted transactions during the same date have been limited to one transaction only, and other transactions named in section 4.3 have been excluded.

Number of Reported Insider Stock Purchases Number of

Stocks Total 2014 2015 2016 2017 2018

Large Cap 102 2 365 412 479 561 447 466

Mid Cap 135 2 568 399 534 466 509 660

Small Cap 94 1 510 259 280 241 342 388

Total 331 6 443 1 070 1 293 1 268 1 298 1 514

Included in the data presented above are however a number of transactions that do not have available stock price data on the day of purchase announcement (the so-called event day – further described in section 5.2), making analysis of that particular event impossible. This is due to the fact that the transaction date is not necessarily the same as the date of purchase but could in some cases be up to 3 days after the date of purchase as described in section 2. These transactions have therefore been excluded from the final sample in which the hypothesis will be tested. Final sample size of insider purchases thus land at 2 228 Large Cap- transactions, 2 363 Mid Cap, and 1 375 Small Cap, presented in Table 4.2.

(16)

13

Table 4.2 Number of insider stock purchases during the period 2014 – 2018, reported through Finansinspektionen’s registry. Presented number of transactions is the final sample size of data to be used for analysis by performing the event study.

Number of Stocks

Number of Reported Insider Stock Purchases Included in

Final Sample

Large Cap 102 2 228

Mid Cap 135 2 363

Small Cap 94 1 375

Total 331 5 966

Final sample of reported insider stock purchases for OMXS, from which the event study will be performed to test the hypothesis that public announcement of insider purchases lead to positive abnormal returns. The divide into announcements of Large Cap, Mid Cap, and Small Cap stock insider purchases will be used to analyze each category of company size separately.

5. Methodology

In examining the effects of insider transactions on stock prices, the hypothesis will be tested by using a methodical framework suitable for finding an answer to the question of the study. To do this, a standard event study framework will be followed in measuring the short-term effects of insider transactions.

5.1 Short-term Event Study

In this study, MacKinlay’s (1997) short-term event study framework has been chosen to examine the effects of insider transaction events on stock prices. MacKinlay’s framework, which is commonly used in studies on the subject, presents the following method to measure the effect of an event on the price of a company.

(17)

14

1. Define the event of interest and identify the period where stock prices of the companies involved in this event will be examined.

2. Determine selection criteria for including companies in the study (presented in further detail under section 4. Data).

3. Select a suitable normal performance model to estimate expected (normal) return over the event window, using a defined window to estimate expected return.

4. Calculate abnormal return to determine the impact of the event, using the estimation of expected return.

5. Aggregate the individual companies’ abnormal returns to test the null hypothesis.

6. Present empirical results.

5.2 Event Window

In the case of this study, the event of interest is the published announcement of the insider transaction, here focusing solely on an insider’s purchase of its company stock. To focus on the possibility for an outside investor to achieve abnormal returns by following the action of an outside investor when buying stock in order to sell the same stock a number of days after the event day, insider sales are disregarded. The event day here corresponds to the date at which the insider stock purchase was published to the public through Finansinspektionen’s insider transaction registry. The publication date is chosen as the event day, rather than the day the insider purchase occurred to again put greater focus on an outside investor’s possibility of achieving abnormal returns by acting as an insider. In order for an outsider to follow insider behavior, the public information available through Finansinspektionen would be required to reliably make investment decisions in cases where no prior announcement has been made by the company in which insider purchases take place. For this reason, the publication day is best suited as the choice of event day, as it could be viewed as the most reliable source of insider purchase information presented to outside investors.

For the purpose of the study, the effects of insider purchases will be measured up to 20 days following the event day, using the following six different event windows lengths to analyze the abnormal return development after the event day.

(18)

15 1. Event window 1: Event day 𝑇₁ + 1 day 2. Event window 2: Event day 𝑇₁ + 3 days 3. Event window 3: Event day 𝑇₁ + 5 days 4. Event window 4: Event day 𝑇₁ + 10 days 5. Event window 5: Event day 𝑇₁ + 15 days 6. Event window 6: Event day 𝑇₁ + 20 days

The purpose of using a number of event windows of different length are to be able to analyze if abnormal return can be gained to a larger degree if the bought stock is held a certain number of days. It allows for better understanding of an outsider’s best course of action if following insider behavior, and also shows if or when the possible returns start to decline.

⏞

𝐸𝑠𝑡𝑖𝑚𝑎𝑡𝑖𝑜𝑛 𝑤𝑖𝑛𝑑𝑜𝑤

⏞

𝐸𝑣𝑒𝑛𝑡 𝑤𝑖𝑛𝑑𝑜𝑤

⏞

𝑃𝑜𝑠𝑡−𝑒𝑣𝑒𝑛𝑡 𝑤𝑖𝑛𝑑𝑜𝑤

𝑇₀ 𝑇₁ 𝑇₂ 𝑇₃

𝐸𝑣𝑒𝑛𝑡 𝑑𝑎𝑦 (𝑇⏞ ₁)

Figure 5.1 Example timeline of an event study, including explanations of the estimation window, event window, and event day relevant to this study.

A common approach to performing an event study is to let a number of the event windows start prior to the event day in order to capture stock price effects that would occur before the publication date due to the possibility of insider activity information leakage prior to the publication date. However, this study choses to focus on the possibility to achieve abnormal returns following only the publication date, and therefore the event windows stretch from the event day forward. Although this means that possible announcements made by the involved company (such as earnings announcements or announcements of large insider purchases) – that could affect the stock price before the event day – are not considered, the event window is limited to the period after the event day.

(19)

16

In an effort to study the possible effects of purchasing stocks as an outsider when only relying on public insider trading information, the event of interest will be limited to time after the public announcement by Finansinspektionen. Prior information is therefore disregarded for the purpose of this study, despite its possible informational value to outside investors.

5.3 Abnormal Returns

Defined as the actual return of the company’s stock over the event window minus the expected return of the company over the event window, abnormal return for the company i encountering an event on event day t is calculated using equation (1)

𝐴𝑅_𝑖𝑡 = 𝑅_𝑖𝑡− 𝐸(𝑅_𝑖𝑡|𝑋_𝑡) (1)

𝐴𝑅𝑖𝑡: Abnormal return for company i in period t 𝑅_𝑖𝑡: Actual return for company i in period t

𝐸(𝑅_𝑖𝑡|𝑋_𝑡): Expected return for company i in period t, in which case the event did not take place 𝑋_𝑡: Variable determining the expected return in period t

Initially, expected return has to be estimated. By choosing to adopt the market model outlined by MacKinlay (1997), the return of company i in period t is related to the return of the market portfolio, and calculated using equation (2)

𝑅_𝑖𝑡 = 𝛼_𝑖 + 𝛽_𝑖𝑅_𝑚𝑡+ 𝜀_𝑖𝑡 (2)

𝑅𝑖𝑡: Return on security i in period t

𝛼_𝑖: Estimated parameter of the market model – the company stock’s intercept

𝛽_𝑖: Estimated parameter of the market model – measure of volatility of the stock compared to the market 𝑅_𝑚𝑡: Return on market portfolio in period t

𝜀𝑖𝑡: Zero mean disturbance term, residual for security i in period t. Expected to equal zero.

(20)

17

The choice of the market model over other popular models, such as the constant mean return model, is made with support from MacKinlay, as it provides a potential improvement of abnormal return variance by removing any part of the model related to market return variation.

Market portfolio return (𝑅𝑚𝑡) will in the case of the study be represented by the Nasdaq OMX Stockholm PI (OMXSPI) index previously mentioned in the data section (4). The index mirrors the price level of all companies listed on the Swedish stock market Nasdaq OMXS and should therefore be able to give a representative view of the companies in question for the study. As market return is the parameter of interest, OMXSPI closing price data is first calculated into daily market return, which is then used to represent the market portfolio return. Using market return based on a large number of the companies included in the final sample of insider purchases should therefore provide for a solid estimate of expected return.

It is worth mentioning that dividend payments are not taken into account in this study. A potential issue in choosing to not do so is that companies with larger dividend payments will likely experience larger price movements as the dividends are payed out, when compared to companies with smaller dividends. It could however be argued that insiders take dividend payments into accounts when purchasing stocks and hesitate to make purchases right before a dividend payment. Such purchases would otherwise appear underperform when compared to companies with smaller dividends and price movements.

Focusing on the short run, and observing a large number of insider purchases, should however prove to give a solid average of what an outsider can expect in terms of abnormal returns. By not considering dividends payments, but only acting as an insider to potentially make money in the short run, the purpose of the study is followed without taking dividends into account.

Market model parameters alpha (𝛼) and beta (𝛽) are subsequently estimated through the use of an ordinary least square (OLS) regression for company i over an estimation window based on a period prior to the event in question. For the purpose of the study, an estimation

(21)

18

window of 120 working (trading) days will be used, consistent with the length of estimation window mentioned by MacKinlay.

𝛼̂_𝑖 = 𝜇̂_𝑖 − 𝛽̂_𝑖𝜇̂_𝑚 (3)

𝛼̂_𝑖: Estimated alpha parameter for company i

𝜇̂_𝑖: Estimated average return for company i over the estimation window L (𝜇̂_𝑖= ¹

𝐿₁∑^𝑇_𝑡=𝑇¹ ₀₊₁𝑅_𝑖𝑡) 𝛽̂_𝑖: Estimated beta parameter for company i

𝜇̂𝑚: Estimated average market return over the estimation window L (𝜇̂𝑚= ¹

𝐿₁∑𝑇₁ 𝑅𝑚𝑡 𝑡=𝑇₀+1 )

𝛽̂_𝑖 = ∑^𝑇_𝑡=𝑇¹ (𝑅_𝑖𝑡− 𝜇̂_𝑖)

0+1 (𝑅_𝑚𝑡− 𝜇̂_𝑚)

∑^𝑇_𝑡=𝑇¹ (𝑅_𝑚𝑡− 𝜇̂_𝑚)²

0+1

(4)

After OLS estimation of 𝛼 and 𝛽 for each observed company i, expected return can be established using the market model equation (2), and abnormal return calculated by following with the abnormal return equation (5)

𝐴𝑅̂_𝑖𝑡 = 𝑅_𝑖𝑡 − 𝛼̂_𝑖− 𝛽̂_𝑖𝑅_𝑚𝑡 (5)

𝐴𝑅̂_𝑖𝑡: Abnormal return for company i in period t

(22)

19

Abnormal return for the individual company i is then aggregated by calculating the average abnormal return for the sample over N number of experienced events by the company

𝐴𝑅̅̅̅̅_𝑡 = 1

𝑁∑ 𝐴𝑅̂_𝑖𝑡

𝑁

𝑖=1

(6)

𝐴𝑅̅̅̅̅_𝑡: Aggregated abnormal return in period t

Lastly, the average aggregated abnormal return is cumulated for each company over each chosen event window using the method in equation (7)

𝐶𝐴𝑅̅̅̅̅̅̅(𝑡₁, 𝑡₂) = ∑ 𝐴𝑅̅̅̅̅_𝑡

𝑡₂

𝑡=𝑡₁

(7)

𝐶𝐴𝑅̅̅̅̅̅̅(𝑡₁, 𝑡₂): Cumulative abnormal return over time periods (event windows) 𝑡₁, 𝑡₂

5.4 Hypothesis Testing

Having established abnormal returns, MacKinlay (1997) seeks to draw inferences from the cumulative abnormal returns by testing the null hypothesis that abnormal returns are zero.

Following MacKinlay’s framework, a test of significance is performed using equation (8)

𝜃₁ = 𝐶𝐴𝑅̅̅̅̅̅̅(𝑡₁, 𝑡₂)

(𝑣𝑎𝑟(𝐶𝐴𝑅̅̅̅̅̅̅(𝑡₁, 𝑡₂)))¹^⁄²~𝑁(0,1) (8)

(23)

20

Aiming to test the hypothesis that published announcements of insider stock purchases lead to positive abnormal returns, one-tailed tests are performed separately over Large Cap, Mid Cap, and Small Cap listed company stocks, where companies belonging to each size group is tested as an aggregate. This means analysis will be able to be done for each category of company size separately, with separate levels of significance.

Before performing the tests using equation (8), the variance of the average abnormal return has to be calculated using equation (10), followed by the variance of the cumulative abnormal return using equation (9)

𝑣𝑎𝑟(𝐶𝐴𝑅̅̅̅̅̅̅(𝑡₁, 𝑡₂)) = ∑ 𝑣𝑎𝑟(𝐴𝑅̅̅̅̅_𝑡)

𝑡₂

𝑡=𝑡₁

(9)

𝑣𝑎𝑟(𝐴𝑅̅̅̅̅_𝑡) = 1

𝑁²∑ 𝜎_𝜀²_𝑖

𝑁

𝑖=1

(10)

In performing the regressions to estimate 𝛼 and 𝛽, calculating expected and abnormal return, as well testing the null hypothesis that cumulative abnormal returns are zero, support is taken from Princeton University (2008A, 2008B) in performing an event study using Stata, helpful in the situation of this study where the sample of events is fairly large.

(24)

21

6. Results

In the following section, results of the study are presented and analyzed. Table 6.1 introduces the found cumulative abnormal return (CAR) for respective Large Cap, Mid Cap, and Small Cap stocks over the six events window lengths following an event. CAR is presented as an aggregate of each company size category to allow analysis of potential differences in company size, of particular interest on the Swedish stock market OMXS as Nasdaq divides listed companies into clear segments of company size (Large Cap, Mid Cap, Small Cap).

The results find that CAR over all the established event windows are significantly different from zero on at least a 5 % significance level, showing that only Mid Cap CAR over the longest lasting event windows of 15 and 20 days is at a 5 % level of significant positive abnormal return, while all other CAR results are significant on a 1 % level.

The highest CAR following insider purchases can be observed in Small Cap listed company stock after 20 days following the event day, where a 2.4 % return can be expected above the market return. The lowest CAR can be observed in in Large Cap stock, 1 day following the event day, where a 0.15 % return is shown above market return. In general, the smallest companies belonging to Small Cap show the highest CAR, exceeding 1 % over all event windows. The largest companies (Large Cap) show the lowest CAR on average over the event windows, although not diverging much from the CAR of Mid Cap stocks. Lowest return as compared to the market can however ne seem in Mid Cap stocks over the 15- and 20-day event windows, apparent when observing the significance level associated with the return.

Observing the standard error shows additionally that returns in Small Cap stocks can be expected to vary more that Large Cap and Mid Cap. Mid Cap stocks once again place between observed value of Small and Large Cap, although at a level comparable to Small Cap 1 and 3 day event windows following 15 and 20 days after purchases in Mid Cap stocks. Where possibilities for higher return is present, so is the risk of experiencing a higher degree of variation in the returns. This is especially true concerning Small Cap stocks over longer event windows where investors have to accept risk to get higher return.

(25)

22

Table 6.1 Cumulative Abnormal Return (CAR) for stock purchases on the Swedish stock market Nasdaq OMX, following the announcement of the insider purchase from Finansinspektionen. CAR is presented as the percentage return over six event windows with different lengths, starting at 1 day after the announcement, up to 20 days after. Presented results are divided into three different levels of company size based on market capitalization, Large Cap representing the largest companies and Small Cap representing the smallest. Also presented are the standard errors and t-value statistic for all presented CAR- percentages.

***: Statistically significant on a 1 % level **: Statistically significant on a 5 % level *: Statistically significant on a 10 % level

As the test performed is a one-tailed (right tail) significance test, the null hypothesis is rejected on a 1 % percent level if t ≥ 2.326, on a 5 % level if t ≥ 1.645, or on a 10 % level if t ≥ 1.282. In the results above, all presented CAR is significant on a 5 % level or below, with the majority of CAR over the event windows are significant at a 1 % level.

Days After Event

(Numbered Event Window in Parenthesis) Number

of Stocks

Total Events

(1)

1 day

(2)

3 days

(3)

5 days

(4)

10 days

(5)

15 days

(6)

20 days

Large Cap 102 2 228

CAR (%) Standard Error t-value

0.150***

0.043 3.47

0.212***

0.058 3.62

0.281***

0.068 4.14

0.437***

0.088 4.94

0.380***

0.106 3.58

0.464***

0.120 3.85

Mid Cap 135 2 363

0.401***

0.070 5.73

0.513***

0.0916 5.60

0.497***

0.107 4.64

0.506***

0.134 3.77

0.312**

0.155 2.01

0.316**

0.181 1.74

Small Cap 94 1 375

1.202***

0.141 8.55

1.711***

0.196 8.72

1.911***

0.234 8.15

2.068***

0.265 7.80

2.289***

0.306 7.48

2.400***

0.371 6.47

(26)

23

Generally, the results show signs of a weak form of market efficiency, where public information is not immediately reflected in the price of securities. Although the abnormal return is not particularly high in some cases and the risk of investing can be quite high, especially if investing in the highest yielding stock segments, abnormal return can be achieved up to 20 days after insider purchases announcements are made public. The outside investor could then conceivably use the signal from the purchasing insider to purchase stocks of their own, expecting to gain positive returns. The difficulty then lies in picking the right stock to buy, which could prove difficult, as well as costly if investing in smaller companies with higher risk. Lakonishok & Lee (2001), in their study of insider trade informativeness, emphasize the notion that implementing investment strategies after insider behavior can be difficult. Even if some insider trading information could show informational value, acting on the right insider trade in the right company could prove a challenge. Lakonishok & Lee (2001) find that higher returns can be expected in companies of smaller size often subject to higher risk, which is consistent with the results found in this study. Investing to expect high returns by following an insider could prove both difficult and expensive.

Although Givoly & Palmon (1985) show no clear connection between transaction size and abnormal return, a possible strategy could be to invest when insiders buy a large number of shares, or a person of particular interest in the company (CEOs, board members, etc.) buys stock of considerable value, if this information is believed to be more informative than other insider trades. As a signal of optimism, a large purchase by a CEO could be believed to be strong, but the potential for abnormal return in the short run is arguable.

Bettis, Vickrey & Vickrey (1997) found that outsiders could follow high-ranking insiders making large-volume trades, but that the holding period after a purchase would have to be above 13 weeks to see significant abnormal returns to compensate for transaction costs. A possibility in considering investment strategy could then be to act on the public announcement of large-volume trades by CEOs, board members, or other executives to see significant return in the slightly longer run. Results of this study show a slight upwards trend in CAR over the event windows, opening up a curiosity of the longer- run abnormal returns following larger-sized insider purchase announcements. This also supports Seyhun’s (2000) results, that conclude outside investors could gain abnormal returns acting

(27)

24

as an insider by holding stocks up to three months to compensate for transaction costs. Not only does the outsider have to accept a high degree of risk taking to act in the way of an insider, the cost of investment would also have to be considered as explained by Seyhun.

If Seyhun’s results are to be believed, the potential for positive return would be hindered by possible transaction costs. If the cost of investment is high, the return from investing after insider behavior in OMXS- stocks would decrease. This would mean that the gain in CAR would be consumed to a degree by costs, possibly strangling the potential for positive economic gain, especially in Large Cap stocks where the CAR already is on the lower end.

7. Conclusion

The study shows potential for positive abnormal return by following an insider’s behavior trading on the main Swedish stock market. The cumulative abnormal return following announcements of insider purchases in Large Cap, Mid Cap, and Small Cap company stock through Finansinspektionen’s public registry show high significance over all three categories of company stock size (Large to Small Cap) up to 20 days after the announcement. Lowest significant return is found after 15 and 20 days following announcements in Mid Cap stocks. Highest significant return is found in Small Cap stocks, but at a cost of accepting higher risk in terms of variation in abnormal returns.

Aiming to answer the hypothesis of whether published announcements of insider stock purchases lead to positive abnormal returns, the results show clear signs that the announcements in fact do so, with signs of a weak form of market efficiency as the outside investor is able to follow insider behavior to gain positive returns. Prices of OMXS company stocks do not immediately reflect the announced information of insider purchases, refuting the semi-strong form of efficiency. The announcement could then in possibility be used by an outsider to make preferable investment choices, aiming to act as an insider does.

Despite high significance in the found results, one important aspect to consider is the cost of investment. As the cumulative abnormal return is rather moderate over a number of the event windows, transaction costs could diminish the possibility of positive return.

(28)

25

Future study could potentially gain from a further focus on the value (volume) of traded stock, as well as who the purchasing insider is. As previous research by Bettis, Vickrey &

Vickrey (1997) showed potential for abnormal return (after transaction costs) as an outside investor investing after insider trades of large volume made by high-interest individuals in the company in question, it could be interesting to consider larger trade volume transactions. It could also prove interesting to explore a time period above 20 days, as the found cumulative return of the study could be believed to increase further over a longer time period, which previous study show in regard to large-volume trades. Considering future research, a stricter sample selection could possibly improve the reliability of the results, as some clustering is present where event windows in some of the observed companies included in this study overlap, potentially affecting the abnormal returns following other events.

In final conclusion, outside investors can gain positive abnormal return acting on announced insider purchases on the Swedish stock market, but with limited potential gain considering transaction costs and the risk of investment.

(29)

26

8. References

Bettis, C., Vickrey, D., & Vickrey, D. W. (1997). Mimickers of Corporate Insiders Who Make Large-Volume Trades. Financial Analysts Journal. 53(5), 57-66.

Bhattarcharya, U., & Daouk, H. (2002). The World Price of Insider Trading. The Journal of Finance, 57(1), 75-108.

European Union (2014). Regulation (EU) No 596/2014 of the European Parliament and the Council. Brussels: Official Journal of the European Union.

Fama, E. F. (1970). Efficient Capital Market: A Review of Theory and Empirical Work.

The Journal of Finance. 25, 383–417.

Finansinspektionen (2019A). New rules for insider reporting and insider lists. Accessed April 30, 2019, from Finansinspektionen. Available at: https://www.fi.se/en/published/

news/2016/new-rules-for-insider-reporting-and-insider-lists/

Finansinspektionen (2019B). PDMR transactions register. Accessed 28 May, 2019, from Finansinspektionen. Available at: https://www.fi.se/en/our-registers/pdmr-transactions/

Finansinspektionen (2019C). Marknadsmissbruk. Accessed June 5, 2019, from

Finansinspektionen. Available at: https://www.fi.se/sv/marknad/om-marknadsmissbruk/

Finansinspektionen (2019D). New rules for insider reporting and insider lists. Accessed June 5, 2019, from Finansinspektionen. Available at: https://www.fi.se/en/published/

news/2016/new-rules-for-insider-reporting-and-insider-lists/

Finansinspektionen (2019E). MAR – reporting PDMR transactions. Accessed June 5, 2019, from Finansinspektionen. Available at: https://www.fi.se/en/markets/reporting/

mar-reporting-pdmr-transactions/

Givoly, D. & Palmon, D. (1985). Insider Trading and the Exploitation of Inside Information: Some Empirical Evidence. Journal of Business, 58(1), 69-87.

Grossman, S. J. & Stiglitz, J. E. (1980). On the Impossibility of Informationally Efficient Markets. The American Economic Review, 70(3), 393-408.

Insiderfonder (2019). Insider-fonder AB – Nytänkande – med rötter i forskarvärlden.

Accessed April 30, 2019, from Insiderfonder. Available from: https://insiderfonder.se/

om-oss/insider-fonder-aktiebolag/

Lakonishok, J., & Lee, I. (2001). Are Insider Trades Informative? The Review of Financial Studies, 14(1), 83, 109.

(30)

27

Lorie, J. H. & Niederhoffer, V. (1968). Predictive and Statistical Properties of Insider Trading. The Journal of Law & Economics, 11(1), 35-53.

MacKinlay, A. C. (1997). Event Studies in Economics and Finance. Journal of Economic Literature, 35, 13-39.

Nasdaq (2019A). Companies Listed on Nasdaq Stockholm. Accessed June 3, 2019, from Nasdaq OMX Nordic. Available at: http://www.nasdaqomxnordic.com/aktier/listed- companies/stockholm

Nasdaq (2019B). OMX Stockholm_PI (OMXSPI). Accessed June 3, 2019, from Nasdaq Global Indexes. Available at: https://indexes.nasdaqomx.com/Index/Overview/OMXSPI Princeton University (2008A). Event Studies with Stata. Accessed May 27, 2019, from Princeton University Library. Available at: https://dss.princeton.edu/online_help/

stats_packages/stata/eventstudy.html

Princeton University (2008B). Data Preparation for Event Studies using Stata. Accessed May 27, 2019, from Princeton University Library. Available at: https://dss.princeton.edu/

online_help/stats_packages/stata/eventstudydataprep.html

Seyhun, H. N. (1986). Insiders’ Profits, Cost of Trading, and Market Efficiency. Journal of Financial Economics, 16, 189-212.

Seyhun, H. N. (2000). Investment Intelligence from Insider Trading. Cambridge, MA:

MIT Press.

Smith, F. P. (1940). Management-Trading and Stock-Market Profits. The Journal of Business of the University of Chicago, 13(2), 103-117.

Spence, M. (1973). Job Market Signaling. The Quarterly Journal of Economics, 83(3), 355-374.

(31)

28

Appendix 1 –

Nasdaq OMX Stockholm Listed Companies and Sample Exclusions

Nasdaq OMXS Large Cap Listed Company Stocks as of April 2019 (Stock Ticker Symbol in Parentheses):

AAK (AAK) ABB Ltd (ABB) Addtech B (ADDT B)

Ahlstrom-Munksjö Oyj (AM1S) Alfa Laval (ALFA)

Arion Banki SDB (ARION SDB) Arjo B (ARJO B)

ASSA ABLOY B (ASSA B) AstraZeneca (AZN) Atlas Copco A (ATCO A) Atlas Copco B (ATCO B) Atrium Ljungberg B (ATRLJ B) Attendo (ATT)

Autoliv SDB (ALIV SDB) Avanza Bank Holding (AZA) Axfood (AXFO)

Beijer Ref B (BEIJ B) Betsson B (BETS B) BillerudKorsnäs (BILL) Boliden (BOL) Bonava A (BONAV A) Bonava B (BONAV B) Bravida Holding (BRAV) Castellum (CAST) Dometic Group (DOM) Elekta B (EKTA B) Electrolux A (ELUX A) Electrolux B (ELUX B) Epiroc A (EPI A) Epiroc B (EPI B) Ericsson A (ERIC A) Ericsson B (ERIC B), Essity A (ESSITY A) Essity B (ESSITY B)

Evolution Gaming Group (EVO) Fabege (FABG)

Fast. Balder B (BALD B)

Fenix Outdoor International B (FOI B) Getinge B (GETI B)

Hemfosa Fastigheter (HEMF)

Hemfosa Fastigheter Pref (HEMF PREF) Hennes & Mauritz B (HM B)

Hexagon B (HEXA B) HEXPOL B (HPOL B) Holmen A (HOLM A) Holmen B (HOLM B) Hufvudstaden A (HUFV A) Hufvudstaden C (HUFV C) Husqvarna A (HUSQ A) Husqvarna B (HUSQ B) ICA Gruppen (ICA) Industrivärden A (INDU A) Industrivärden C (INDU C) Indutrade (INDT)

Intrum (INTRUM) Investor A (INVE A) Investor B (INVE B) JM (JM)

Kindred Group (KIND SDB) Kinnevik A (KINV A) Kinnevik B (KINV B) Klövern A (KLOV A) Klövern B (KLOV B) Klövern pref (KLOV PREF) Kungsleden (KLED) Latour B (LATO B) Lifco B (LIFCO B) Loomis B (LOOM B)

Lundbergföretagen B (LUND B) Lundin Mining Corporation (LUMI) Lundin Petroleum (LUPE)

Millicom Int. Cellular SDB (TIGO SDB) Modern Times Group A (MTG A) Modern Times Group B (MTG B) Munters Group (MTRS)

NCC A (NCC A) NCC B (NCC B) NetEnt B (NET B)

NIBE Industrier B (NIBE B) Nobia (NOBI)

Nolato B (NOLA B) Nordea Bank Abp (NDA SE)

Nordic Entertainment Group A (NENT A) Nordic Entertainment Group B (NENT B) Nyfosa (NYF)

(32)

29

Oriflame Holding (ORI) Peab B (PEAB B) Pandox B (PNDX B) Ratos A (RATO A) Ratos B (RATO B) Resurs Holding (RESURS) SAAB B (SAAB B) Sagax A (SAGA A) Sagax B (SAGA B) Sagax D (SAGA D) Sagax pref (SAGA PREF) Sandvik (SAND) SCA A (SCA A) SCA B (SCA B) SEB A (SEB A) SEB C (SEB C) Securitas B (SECU B) Skanska B (SKA B) SKF A (SKF A) SKF B (SKF B) SSAB A (SSAB A) SSAB B (SSAB B) Stora Enso A (STE A)

Stora Enso R (STE R) SWECO A (SWEC A) SWECO B (SWEC B) Swedbank A (SWED A) Swedish Match (SWMA)

Swedish Orphan Biovitrum (SOBI) Svenska Handelsbanken A (SHB A) Svenska Handelsbanken B (SHB B) Tele2 A (TEL2 A)

Tele2 B (TEL2 B) Telia Company (TELIA) Thule Group (THULE) Tieto Oyj (TIETOS) Trelleborg B (TREL B) Wallenstam B (WALL B) Wihlborgs Fastigheter (WIHL) Veoneer SDB (VNE SDB) Vitrolife (VITR)

Volvo A (VOLV A) Volvo B (VOLV B) ÅF Pöyry B (AF B)

Nasdaq OMXS Mid Cap Listed Company Stocks as of April 2019 (Stock Ticker Symbol in Parentheses):

AcadeMedia (ACAD) Acando B (ACAN B) AddLife B (ALIF B)

Addnode Group B (ANOD B) Africa Oil (AOI)

Alligator Bioscience (ATORX) Alimak Group (ALIG) Ambea (AMBEA) AQ Group (AQ) Beijer Alma B (BEIA B) Bergman & Beving B (BERG B) Besqab (BESQ)

Better Collective (BETCO) Bilia A (BILI A)

BioArctic B (BIOA B) BioGaia B (BIOG B) Biotage (BIOT) Boozt (BOOZT) BTS Group B (BTS B) Bufab (BUFAB) Bulten (BULTEN) Bure Equity (BURE)

Bygghemma Group First (BHG) Byggmax Group (BMAX)

Calliditas Therapeutics (CALTX) Camurus (CAMX)

Catella A (CAT A) Catella B (CAT B) Catena (CATE) Catena Media (CTM) Cavotec (CCC) CellaVision (CEVI) Clas Ohlson B (CLAS B) Cloetta B (CLA B)

CLX Communications (SINCH) Concentric (COIC)

Collector (COLL)

Coor Service Management Holding (COOR) Corem Property Group A (CORE A) Corem Property Group B (CORE B) Corem Property Group Pref (CORE PREF) Creades A (CRED A)

Diös Fastigheter (DIOS) Duni (DUNI)

Dustin Group (DUST)

Investing Like an Insider: An Event Study Exploring the Possibilities of Positive Return for Outside Investors Following an Insider's Behavior