An Empirical Analysis of the Information Content of Aggregate Insider Trading in the French Context

1

An empirical analysis of the information content of aggregate Insider

trading in the French context

Meriem JERBi

Phd student, Prism, University of Paris I Sorbonne

[email protected]

Bahram SOLTANI (corresponding author)

Associate Professor University of Paris I Sorbonne

[email protected]

mailto:[email protected]

mailto:[email protected]

2

ABSTRACT

This paper examine corporate insider trading using the data extracted manually for French

companies during the period of 2007-2010. The major objective of the study is to empirically

investigate several important issues concerning the characteristics of corporate insider trading

and the conformity of transactions with mandatory regulations of the European Directive of

2003 on market abuse.

We investigate the time series trend change of corporate insider transactions during the period

of financial crisis. Results show that under favorable market conditions, they increase their

sell transactions, and when market performance falls they intensify their purchase

transactions.

We also investigate the reasons of increasing number of insider transactions to determine

whether this is due to their contrarian strategy (market to book value and volume) or private

information on firm future performance (return variation and ROA variation between 2009

and 2008). Results are significant for contrarian strategy assumption as we find negative

significant relation between insider trading ratios and trading volume and market to book

ration. It appears that corporate insiders select the undervalued stocks with high future returns.

This study is a contribution to literature in insider trading and may have practical implications

for regulatory and professional bodies.

JEL Classification: G01, G02, G11, G12, G14

Key words: insider trading, MAD EC directive, financial crisis, contrarian strategy,

private information, company’s future performance

3

An empirical analysis of the information content of aggregate Insider

trading in the French context

Introduction

There has been an overwhelming debate on insider trading since the inception of the financial

market. The recent spate of corporate and financial crises have highlighted in dramatic

fashion further reflection on this issue. However, an important aspect of insider trading is

related to the transactions made by corporate management and those who represent the major

shareholders within company. For various reasons, particularly the influential position of

these groups, it is essential to identify the circumstances in which corporate insiders have

traded. Whether they have bought on balance before abnormally good price movements and

sold on balance before poor market periods for the stocks of their companies. It is equally

important to investigate about the incentives of corporate traders and ethical issues involving

the insider transactions.

The interest in insider trading is threefold. First, from the academic perspective, it

involves several theoretical discussions regarding particularly the economics of information,

theories of justice, information asymmetry and informational advantage, signaling effect,

pricing and market performance, predictive value of information it may convey and its effect

on market efficiency and quality. Insider trading has also several practical implications

notably in terms of contrary-opinion rules, market overreaction, top management role in

corporate reporting, earnings announcements and cash flows prospects. Finally, the topic of

insider trading concerns ethical considerations and corporate management behavior,

managers’ incentives and their personal interests, fraudulent actions and conflicts of interest.

Because of its multi-dimensional effects on market performance, perceptions of investors and

analysts and their decision-making processes, the regulatory bodies have shown particular

4

interest in shaping the characteristics of insider trading and the effective policies in this area.

The regulatory efforts undertaken by the SEC in the United States and the European

Commission in Europe have accelerated particularly since the beginning of this century and

following several high profile financial scandals and management misconduct that began to

surface from late 2001 to 2003.

This paper aims to examine corporate insider trading in the European context by using

the data extracted from French market. The major objective of the research study is to

empirically investigate several important issues such as the reasons of shares’ trading

undertaken by members of top management and those representing major shareholders,

whether such transactions produce favorable outcome, the effect of these transactions on

market liquidity and the conformity of transactions with mandatory regulations of the

European Directive 2003 on insider dealing and market manipulation (market abuse). This

directive was among several other regulatory initiatives introduced, in the form of directive or

regulation, by the European Commission in 2002 and 2003. They essentially deal with buy-

back programs and stabilization of financial instruments (EC 2273/2003), quality of

disclosure prospectus of listed companies EC 203/71), fair presentation of investment

recommendations (EC 2003/125) and the application of international accounting standards

(EC1606/2002). All these mainly aim at improving the quality of information disclosed by

public companies and market functioning.

The remainder of the paper is organized as follows. Section II contains a discussion on

theoretical framework including literature review. Section III presents an overview of

regulatory framework especially in the European context. Data sources, sample companies,

research design and questions are presented in Section IV. This will be followed by the

analysis of our empirical results. The final Section discusses the concluding remarks,

contributions and limitations of study.

5

Theoretical framework

The academic debate and early research work on insider trading date back to the 1960s and

1970s, the period of the development of several fundamental issues in finance. The early

studies on this topic are Lorie and Niederhoffer (1968), Finnerty (1976a and b), The

expansion of financial markets in the 1980s and the subsequent periods in terms of number of

listed companies, their size and importance has been a determinant factor in growing interest

in academic research in the area. After almost five decades, there are still no clear-cut

academic explanations for a number of research questions on insider trading. Part of the

problem relates to mitigating results the research studies on this topic report. For instance,

several studies discussed in this section provide mixed support for the efficient market

hypothesis (e.g., Roddenberry and Bacon, 2011). On the other hand, the ever-increasing

complexity of such transactions due to increasing role of corporate management and

executive board members in capital market economy does not contribute to providing

satisfactory explanations for a number of outstanding questions on insider trading.

The early research of Jaffe (1974) argued that public investors who consistently traded

with the insiders based on announced insider transactions would have enjoyed excess risk-

adjusted returns. This argument was supported by the study of Trivoli (1980) in favor of

combining insider trading information with key financial ratios to increase the investors’

returns. The study of Nunn et al. (1982) went on saying that there are some insiders who are

more ‘inside’ than the others and the investors should consider in their strategy which group

of insiders (board chair, officers, directors versus other insiders) is involved in buying and

selling the companies’ shares. In contrast, Seyhun (1986) stated that the realizable return to

investors who attempt to use insider reports was not positive after considering total

transaction costs. The work of Lee and Solt (1986) supported this argument stating that it is

6

not possible to use aggregate insider trading activity as a guide to market trading. This was

later confirmed by the results provided in the studies of Seyhun (1988) and Chowdhury et al.

(1993).

Several research papers examine whether the buying and selling transactions of

corporate insiders convey information for market participants. Seyhun (1988) provides

evidence on the positive correlation between net aggregate insider trading activity and market

portfolio. Similarly, the study of Petit and Venkatesh (1995) showed a significant relationship

between insider trading and longer-term security performance indicting that on average,

insider trades are associated with substantial changes in share valuation. The study of Seyhun

(1988) shows that on an overall basis, “insiders increase their stock purchases prior to

increases in the stock market and decrease their stock purchases following increases in the

stock market” (p. 22). This may provide evidence on the ability of inside traders to predict, at

least partially, the favorable or unfavorable effects of economy-wide activity. However, the

author does not provide clear evidence on the extent to which the outcome of insiders’

transactions depends on their anticipation capacity of economic conditions or the nature and

quality of private information they possess because of influential position they hold in

company. In his following paper, Seyhun (1990) sheds light on the position of insider traders

who did not systematically foresee the market crash of 1987. “As stock prices began to

decline during the week of October 12, 1987, insiders became buyers rather than sellers” (p.

1386). The author believes that the corporate insiders’ attitude is mainly due to overreaction

effect in market pricing. In our opinion, this may also be related to the reluctant position of

corporate insiders at the time of crisis and their defensive strategy to protect their interests.

The issue of overreaction has been examined by Rozeff and Zaman (1998) who

provide evidence on increasing purchasing transactions by corporate insiders as stocks change

from growth to value categories and that they increase their purchases after low stock returns

7

and decrease their purchases after high stock returns. The authors state that this strategy of

corporate insiders is in contrast to the position of outside investors who overvalue growth

stocks and undervalue value stocks. Rozeff and Zaman (1998) acknowledge that “corporate

insider, who presumably have superior information, have incentives to take advantage of the

misvaluations, to the legally permissible extent, by buying value stock more heavily and/or

selling growth stocks more heavily (1998, p. 702). Although not explicitly expressed in the

paper of Rozeff and Zaman (1998), we believe that the potential profit arising from private

information the corporate insiders hold will result in a serious detriment to investors and

competitive market conditions.

The concept of contrarian beliefs is one the major topics studied in literature on

corporate insider trading. This concept is based on contrary-opinion rules widely used by

technical analysts. These technical trading rules are based on the premise that the majority of

investors are wrong as the market approaches peaks and troughs. Consequently, the traders

take advantage of the opportunities and engage in the opposition direction when the majority

of investors is either very bullish or very bearish. In line with this proposition, several papers

for example Jiang and Zeman (2009), Piotorski and Roulstone (2005), Gregory et al. (2011)

and Lakonishok and Lee (2001) examined the issue of insider trading from the viewpoints of

contrarian beliefs or superior information. Although different in research design and sample

data, they provide similar conclusion showing that insiders are able to predict the market

returns either on the basis of contrarian beliefs or because they benefit from superior

information on future cash-flows prospects.

Insider trading has been also studied in relation to other topics. For instance, it is

interesting to examine the relationship between information disclosure of the insiders’

transactions on the valuation implications of past and future earnings information. Very recent

study of Veenman (2012) provides evidence on the signaling effect of insider share purchases

8

on the future earnings information (similar to Piotroski and Roulstane 2005 and Roulstone

2008) but also has the valuation implications of past earnings signals for rational investors.

The aforementioned comments clearly show the importance of ethical considerations

regarding the insider trading. Indeed, several research studies provide evidence that members

of the board and top management who engage in such operations significantly benefit from

informational advantage and that these operations are mainly driven by superior information.

Examining the corporate fraud cases in the United States and Europe, Soltani (2012) sheds

light on significant insider trading of the chairman and several board members (for example in

Enron and HealthSouth). The study also referred to the cases of market abuse and unethical

behavior of several members of top management before the collapse of several multinational

companies. Several other studies (e.g., McGee 2008 and Show, 1990) emphasized the

importance of ethical principles in insider trading. However, this discussion goes beyond the

scope of the present study.

Regulatory framework

Following high profile financial failures, particular attention has been paid by the regulatory

bodies in developed capital markets to corporate issues, market functioning and quality of

information. One of the critical topics is related to insider trading which concerns both legal

and illegal conduct. The regulatory bodies are mainly concerned with transactions made by

corporate officers, directors, and employees when they buy and sell stock in their own

companies. Since 2002, most regulatory bodies require the mandatory reporting on

transactions made by corporate insiders. The distinction between legal and illegal versions

should be usually based on whether the corporate insiders when buying and selling their own

companies are in breach of a fiduciary duty or other relationship of trust and confidence,

while in possession of material, nonpublic information about the security. However there are a

9

number of problems in making such distinction including the difficulties associated with the

conditions under which insider trading violations may occur, the problems regarding the

trading by those who misappropriate such information (e.g., company’s management), and

"tipping" related information. Evidently, the regulatory bodies should treat the detection and

prosecution of insider trading violations as one of their enforcement priorities because of

tremendous effect the breach of a fiduciary duty may have on investor confidence in the

fairness and integrity of the securities markets and corporate disclosure.

In Europe, an initial step was taken in 1989 within the framework of the Directive on

Coordinating Regulations on Insider Trading (EC Directive 89/592). However, the

introduction of a comprehensive review of European Company Law and the EU Action Plan

in 2002 (Report of the High Level-Winter Group 2002) and recommendation 2003/284

emphasized the changes in redefining the term of insider trading and its disclosure policies. In

2003, following the recommendations of the High Level Group, three other initiatives were

put on the agenda by the European Commission. One was the introduction of a special

investigation to be requested by shareholders, another was the development of an EU wide

wrongful trading rule and the third was the imposition of directors' disqualification across the

EU.

The Directive on insider dealing and market manipulation (market abuse) (2003/6/EC)

was among the major initiatives of the Commission. It was aimed at reinforcing market

integrity by addressing the issues of price manipulation and the dissemination of misleading

information. The Commission acknowledges that “insider dealing and market manipulation

prevent full and proper market transparency” (Art. 21). By providing the detailed definitions

of the terms such as ‘insider trading’, ‘market manipulation’, and ‘financial instruments’, the

introduction of this Directive was a major step to reinforce the market quality. However, this

10

Directive similar to several others lacks an in-depth analysis of the conditions and disciplinary

policies which may be determinant factors in achieving final outcome. This analysis goes

beyond the scope of this paper but as highlighted by Soltani (2005), the European Directives

and proposed regulations particularly those dealing with financial market functioning and

quality of corporate disclosure policies do not fully respond to the imperatives in this respect.

Moreover, The EC Directives and national laws and regulations can be considered as

minimum requirements and this does not prevent the corporations from going further by

providing additional oversight measures. This point was also raised in the Green paper of the

European Commission (2011).

In the U.S. context, the term of insider trading was introduced in the SEC Act of 1934

directly through Section 16(b) and indirectly through Section 10(b). Section 16(b) defines

‘insiders’ as officers, directors, and large shareholders of more than 10 percent of any equity

class of securities of an issuing company. Before the introduction of SOX Act in 2002, the

information regarding insider trading was publicly available through EDGAR system (SEC’s

Electronic Data Gathering, Analysis and Retrieval). The introduction of SOX Act, particularly

the provisions underlined in section 403, gave new prominence to disclosure requirements

regarding the transactions involving management and principles shareholders. Based on these

requirements all insider trading activities should be publicly available on form 4 through the

EDGAR system within two business days. This short delay compared to five days in major

European market (e.g., France) and the requirement to disclose the information in electronic

formats of XBRL (eXtensible Business Reporting Language) and XML (eXtensible Markup

Language) provides several benefits for the American investors in terms of transparency,

timeliness and lower cost of information-processing.

11

Research design

We present in this part the hypotheses formulated in this study, sample selection and research

methodology and results.

Hypotheses

In line with the objectives outlined in the study, we test three following hypotheses

concerning the insider trading behavior measured in terms of increase or decrease of number

of transactions on comparative basis (before and after financial crisis), contrarian strategy and

the favorable (unfavorable) outcome of possible private information they may hold on their

transactions.

Hypothesis 1

The following hypothesis is defined to examine whether the insider traders increase

(decrease) their transactions during the time interval of 2007-2010 taking into account the

price movements. We test the trading behavior of insiders in terms of aggregate number of

transactions before and after crisis to make comparison between average number of

transactions with market movements represented by SBF 120 index.

H1: Do the inside traders, in anticipation of price movements, increase (decrease) their

purchases and sells before and after the periods of crisis.

Hypothesis 2

We investigate the reasons of increasing number of transactions made by insider traders to

determine whether this is due to their contrarian strategy and the selection of undervalued

stocks with high future returns.

12

H2: Based on contrarian strategy hypothesis, and considering the insiders as contrarian

investors, is there a negative relation between insiders trading activity and firm market to

book ratio and transaction volume during the periods of crisis.

Hypothesis 3

The corporate insiders have influential position in company and are in charge of financial

reporting and control mechanisms. This provides them with superior information compared to

other economic agents including the current and potential shareholders. It is interesting to

examine whether this may have a favorable effect on the performance of their transactions

compared to companies’ performance measured by ROA and the variation of firms’ during

2008 and 2009. This analysis is based on a cross sectional regression of insider trading

purchases using different proxies of contrarian strategy and superior information.

H3: Considering the influential position of corporate insiders in having better quality of

information, is there a positive relation between insiders trading activity and future firm

performance as measured by ROA variation and firm return variation between 2009 and 2008.

Sample and Data:

The period of study includes 2007 to 2010. This provides the possibility to examine the

insider trading behavior before and after financial crisis. The time series analysis is used to

detect any abnormal behavior in response to the variation of global market performance

during this period. For the second part of our analysis, we focus on time interval of 2008-

2009, considering year 2008 as ex-ante and 2009 as ex-post crisis period.

The sample consists of all companies included in index SBF 250. In the absence of database

regarding the information we intend to use for the purpose of this study, We collect manually

13

all the transactions of corporate insiders for sample companies, We eliminate from our sample

all firms that do not have a trading activity (buy or sell) during the period of study. Our final

sample contains 126 firms of SBF 250. The sample size includes 7418 transactions (4140

purchase transactions and 3278 sell transactions). For year 2008, our sample consists of 2181

transactions (1735 purchase transactions and 446 sell transactions).

We have collected our data from the official site of French financial market Authority (AMF)

which is a reliable source for the purpose of this study because the disclosed information on

this website is based on mandatory reporting of corporate insiders. In this study, we examine

only the purchasing and selling transactions undertaken by corporate insiders. We do not

consider the other types of insider transactions such as exercise of options; shares acquired

from compensation plan, private transactions etc. We have collected a number of information

from other sources. For instance the daily market variables including stock opening price,

stock volume, stock market value, stock market to book value, SBF120 price are collected

from DataStream database and ROA stock end year value from Thomson database.

Variable definitions

Insider trading activity:

We use the following measures of insider activity as defined by Seyhun (1990):

NP (Number of Purchase): denotes the number of total insider purchase transactions.

NS (Number of Sell): denotes the number of total insider sell transactions.

SP (Share Purchased): denotes the total number of share purchased by insiders

SS (Share Sold): denotes the total number of share sold by insiders

PRAT (Purchase Ratio to All Insider Transaction): is the number of purchase of firm i during

year t as a fraction of all purchases and sales by insiders.

14

SPRAT (Share Purchase Ratio to all insider Transaction): is the number of purchased sell of

firm i during year t as a fraction of all shares purchased and sold by insiders.

This ratio was also used by Piotroski and Roulstone (2005) to proxy insider trading behavior.

Seyhun (1990) argues that these ratios (PRAT and SPRAT) are not sensitive to changes in the

number of firms or trading activity over the time and it do not display heteroscedasticity or

extreme outliers.

Firm performance Measure

The performance measures are usually used in accordance with the objectives of the study.

Accounting variables such as ROA and ROE denote return on efficient utilization of firm’s

assets. It is a good performance indicator. Pandya and Rao (1998) argue that these measures

include depreciation and inventory costs and affect the accurate reporting of earnings. For this

reason, these data should be used in conjunction with financial measures such as return

change.

We measure the future company’s performance by using two accounting and financial

indicators:

ROA: future changes in Return on Assets of firm i= ROAt+1 - ROAt.

ROA is defined as net income (income available to common stockholders) divided by the

book value of total assets.

ROAt+1 is the Return on asset of firm i at year t+1 (2009) and ROAt is the Return on asset of

firm i at year t (2008).

A positive change is a sign of well-performed firm in year t+1 relative to year t.

ROA is the performance indicator most frequently used in previous studies (e.g., Piotroski

and Roulstone (2005) and Pandya and Rao (1998)).

15

returni: future changes in firm Return = return t+1 – return t.

lnr 2008: the contemporaneous 12 months market-adjusted return and is calculated as firm is

12 months log return for 2008 less the corresponding 12 months on the value weighted market

index, lnr2009 signifies future 12 months adjusted return measured as firm is 12 months log

return for 2009 less the corresponding 12 months on the value weighted market index.

To measure firm performance, Piotroski and Roulstone (2005) use firm future return.

We analyze change return level to proxy firm future performance because it reflects how

firms perform relative to last year. This measure was also used by Pandya and Rao (1998).

Measurement of contrarian factors

The measurement of contrarian factors should reflect the impact of current beliefs and

sentiment of investor on insider trading. The Book to market ratio and current return are

mainly used in previous studies [(Seyhun (1990), Rozeff and zaman (1998) and Piotroski and

Roulstone (2005)]. For the purpose of our study, we choose two measures:

Market to Book Ratio: denotes the market relative value of a company by comparing the

market value of a firm to its book value. The lower market to book ratio, the better the value

as this suggest a company’s assets are undervalued or that the company’s prospects are good.

Market to Book Ratio = Market price per stock ÷ Book value per stock

Rozeff and Zaman (1998), Piotroski and Roulstone (2005) and Gregory et al (2009) use book

value to market which reflects the market pricing error. However, this measure may also

indicate future performance measure. La Porta et al (1997) show that firms with greater book

value to book tend to have future earnings announcements periods returns. Thus, we add

another variable used in contrarian strategy literature to proxy investor sentiment.

16

LN VO i: is the second contrarian strategy measure representing the trading volume for firm i

during year 2008.

Blume et al (1994) use volume information and historical price information to predict future

prices change and show a signaling role of volume in return predictability. Datar et al (1998),

Brennan et al (1998), and Chordia et al (2001) show that stocks with lower trading volume

earn higher expected returns. Baker and Stein (2004) suggest that turnover or liquidity, can

serve as a sentiment index. It is thus conceivable that when investor sentiment becomes high

(low), trading volume is likely to increase (decrease). Baker and Stein (2004) show that the

increase in trading volume reflects a rise in investor sentiment. Lei (2005) investigates the

ability of past volume on predicting stock returns and show that trading volume trend has a

negative and significant relation with expected stock returns,

Other variables

Index return: this represents the SBF 120 return measured by LN (Pt+1/Pt) with Pt+1 is the

SBF120 opening price of day t+1 and Pt is the SBF120 opening price of day t.

Beta i: denotes a measure of past stock return volatility for stock i and represents the market

model slope coefficient. Beta is computed using daily return from January 2006 to December

2007. This measure was used by Seyhun (1990) in his study on insider response to the market

crash of 1987. Seyhun (1990) argued that pre crash risk should be an important predictor on

stock price during and after the crash, thus insider in the higher market risk firms are more

likely to observe and trade on the basis of mispricing caused by economy wide factors.

However, he found no evidence on this issue. Jiang and Zaman (2010) show that firms with

high incertitude for future are more associated with contrarian strategy.

MV i: Firm size i is measured by average Market value during year 2008.

17

Lakonishok and Lee (2001), show that insider in small firm are more able to predict future

market return on US market from 1975 to 1995. This result was later confirmed by Piotroski

and Roulstone (2005).

Results of empirical analysis

Time series Analysis of insider behavior and its comparison with market fluctuation:

First, we investigate aggregate insider trading activity during the period of study considering

global market fluctuations. The objective is to determine whether the corporate insiders react

to market performance particularly during the financial crisis.

Summary of descriptive statistics

Table 1 indicates different statistics on monthly insider trading variables (NP, NS, PRAT, and

SPRAT) and monthly market fluctuations (log of market index return of SBF120) during the

period 2007-2010 and for each year (2007, 2008, 2009 and 2010). Table 1 is completed by

table 2 in which we present average comparison tests of insider activity between periods.

[Insert table 1 here]


Table 1 shows that for the whole period (2007-2010) the PRAT and SPRAT ratios are slightly

higher than 50%. During this period of study, insiders tend to purchase more shares than what

they sell with a slight difference. The average number of purchase transaction which consist

of 86 (5823793) is higher than the number of sell transactions 68 (2930729). This purchase

18

activity corresponds to negative market return during the period. We also investigate this

results based on sub period analysis.

The difference in terms of trading activity is due, to some extent, to the level of trading in

2007 during which we observe a minor difference between buy and sell activity. Insider buy

in average 1152361 shares (100 buy transactions) and sell in average 2800475 shares (79

transactions). According to table 2, this difference is not significant for transaction number

and leads to a PRAT ratio close to 50% but the difference is significant between number of

purchased and sold shares and leads to relatively high SRPAT. The second important factor to

the purchase activity during the whole period is the insider trading activity (volume and

frequency) during 2008. As table 1 indicates, PRAT and SPRAT ratios show a large and

significant increase with a median of 81, 7% and 91, 9% respectively. This intensive buy

activity occurs during the unfavorable market performance period in which global return

market shows in average a negative trend of about -0,045%. According to table 2, the insider

purchase activity during 2008 is significantly higher than purchase activity during all analysis

periods.

Having considered market a sign of market recovery in 2009, we notice a reversal insider

activity as the PRAT and SPRAT ratio decrease to even fewer than 50%. Table 2 shows no

significant difference between sell and buy transactions during 2009. The activity trend was

supported during 2010 in which we observe a positive market return (0,001) and an intensive

selling activity significantly superior than insider buy transaction, as insider purchase in

average 41 shares and sell more than twice and low ratio as PRAT and SPRAT ratio reach in

average 31, 9% and 23, 3% respectively. Consequently, this may provide evidence that insider

purchase activity during the whole period is largely explained by the intensive buy activity in

2008 which has compensated the increase in sell activity especially in 2010.

19

Monthly trading insider activity:

Table 3 presents different measures of insider trading activity and global market variations for

every month from January 2007 to December 2010. We try also to report results on figure 1, 2

and 3 indicating insider behavior in time based on market variations.


[Insert figure 1 here]



As reported in table 3 and figures (1, 2 and 3), for the first half of 2007, we observe no

significant difference between buy and sell activity. In November 2007, the insider activity

purchase increases and the sell activity decreases matching with first sign of crisis in the

European market. Then insider buy activity shows a pick on January 2008 thus PRAT and

SPRAT ratio reach 92% and 95% respectively. This abnormal purchase activity corresponds

to market crash as we observe the unfavorable market return performance in this month (-0,

12). The relative market recovery decreases insider purchase transactions which still

significantly higher than insider sell activity. The second insider purchase activity pick

occurred in October 2008 (PRAT and SPRAT reach 92% and 93% respectively) which in fact

follow a second crash indicating significant negative return value (-0,155).

Insider purchase activity was at high levels until the first quarter of 2009. Following the

market recovery, the insider activity shows a reversal trend. As we observe, picks of insider

sell activity occur in September 2009 and March 2010 as PRAT and PRAT ratio reach in

average 17,3% and 18,4% respectively for September 2009 and 17% and 12, 9% respectively

20

for march 2010 with a favorable market performance (0,039 for September 2009 and 0,067

for march 2010). Selling activity was supported along the year 2010.

These results provide evidence for contrary position of the insider vis-à vis the market. For

this reason, when market performance is high, they increase their sell transactions and when

the market is low they increase their insider trading purchased. This strategy is mainly chosen

during financial crisis period with possible crashes. The results of this part of our analysis are

supported by Seyhun (1990) findings. He investigated insider trading response to market

crash of 1987 and found a higher PRAT and SPRAT ratios (80% and 62% respectively)

during October 1987.

Reasons of insider trading

In this section, we examine the reasons of aggregate insider purchase activity during crisis

period by using cross sectional analysis. The reasons may be either related to contrarian

strategy of corporate insiders, the quality of information they hold about future firm

performance or both.

To examine these questions we construct a model which considers variables reflecting

contrarian strategy and future firm performance.

Summary statistics and variables correlations:


Table 4 presents statistic summury (average, median and stdev) of insider trading activity

cross firms (NP, SP, NS, SS PRAT and SPRAT) and different variables (ROA, Beta, VO,

MV, MB and stock return). Our results show an average of 13.77 purchase transactions

(corresponding to 805117 share purchased) relative to 3.54 sell transaction (corresponding to

21

144193 share selled) in 2008. This indicates an average PRAT and SPRAT ratio of the order

69% for 2008.


Table 4 is completed by the infomation on table 5 in which we present the correlation matrix

between all variables studied. Based on Table 5, SPRAT present a negative significant

correlation with lnvo2008, MB2008, MV 2008 and current return and a positive significant

correlation with variation return and futur return. This correlation shows the same trend for

PRAT with no significant sign for current and futur returns. Results are in line with both

contrarian strategy and private information. Moreover, this indicates that insider activity is

more significant for small firms. We can conclude from the matrix correlation that insider

intensify their purchase activity on shares that are most undervaluated during the crisis period

(negative correlation with MB and return in 2008) and recover first post crisis (positive

correlation with return in 2009).

The results indicated on table 5 show a negative significant relation between MB ratio and the

variation of return. These observations are also made in Rozeff and Zaman (1998) and

Piotroski and Roulstone (2005) showing the ability of insider to detect overreaction and to

trade accordingly.

As shown in table 5, there is a negative relation between past volume traded and futur return.

This finding is supported by the resulst observde in Subrahmanyam and Anshuman (2001)

and Lei (2005) who show that stocks with lower trading volume provide higher expected

return.

22

Cross sectional analysis:

We analyze cross sectional insider trading using this regression model to investigate the

reasons of insider purchases:

(Insider activity) i = β 0 + β 1 MB i + β 2 LNVO i + β 3 Return i + β 4 ROA i

+ β 5 Beta i + β 6 i + i

(Insider activity) i : Reflect PRAT and SPRAT ratios which denote the number of purchases

as a fraction of all insider transaction (sell and buy) and the number of share purchased as a

fraction of share purchased and sold repectively, MBi is the Market to Book ratio of firm i

during year 2008, LNVO i is average Turnover volume during year 2008, Return i is the

difference between average return during year 2009 and average return during year 2008,

ROA I is the difference between Return On Asset during year 2009 and Return On Asset

during year 2008 Beta is the market model slope coefficient using daily return from 2006 to

2007 and MVi is the firm market value of firm i in 2008. Regression results are represented

on table 5.


Results show a negative significant relation between insider purchase activity and insider

ability to detect pricing error measured by MB ratio. This is in line with Rozeff and Zaman

(1998) findings as they found a high purchase ratio for highest BM portfolio and the lowest

purchase ratio for the lowest BM value. This is also observed by Piotroski and Roulstone

(2005) who show that insider purchase is inversely related to firm’s book to market ranking.

23

Furthermore, results show a negative significant relation between insider trading activity and

Lnvo2008, which reflects investor sentiment. Thus insider trade intensively on stocks trated

less (less liquid) on the market as they expect a market reversion.

Results on contrarian investor strategy are significant as it indicates that insiders can detect

pricing error on their own stocks and they trade accordingly. Moreover they trade against

investor sentiment in the anticipation for a futur reversal market.

Considring the regressionn analysis, our results show a positve non significant relation

between insider trading activity and variation on ROA. Thus insider trade more on stcoks

having good future performance prospects. Moreover, a positive relation between insider

purchase activity and return variation indciates that insider purchase more the stocks with

favorabler futur performance in the market. Howver, this relation is not significant.

The results show also that insider trading is positively correlated to firm risk and negatively

related to firm market value, although these are not significant. This may indicate that insider

invest more in small and risky firms.

Overall conclsuion is that insider purchase activity in crisis period is mainly motivated by

contrarian strategy based on detecting the error pricing rather than using private information.

It seems that insiders are able to detect overreaction and a rise of pessimism sentiment of

uninformed investor in the market which contribute to take away prices from their intrinsec

values and a decrease on traded volume. The fact that they act as contrarian investor, this may

rpovide a signal to the market and leads to a rapide price correction in favor of market

recovery.

24

Contribution and limits:

This study may have several contributions mainly because it is interested in the potentially

informed investor reaction in critical market conditions. During this period it is important to

study the response of specific category of market participants especially informed traders as

their behavior may provide a signal to other market participants. This may also be useful in

correcting the pricing error in market more rapidly. This study has also academic implications

for different reasons. First, most previous studies were conducted in English speaking

countries such as the U.S. and the UK. This research is conducted in the context of French

market which is one of the most important European financial centers. Besides that to our

knowledge there is no study in this area using an extensive database and variables employed

for French market. The study considers several variables related to contrarian investor

hypothesis and private information hypothesis and introduce a measure of investor feeling

(trading volume) as contrarian strategy proxy to reflect irrational investor pessimism during

the crisis period.

The study has several practical implications as it contributes to better understanding of market

participants mainly as it provides better insights into the effect of the informed investor during

the crisis and their contribution to market recovery. This may provide useful information for

the regulatory bodies and the organizations in charge of the market when supervising insider

transaction.

However, our research study is also subject to several limitations. The first is related to

sample size as the number of firm in this study is relatively low to other studies on U.S. This

is due to the difficulty related to data collection. Thus we did not dispose of insider database

and we need to collect transactions from AMF site. Secondly, our paper reflects insider

25

reaction in a critical period thus they may have different behavior in normal period than we

can enlarge our analysis by increasing analysis period to compare their reaction in normal and

critical periods.

Conclusion

The major objective of this study is to understand better about insider trading strategy in

medium-sized market such as France during financial crisis. We investigate the time series

trend change of corporate insider transactions to provide explanations about their behavior

and perceptions towards financial crisis. Results show that insider purchase activity is

negatively correlated to global market variation. Under favorable market conditions, they

increase their sell transactions, and when market performance falls they intensify their

purchase transactions. Seyhun (1990) shows the same insider behavior in US market

regarding the insider response to market crash of 1987.

We have examined a number of questions regarding the strategy of corporate insiders and the

reasons behind their strong intervention during the crisis period. Is it because they detect a

pricing error prorogated by deviation from fundamental value following noise trade? Is it

because they hold private information, view their positions and their proximity to private

information, about future firm performance? We used a model considering the two

assumptions and we use regression concerning insider purchase measures based on contrarian

strategy proxy and using the trading volume in 2008 and market to book ration in 2008. We

also test for private performance information measure proxy calculated by return variation

and the variation on ROA between 2009 and 2008. Results are significant for contrarian

strategy assumption as we find negative significant relation between insider trading ratios and

trading volume and market to book ration. Results on private performance information are

26

positive suggesting that insider trade intensively on stocks that perform well in future but are

not significant.

This supports the idea that corporate insiders have a clear strategy in terms of purchasing,

selling, the timing and conditions under which they make transactions in financial market.

The strategy based on anticipation is not obviously limited to corporate insiders and concerns

all categories of market participants. However, there are several major differences between

the quality and the conditions of insider trading and other market participants. One major

difference is related to the influential position of corporate insiders and informational

advantage they have compared to others. Another difference may be related to their

knowledge and expertise of market conditions and environmental factors. Is this added value

considered ‘private information’? Does this outstanding position affect their incentives to

intervene in market in specific period or time interval? Do they behave ethically and if not, to

what extent they are responsible for using private information to detriment of others. This

paper does not claim to provide answers to these questions. However, we have tried to shed

light on strategy of corporate insiders in relation to market conditions, the position of other

investors and regulatory framework.

Bibliography:

Baker, M, and Stein, J. (2004): “Market Liquidity as a Sentiment Indicator”, Journal of

Financial Markets, 7, pp. 271-299.

Blume, L., Easley, D., and O'Hara, M., (1994): “Market Statistics and Technical Analysis: the

Role of Volume”, Journal of Finance 49, pp. 153-181.

27

Brennan, M.J., Chordia.T. and Subrahmanyam, A. (1998): “Alternative Factor Specifications,

security Characteristics, and the Cross-Section of Expected Stock Returns”, Journal of

Financial Economics 49, pp. 345-373.

Chordia, T., Subrahmanyam. A. and Anshuman, V.R. (2001): “Trading Activity and Expected

Stock Returns”, Journal of Financial Economics 59, pp. 3-32.

Chowdhury, M., How, J. S. and Lin, J. C. (1993): “The Relation between Aggregate Insider

Transactions and Stock Market Return”, Journal of Financial and Quantitative

Analysis, 28, pp. 431-438.

Council Directive. (1989): “Coordinating Regulations on Insider Trading”, Common Mkt.

Rep. (CCH), 1761 ("EC Directive 89/592").

Datar, V.T., Naik. N.Y. and Radcliffe, R. (1998): “Liquidity and Stock Returns: An

Alternative test”, Journal of Financial Markets 1, pp. 203-219.

European Commission. (2011): Green Paper- the EU corporate governance framework. COM:

(2011) 164 final, pp. 24.

European Parliament and of the Council. (2003): “Directive on Insider Dealing and Market

Manipulation (Market Abuse)” 2003/6/EC, 28 January, pp. 10.

Finney, J. E. (1976a): “Insiders Activity and Inside Information: A multivariate Analysis”,

Journal of Financial and Quantitative Analysis, 11 (2), pp. 205-215.

Finney, J. E. (1976a): “Insiders and Market Efficiency”, Journal of Finance, 31, pp. 1141-

1148.

Gregory, A., Tharyon, R. and Tonks, I. (2009): “Insider trading in Glamour and Value

Firms “, working paper, 09/04, Xfi Centre for Finance and Investment.

Jaffe, J. F. (1974): “Special Information and Insider Trading”, Journal of Business, 47 (2), pp.

410-428.

Jiang, X. and Zaman, M. (2010): “Aggregate insider trading: Contrarian beliefs or superior

28

information?”, The Journal of Banking and Finance, 34, pp. 1225-1236.

La Porta, R., Lakonishok, J., Shleifer, A. and Vishny, R., (1997) : “Good news for value

stock: Further evidence on market efficiency”. Journal of Finance 52, pp. 859–874.

Lakonishok, J., Lee, I. (2001): “Are insider trades informative?”, The Review of Financial

Studies, 14, pp. 79–111.

Lee, W.Y. and Solt, M.E. (1986): “Insider Trading: A Poor Guide to Market Timing”, Journal

of Portfolio Management, 12, pp. 65-71.

Lei, Y.C. (2005): “The trading volume trend, investor sentiment, and stock returns”, PHD

Dissertation, Louisiana State University, 2004.

Lorie, H. and Niederhoffer, V. (1968): “Predictive and Statistical Properties of Insider

trading”, Journal of Law and Economics, 11, pp. 35-53.

McGee, R. W. (2008): “Applying Ethics to Insider Trading”, Journal of Business Ethics, 77,

pp. 205-217.

Nunn, K. P, Madden, G. P. and Gombola, M. J. (1982). “Are Some Insiders More ‘Inside’

than Others?” Journal of Portfolio Management, 9(3),

Pandya, A.M. and Rao, N. (1998): “Diversification and firm performance: an empirical

evaluation”, Journal of Financial and Strategic Decisions, 11, pp. 67-81.

Petit, R.R. and Venkatesh, P.C. (1995): “Insider Trading and Long-Run Return Performance”,

Financial Management, 24, No. 2, pp. 88-103.

Piotroski, J.D and Roulstone, D.T. (2005): “Do insider trades reflect both contrarian beliefs

and superior knowledge about future cash flow realizations?” Journal of Accounting

and economics, 39, 55-81.

Roddeenberry, S. and Bacon, F. (2011): “Insider Trading and Market Efficiency: Do insiders

29

buy Low and Sell High?”, Journal of Finance and Accountancy, 8, pp. 16.

Roulstone, D. T. (2008): “Insider Trading and the Information Content of Earnings

Announcements”. Working Paper, The Ohio State University.

Rozeff, M., Zaman, M. (1998) “Overreaction and insider trading: evidence from growth and

value portfolios”. The Journal of Finance, 53, pp. 701–716.

Seyhun, N. (1986): “The January effect and aggregate insider trading”, The Journal of

Finance, 43, pp. 129 - 141.

Seyhun, N. (1988): “The information content of aggregate Insider trading”, The Journal of

Business, 61, pp. 1-24.

Seyhun, N. (1990): “Overreaction or Fundamentals: some lessons from insiders’ response to

the market crash of 1987”, The Journal of Finance, 45, pp. 1363-1388.

Show, B. (1990): “Shareholder Authorized Inside Trading: A Legal and Moral Analysis”,

Journal of Business Ethics, 9, pp. 913-928.

Soltani, B. (2005): “Factors Affecting Corporate Governance and Audit Committees in

Selected Countries”, the IIA Research Foundation, pp. 200.

Soltani, B. (2012). “The Anatomy of Corporate Fraud”, Paper presented at the annual Meeting

of the American Accounting Association. August 2012, Washington DC, pp. 75. (New

version forthcoming in Journal of Business Ethics).

Trivoli, G. W. (1980): “How to profit from Insider Trading Information”, Journal of Portfolio

Management, 6(4), pp. 51-56.

Veenman, D. (2012): “Disclosures of Insider Purchases and the Valuation Implications of

30

Past Earnings Signals”. The Accounting Review, 87(1), pp. 313-342.

Winter Group. (2002): “Report of the High Level Group of Company Law Experts on A

Modern Regulatory Framework for Company Law in Europe”, Brussels.

31

Table 1: Statistics summary of trading activity

NP is insider purchase number, NS is insider sell number, SP is number of shares purchased, SS is number of shares

sold, PRAT is number of purchases as a fraction of all insider transaction (sell and buy) and SPRAT is number of share

purchased as a fraction of share purchased and sold.

DATE LNRM NP SP NS SS PRAT SPRAT

Sum 2007-2010 - 4140 3278 279542078 140674971 - -

average 2007-2010 -0,0069 86 68 5823793 2930729 0,53085 0,57807

median 2007-2010 -0,0060 73 67 2367954 1248040 0,48429 0,67461

stdev 2007-2010 0,0641 57 40 9232702 6367672 0,24030 0,33564

sum 2007 - 1198 947 133828335 33605695 - -

Average 2007 -0,0002 100 79 11152361 2800475 0,55848 0,70468

Median 2007 0,0021 96 64 7580370 1776463 0,52820 0,78074

stdev2007 0,0342 39 38 13710029 3784584 0,16076 0,22665

sum 2008 - 1735 446 101444722 18168377 - -

Average 2008 -0,0458 145 37 8453727 1514031 0,78092 0,80008

Median 2008 -0,0273 146 37 6488898 844115 0,81766 0,91982

stdev2008 0,0700 56 20 8890843 1639817 0,13381 0,24399

Sum 2009 - 715 804 36858205 10106768 - -

Average 2009 0,0174 60 67 3071517 842231 0,46440 0,57378

Median 2009 0,0429 37 70 1381215 751660 0,36835 0,68831

stdev2009 0,0752 43 42 4490619 640371 0,25635 0,36044

Sum 2010 - 492 1081 7410817 78794130 - -

Average 2010 0,0010 41 90 617568 6566178 0,31962 0,23373

Median 2010 -0,0100 38 77 371381 1604999 0,30196 0,17540

stdev2010 0,0586 18 38 829960 11594863 0,12131 0,19974

32

Table 2: Insider activity comparison between periods



purchased as a fraction of share purchased and sold. This table presents the difference between average insider

purchases and sells activity in the same year and between different years of study.

NP2008 t-stat SP2008 t-stat SP2007 t-stat SP2009 t-stat SP2010 t-stat

SP2008 107* 5,7

SP2007 66* 3,05 -42* -4,14

SP2009 77* 3,8 -30* -2,68 12 0,66

SP2010 54* 2,85 -53* -4,7 -11 -0,63 -23 -1,26

NP2007 45* 2,074 -63* -4,82 -21 -1,11 -33* -2,2 -10 -1,26

NP2009 85* 4,87 -23* -1,57 19 1,55 7 0,34 30* 2,04

NP2010 103* 5,1 -4 -0,77 38* 3,82 26* 2,1 49* 3,9

NS 2008 t-stat SS 2008 t-stat SS 2007 t-stat SS2009 t-stat SS2010 t-stat

SS2008 7E+06* 2,59

SS2007 6E+06* 2,068 -1,00E+06 -1,04

SS2009 8E+06* 2,84 7,00E+05 1,5 2,00E+06 1,7

SS2010 2,00E+06 0,41 -5,00E+06 -1,44 -4,00E+06 -0,98 -6,00E+06 -1,7

NS2007 -3,00E+06 -0,58 -1E+07* -2,35 -8E+06* -2,1 -1E+07* -2,6

-

5,00E+06 -1,04

NS2009 5E+06* 2,87 -2,00E+06 -1,3 -3,00E+05 -0,15 -2,00E+06 -1,61 3,00E+06 0,9

NS2010 8E+06* 2,96 9,00E+05 1,73 2,00E+06 1,9 2,00E+05 1,31 6,00E+06 1,7

PRAT 2008 t-stat PRAT 2007 t-stat PRAT 2009 t-stat

PRAT 2007 0,23* 4,41

PRAT 2009 0,32* 3,91 0,09 0,87

PRAT 2010 0,47* 7,41 0,24* 3,77 0,18 1,59

SPRAT 2008 t-stat SPRAT 2007 t-stat SPRAT 2009 t-stat

SPRAT 2007 0,1 1,2

SPRAT 2009 0,23 1,78 0,13 0,88

SPRAT 2010 0,57* 5,9 0,47 0,23 0,31* 2,63

*significant at 5% level

33

Table 3: Insider trading activity in 128 firms from 01/01/2007 to 31/12/2010



purchased as a fraction of share purchased and sold and LNRM is log return of market (SBF120).

DATE LNRM NP NS SP SS PRAT SPRAT

january-07 0,017876 69 83 8955112 1933500 0,4539 0,8224

february-07 -0,02728 63 115 2580082 3743071 0,3539 0,4080

march-07 0,025991 97 68 1637226 1619426 0,5879 0,5027

april-07 0,056474 71 97 1307221 1368085 0,4226 0,4886

may-07 0,020987 99 139 18749400 2266267 0,4160 0,8922

june-07 -0,01821 108 148 12261140 3512941 0,4219 0,7773

july-07 -0,06628 94 49 19062240 14306572 0,6573 0,5713

august-07 0,00859 146 35 10433738 834988 0,8066 0,9259

september-07 -0,00446 74 59 1271263 2348332 0,5564 0,3512

october-07 0,032298 60 60 1855704 510701 0,5000 0,7842

november-07 -0,03885 192 34 6205628 335707 0,8496 0,9487

december-07 -0,00994 125 60 49509581 826104 0,6757 0,9836

january-08 -0,12827 249 20 32518771 1657219 0,9257 0,9515

february-08 -0,03609 125 19 17788282 366215 0,8681 0,9798

march-08 -0,00752 173 39 9496182 879814 0,8160 0,9152

april-08 0,074862 61 52 729512 5557232 0,5398 0,1160

may-08 -0,00776 79 69 3234921 1260871 0,5338 0,7195

june-08 -0,12803 156 48 6930685 2858819 0,7647 0,7080

july-08 -0,01859 136 30 7015420 564160 0,8193 0,9256

august-08 0,026801 90 10 2918132 93041 0,9000 0,9691

september-08 -0,09205 167 69 6047111 3383120 0,7076 0,6412

october-08 -0,15583 222 17 2323697 161067 0,9289 0,9352

november-08 -0,07331 169 34 2553503 578403 0,8325 0,8153

december-08 -0,00432 108 39 9888506 808417 0,7347 0,9244

january-09 -0,09342 97 18 14966406 74475 0,8435 0,9950

febrary-09 -0,10594 134 37 3937402 168646 0,7836 0,9589

march-09 0,057065 140 20 2536175 107567 0,8750 0,9593

april-09 0,133189 37 20 8885229 1256550 0,6491 0,8761

may-09 0,046428 48 71 145715 2199430 0,4034 0,0621

june-09 -0,04977 82 93 2412210 802289 0,4686 0,7504

july-09 0,080609 32 64 1588022 406864 0,3333 0,7960

august-09 0,069886 23 82 237931 987440 0,2190 0,1942

september-09 0,039367 36 171 343840 1517382 0,1739 0,1847

october-09 -0,05511 29 69 1174408 701030 0,2959 0,6262

november-09 0,027965 30 81 170463 621492 0,2703 0,2152

december-09 0,059029 27 78 460403 1263603 0,2571 0,2671

january-10 -0,05284 22 82 206250 1033932 0,2115 0,1663

februay-10 0,007539 25 48 335744 529760 0,3425 0,3879

march-10 0,06771 36 175 296718 1989687 0,1706 0,1298

april-10 -0,04383 49 130 529962 1826544 0,2737 0,2249

may-10 -0,08327 85 66 3171408 1239530 0,5629 0,7190

june-10 -0,02751 54 69 755385 1383455 0,4390 0,3532

july-10 0,080077 43 60 200618 886815 0,4175 0,1845

august-10 -0,04414 55 75 407018 681694 0,4231 0,3739

september-10 0,061721 40 92 448573 4028744 0,3030 0,1002

october-10 0,038551 17 67 74966 36864565 0,2024 0,0020

november-10 -0,05538 32 138 720345 4006581 0,1882 0,1524

december-10 0,063017 34 79 263831 24322823 0,3009 0,0107

34

Figure 1: Daily insider trading number of buy and sell and return market stock in 128 firms from 01/01/2007 to

31/12/2010

NP is insider purchase number, NS is insider sell number and LNRM is the log of SBF120 market return

35

Figure 2: Daily insider trading number of share purchased and sold and return market stock in 128 firms from

01/01/2007 to 31/12/2010

SP is number of shares purchased, SS is number of shares sold and LNRM is the log of SBF120 market return

36

Figure 3: Daily insider trading activity (ratio) and return market stock 128 firms from 01/01/2007 to 31/12/2010

PRAT is number of purchases as a fraction of all insider transaction (sell and buy) and SPRAT is number of share

purchased as a fraction of share purchased and sold and LNRM is the log of SBF120 market return

37

Table 4: Statistics summary of Cross sectional study



purchased as a fraction of share purchased and sold, ROA 2008 is Return On Asset of 2008, ROA 2009 is Return On

Asset of 2009, ROA is the difference between return ROA 2009 and ROA 2008, BETA is the market model slope

coefficient using daily return from 2006 to 2007, Lnvo 2008 is the log of 2008 volume, MV is the firm Market Value,

lnr2008 is the contemporaneous 12 months market-adjusted return and is calculated as firm is 12 months log return for

2008 less the corresponding 12 months on the value weighted market index, lnr2009 is future 12 months adjusted return

measured as firm is 12 months log return for 2009 less the corresponding 12 months on the value weighted market index

and Return is the difference between lnr2009 and lnr2008 and MB 2008 is the Market to Book ratio in 2008. For each

of these variables we present statistics (average, median and standard deviation)

Average

median stdev

NP 13,77 4 23,80

SP 3,54 1 5,95

NS 805116,84 17478,696 3662387,35

SS 144193,47 1310,006 546006,79

PRAT 0,69 0,828 0,36

SPRAT 0,69 0,950 0,40

ROA 2008 2,78 3,790 7,32

ROA 2009 1,69 3,040 10,44

ROA -1,09 -0,075 7,92

BETA 0,12 0,101 0,31

lnvo2008 4,32 4,456 2,87

lnr2008 -0,00072 -0,00035 0,00189

m2b2008 1,81 1,540 1,61

mv2008 7380,00 1078,558 15863,53

Lnr 2009 0,00077 0,00057 0,001527

Return 0,00149 0,00088 0,0031

38

Table 5: Sample Cross Sectional correlations

PRAT is number of purchases as a fraction of all insider transaction (sell and buy) and SPRAT is number of share

purchased as a fraction of share purchased and sold, BETA is the market model slope coefficient using daily return from

2006 to 2007, Ln vo 2008 is the log of 2008 volume, MV is the firm Market Value, lnr2008 is the contemporaneous 12

months market-adjusted return and is calculated as firm is 12 months log return for 2008 less the corresponding 12

months on the value weighted market index and Return is the difference between lnr2009 and lnr2008 and MB 2008 is

the Market to Book ratio in 2008.

PRAT SPRAT ROA BETA LNVO MB REND lnr2009 lnr2008

PRAT 1

SPRAT 0.919** 1

ROA 0.087 0.079 1

BETA 0.148 0.147 0.086 1

lnvo -0.345** -0.354** -0.051 -0.069 1

MB -0.354** -0.347** 0.33** -0.079 0.104 1

Return 0.178* 0.243** -0.014 0.27** -0.166 -0.240** 1

lnr2009 0.172 0.225* 0.013 0.219* -0.150 -0.200* 0.907** 1

lnr2008 -0.159 -0.225* 0.034 -0.275** 0.157 0.240** -0.940** -0.71** 1

126 126 126 126 126 126 126 126

** Correlation is significant at 0.01 levels (bilateral) and * Correlation is significant at 0.05 levels (bilateral)

39

Table 6: Insider trading reasons: regression analysis

(Insider activity) i = β 0 + β 1 MB i + β 2 Volume i + β 3 Return i + β 4 ROA i + β 5 Beta i + β 6 i + i

Insider activity i is the insider trading activity for firm I during 2008 represented by PRAT and SPRAT, PRAT is number

of purchases as a fraction of all insider transaction (sell and buy) and SPRAT is number of share purchased as a fraction

of share purchased and sold, MB is the Market to Book ratio for firm i in 2008, Lnvo 2008 is the log of 2008 volume,

Return is the difference between lnr2009 and lnr2008 ROA is the difference between return ROA 2009 and ROA

2008, BETA is the market model slope coefficient using daily return from 2006 to 2007, MV is the log of firm Market

Value in 2008, observation nb indicates the number of observation per regression and R2 is R-squared for the regression.

Values in parenthesis indicate t-stat.

Insider trading activity

PRAT SPRAT

β 0

MB

Ln VO

Return

ROA

Beta

MV

Observation nb

R2

0,94* 0,95*

(14,0) (12,97)

-0,07* -0,077*

(-3,54) (-3,32)

-0,030* -0,033*

(-2,47) (-2,43)

2,07 11,09

(0,21) (1,09)

0,00008 0,00033

(0,02) (0,08)

0,086 0,07

(0,89) (0,64)

-2,461e-06 -2,95e-06

(-1,06) (-1,15)

126 126

0.24 0,25

*significant at 5% level