Insider trading and stock prices

Insider trading and stock prices

International Review of Economics and Finance 22 (2012) 254–266 Contents lists available at SciVerse ScienceDirect International Review of Economics...

214KB Sizes 1 Downloads 192 Views

International Review of Economics and Finance 22 (2012) 254–266

Contents lists available at SciVerse ScienceDirect

International Review of Economics and Finance journal homepage: www.elsevier.com/locate/iref

Insider trading and stock prices Manouchehr Tavakoli a,⁎, David McMillan b, 1, Phillip J. McKnight c, 2 a b c

University of St. Andrews, School of Management, St. Andrews, Scotland, KY16 9RJ UK University of Stirling, Management School, Stirling, Scotland, FK9 4LA UK University of Wisconsin, Lubar School of Business, Milwaukee, WI 53211, United States

a r t i c l e

i n f o

Article history: Received 5 January 2010 Received in revised form 8 March 2011 Accepted 3 October 2011 Available online 15 November 2011 Keywords: Insider trades Abnormal returns Market efficiency

a b s t r a c t We examine the informational content of insider trades and its value to market investors using a US dataset. Overall, our results support the view that insider actions have positive predictive power for future returns. However, these results may come with some caveats. First, it is not the actions of all insiders (directors, officers and large shareholders) that have predictive power for future returns, but typically only those of directors and officers (senior management). Second, while director actions have predictive power for firm of all sizes, officers only have predictive power for small firms. The signal emanating from buys is stronger than the signal emanating from sells. Finally, the trading actions of directors, and to a lesser extent, officers have significant effects on the trading behaviour of other groups of insiders. © 2011 Elsevier Inc. All rights reserved.

1. Introduction In this paper, we examine the informational content of insider trades and its value to market investors. The demand for credible yet lawful information that could assist investors in beating market averages is enormous, as typified by, for example, a number of data vendors, such as CDA/Investnet, who use insiders' trades to predict returns, for institutional and individual investors (Lakonishok & Lee, 2001). Moreover, one of the best known signalling theories is the insider trading of corporate management team. However, there remains uncertainty as to the predictive content of insider trading. Thus, the motivations for this paper are several. First, it contributes to and extends the existing literature by re-considering the predictive content of insider buy and sells across four different categories of insider trader using a new dataset that covers a more recent time period than previously considered. In particular, the dataset considered here extends over the period from 2000 to 2007 and thus takes in the bear market of the early 2000s and the market recovery that extended from 2003. Previous studies have typically not included large bear market periods such as the early 1970s or in 1987 (one exception would be Lakonishok & Lee, 2001). Thus, at the very least, this paper will serve as a check on previous results, but also extends the weight of the information set in regard to whether insider information has predictive power by considering a sample that incorporates both bull and bear markets. Furthermore, the dataset covers periods marked by both economic expansion and contraction and thus implicitly considers whether insiders can predict the state of the economy, the future prospects of their firm and signal this ability through their own trading. Utilising such a data set, we believe, helps provide a robust conclusion. Moreover, we consider the actions of four groups of insiders and, distinct from the majority of the literature, we split management into directors and senior management. In general we believe that the majority of meaningful insider trading will come from

⁎ Corresponding author. Tel.: + 44 1334 462810; fax: + 44 1334 462812. E-mail addresses: [email protected] (M. Tavakoli), [email protected] (D. McMillan), [email protected] (P.J. McKnight). 1 Tel.:+44 1786 467309. 2 Tel.: + 1 424 644 4235. 1059-0560/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.iref.2011.11.004

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

255

these groups and whereas the existing literature tends to have a single management category (e.g., Lakonishok & Lee, 2001), splitting them may provide greater insight. Hence, a second motivation is that while the management group may have the similar levels and type of information, their trading behaviour may not be the same indicating a hierarchy of information content. Additionally and third, we can examine whether insider actions are, in part, themselves conditioned by the actions of other insider groups. We also, and fourth, examine the effects based on firm size and whether the insider actions are buy or sell and whether the actions of insiders are affected by past returns. This latter issue has not been typically considered in the previous work (in partial exception Iqbal and Shetty, 2002 and Chowdhury, Howe, and Lin, 1993 consider VAR estimation to test for Granger causality between returns and insider actions). The rationale for the use of insider information is that managers know more about their companies than any outsider, including Wall Street analysts and as such investors could benefit from observing the behaviour of insiders. Studies of managerial decisions suggest that insiders are better informed about their companies' prospects and that the market is slow in adjusting to managerial signals. 3 Prior US based research has examined the relationship between insider trading and the subsequent behaviour of share returns (Finnerty, 1976; Jaffe, 1974; Jeng, Metrick, & Zeckhauser, 2003; Lakonishok & Lee, 2001; Lin & Howe, 1990; Lorie & Niederhoffer, 1968; Penman, 1982; Rozeff & Zaman, 1998; Seyhun, 1986; Seyhun, 1988). Moreover, insider active trading could suggest features of over-confidence (Kumar, 2009), which in turn could provide confidence for others (Givoly & Palmon, 1985). Furthermore, given their privileged information the trading patterns of these individuals are considered to be different from the normal individual (noise) traders (Fidrmuc, Goergen, & Renneboog, 2006; Kaniel, Saar, & Titman, 2008; Kyle, 1985). The evidence from this body of work appears to suggest that there is information content in corporate insider trading strategies using private information to earn abnormal returns (Carter, Mansi, & Reeb, 2003). Furthermore, it also suggests that the average investor can make an abnormal return by simply observing the trading behaviour of certain directors and managers of the firms they manage. In addition, it is argued that that nature of any link between insider activity and returns can depend on whether the insiders buy or sell. This is because while an insider buy can convey positive and thus favourable information on the firm's prospects, it is less clear about the information content of an insider sell as it may represent either unfavourable information about the firm's prospects or it could simply be to meet the liquidity needs of the insider (Fidrmuc et al., 2006). Nonetheless, this positive finding is not universal. Eckbo and Smith (1998) report that the insiders of firms listed on the Oslo Stock Exchange do not earn abnormal profits. While, Chakravarty and McConnell (1999) found that there were no distinguishable price effects between a confessed insider trader, Ivan Boesky, in Carnation's stock in 1984 prior to Nestlé's acquisition of Carnation and that of non-insider. In line with previous studies (Iqbal & Shetty, 2002; Lakonishok & Lee, 2001) in order to address the issues outlined above, this paper examines whether there exists a time-series relationship, and in particular, whether there exists predictive power, between insider transactions and stock returns over the period 2000–2007. Moreover, as in Iqbal and Shetty (2002), it is assumed that the time series relationship between insider transactions and stock returns will reveal any firm-specific information without needing to examine any corporate events. Thus, we will test for the existence of a positive relationship between insider transactions and subsequent stock returns. Such a relationship will show that insiders purchase (sell) before stock prices rise (decrease). Moreover, we will consider whether there exists a negative association between stock returns and subsequent insider transactions, which would indicate that insiders sell (purchase) after stock prices rise (decrease). In addition, as noted by Seyhun (1988), there is a belief that firm size is important, thus, we consider the effects of insider trades on return by size. Further, we consider the separate influence of buy and sell signals. As noted above, it is believed that buy signals may carry more information than sells. Finally, we consider whether the actions of one group of insiders affect the actions of another group of insiders. Of note, Fidrmuc et al. (2006) argued that the actions of directors and officers (senior managers) have predictive power for not only future returns, but also for the behaviour of other insiders. 2. Brief review of past literature and motivation Several studies have shown that insiders can use privileged insider information for their personal gain. Seyhun and Bradley (1997) presented evidence that insiders sold shares of their firms before filing for bankruptcy and bought after prices had fallen. Other studies have presented evidence that insiders are able to strategically trade their own shares to earn abnormal returns around major corporate announcements or events such as new issue announcements (Karpoff & Lee, 1991), stock repurchases (Lee, Mikkelson, & Partch, 1992), dividend announcements (Cheng, Davidson, & Leung, 2011; John & Lang, 1991), listing and delisting (Lamba & Khan, 1999), takeover announcements (Seyhun, 1990), and the event itself (Ma, Sun, & Tang, 2009; Rozeff & Zaman, 1998; Seyhun, 1986; Seyhun, 1992). These studies observe positive returns following insider purchases and negative returns following sales. Also, negative returns precede purchases and positive returns precede sales. It is also suggested that insiders, in aggregate, are able to predict market movements and thus able to time the market (Lakonishok & Lee, 2001; Seyhun, 1988). Hence, when insiders are optimistic, markets tend to do well and when they are pessimistic, markets tend to do poorly. Insiders also predict aggregate movements of small companies better than of large companies.

3 For example, Ikenberry, Lakonishok, and Vermaelen (1995) found evidence of abnormal returns upon companies announcement of their share repurchases, as insiders perceive this as undervalued shares.

256

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Lakonishok and Lee (2001) and a more recent study by Hotson, Singh, and Singh (2008) demonstrated that the information content and ability to predict stock returns depends on company size and that large companies are priced more efficiently than small companies, suggesting that smaller companies would provide more opportunities for insiders to predict and to earn abnormal returns. Furthermore, it is argued that the information content of insiders' activities is coming from purchases and their ability to predict returns in smaller firms. On the other hand insider selling appears to have no predictive ability. However, while it is generally agreed that insiders are better informed and can make abnormal returns on trading on their company's shares it is still debatable whether outsiders can benefit by observing the behaviour of insiders. For example, Rozeff and Zaman (1988) and Seyhun (1986) showed that outsiders would not be able make abnormal return by using available information on large insider trading, while Bettis, Vickrey, and Vickrey (1997) provided evidence that they could benefit by imitating behaviour of top executives. Iqbal and Shetty (2002) examined the causality between insider transactions and stock returns and found that the causality is stronger from stock returns to insider transactions than insider transactions to stock returns. However, their findings at both aggregate and firm levels are at odds with other studies (Karpoff & Lee, 1991; Seyhun, 1988; Seyhun & Bradley, 1997) and the view that, on average, insiders profit from mispricing of firm's stock associated with macroeconomic and firm-specific information. They examined the association between stock returns and insider transactions at both aggregate and firm-specific levels and found a large negative impact of stock returns on subsequent insider transactions which suggest that insiders buy after stock price decreases and sell after stock price rise. Chowdhury et al. (1993) use a similar approach at the aggregate level. While trading at the aggregate level occurs due to macroeconomic factors, our analysis sheds light on insider trading associated with firm-specific information for a large sample of firms. In particular, the analysis below can provide evidence whether stock returns have a strong effect on subsequent insider transactions. 3. Data and methodology 3.1. Data Insider trading data in this paper has been complied from EDGAR Online, Insider Trades Data Feed, a new database on insiders' trading, covering the period from January 2000 to March 2007, a total of 87 calendar months. This period incorporates both a recessionary and growth period for the economy. As such the question arises as to whether the directors are knowledgeable enough about the economy in predicting the future prospects of their own firms and signalling this ability through the buying and selling of their own securities. The database contains more than 4 million daily transactions by insiders in over 9430 firms. The data used in this paper includes only purchases and sales that trade on the NYSE, AMEX and NASDAQ markets. The data were aggregated to the monthly frequency and following Lamba and Khan (1999), Iqbal and Shetty (2002) and Lakonishok and Lee (2001) we have included only firms with open market transactions of 100 shares or more. Similarly, following Conrad and Kaul (1993) and Lakonishok and Lee (2001) we have excluded shares whose stock prices were less than $2, non-common shares (shares with CRSP share codes other than 10 or 11), American Depository Receipts, closed-end funds, real estate investment trusts, convertible debt, exchange notes and options (purchase or sale of share through the exercise/conversion, warrants, or convertible bonds) in this study. Finally, firms with less than 12 months (not necessarily consecutive) transactions were also excluded. Insiders are classified into four groups: the Directors (including President and Chief Executive Officer, Chairman, Executive Vice President and Vice Chairman); Officers (senior management group including Chief Financial Officer, Chief Operating Officer and Controller); large shareholders, those who own more than 10% of shares and are not in management group; and Other groups, those who are required to report their trading to the SEC but are neither in management group nor are large shareholders (e.g. company lawyers). Following Seyhun (1988) throughout this paper, we have divided the sample firms into three sizes: small firms with a market value of less than $250 million, medium size firms with a market value of between $250 million and $1 billion and large firms with market values of greater than $1 billion. This classification also ensured that each size group has more than 1000 firms with transactions about 200,000, 364,703 and 650,619 in each group respectively. Finally, all the stock returns and corporate financial information came from DataStream. In particular, in addition to the stock return data, we consider a selection of variables that are believed to have predictive power for returns. This allows us to examine whether the insider trading information has predictive power over and above information that would be publicly available. This data set includes the dividend yield, price-earnings ratio, book-to-market ratio, the company's beta, the debt-to-equity ratio, the market returns given by the S&P500 index and a short-term interest rate. 3.2. Methodology Following Lakonishok and Lee (2001), the main insider variable of interest is the net purchase ratio, which is the ratio of net purchases to total insider trading activity. Moreover, following John and Lang (1991), Yur-Austin (1998), and Iqbal and Shetty (2002) we define the net purchase in different ways according to the number and volume of net purchases, while we also include a measure for the value of net purchases. In addition, we also include the purchase and sale sides only as measures of insider

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

257

activity. Hence, we constructed the following ratios to measure the degree of insider transactions per month: net number index (NNI), net share index (NSI), net value index (NVI), insider purchase index (PI) and insider sale index (SI): NNI NSI NVI PI SI P S PV SV PVA SVA

(P − S) / (P + S), this reflects trading activities from two directions; (PV − SV) / (PV + SV), this controls for the volume of transactions; (PVA − SVA) / (PVA + SVA), this reflects the value of transactions; P / (P + S), this measures insider purchases only; S / (P + S), this measures insider sales only; total number of insider purchase transactions in a given month, total number of insider sale transactions in a given month, total number of shares purchased by insiders in a given month, total number of shares sold by insiders in a given month, total value of shares purchased by insiders in a given month, total value of shares sold by insiders in a given month.

Separate examinations of purchases and sales are considered because it is argued (Chowdhury et al., 1993; Lakonishok & Lee, 2001) that insider purchases may have more information content than insider sales. In order to examine whether insider information has any predictive power for returns the basic equation we use is the standard predictive regression given by: r tþ1 ¼ α þ δIT t þ εtþ1

ð1Þ

where rt is the return on the stock associated with the insider information and ITt is the insider information variable as defined above. 4 If insider information contains predictive content for returns we would expect to see, given the definition of insider information as some form of net purchases, the δ coefficient both significant and positive. Of course, it is argued within the existing literature that a number of variables may have predictive power for returns, and thus, we wish to see if insider information has predictive power over and above any predictive power contained within publicly available information. Thus, we augment the equation in (1) with a variety of variables put forward within the literature that are argued to have predictive power. These include the dividend yield, the price-earnings ratio, the price-to-book ratio, the companies' beta, the equity to debt ratio and the movement of short-term interest rates. Thus, we estimate the following: r tþ1 ¼ α þ δIT t þ

k X

γ i xi;t þ εtþ1

ð2Þ

i

where the xi,t terms refer to alternative explanatory variables. Again, the key point of interest remains the significance of the δ coefficient. 4. Results 4.1. Summary statistics The Edgar insider trading data contains over 4 million transactions and over 9430 firms. The insider trading data was cleaned up using above filters and excluded transactions for which there was no exact trading or reporting dates. The insider data was then linked to DataStream for all the economic data used in this paper and excluded firms that did not have share price information. The cleaning up the dataset resulted in 3565 firms (issuers) with over 1.2 million transactions. Data were aggregated to generate monthly panel data and non-trading cases during a given month were set to zero. Table 1 shows the breakdown of the number of firms, transactions and mean value for small, medium and large companies. While the mean values of small and medium firms could be considered relatively close, the mean value of large companies is substantially large and different from the other two firm sizes. This has consequently been reflected in the number, volume and value of acquisitions and disposals. While the average values of shares traded as proportion of firms' market value are rather small, the average values of shares traded per firm per year ranges from $2 m in small firms to $47 m in large companies. Also the value of shares traded over the sample period ranged from over $21.4 billion to over $391.1 billion. Furthermore, over this period over 21.5 billion shares were traded by insiders. Tables 2 and 3 show the level of trading activity of various insider groups broken-down into small, medium and large companies over the sample period and in more recent years. Table 2 shows that directors are more active insider traders than officers and other groups. They tend to sell more than buy as observed in other studies (Fidrmuc et al., 2006). This may suggest that

4 In respect of returns we consider raw returns, excess (over the short-term Treasury bill) and abnormal returns (over the S&P500). The results are both qualitatively and quantitatively similar. The effect of using excess and abnormal returns is to remove the influence of cyclical economic-wide factors and to control for the belief that insiders may act in a contrarian fashion (for example, Lakonishok & Lee, 2001).

258

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Table 1 Stock, January 2000–March 2007.

1. Number of firms (issuers) 2. Number of traders (owners) 3. Mean firm value ($m) 4. Number of acquisitions 5. Number of disposals 6. Number of transactions 7. Volume of acquisitions (millions) 8. Value of acquisitions (million$) 9. Volume of disposal (millions) 10. Value of disposal (million$) 11 Number of shares traded (millions) 12. Average shares traded per firm per year 13. Total value of shares traded (million$) 14. Average value of shares traded per firm per year (million$) 15. Average values of shares traded as proportion of firms' market value (%)

Total

Small (b$250 m)

Medium ($250–999 m)

Large (≥$1000 m)

3565 49,387 3435 348,502 866,648 1,215,150 5338 80,643 16,188 423,460 21,526 832,836 504,104 20 0.57%

1402 14,980 100 83,803 116,025 199,828 873 6565 1468 14,893 2340 230,250 21,458 2 2.11%

1026 14,502 539 107,955 256,748 364,703 1498 15,552 3988 75,971 5486 737,542 91,523 12 2.28%

1137 19,905 10,162 156,744 493,875 650,619 2967 58,527 10,732 332,596 13,699 1,661,858 391,123 47 0.47%

directors might believe that their companies' stocks are over-valued or they are sold to meet liquidity needs. Over the sample period, on average directors traded in their companies' shares of over 1.14%, 1.32% and 0.23% of their respective market values in small, medium and large firms. However, relative to their market values, medium sized firms tend to have higher insider trading followed by smaller and larger firms. Table 3 shows ratios of average value of acquisitions and disposals per firm per year to average firm sizes disaggregated by insider traders over more recent years. The time series data suggest that trading over the period 2003–2006 has been relatively stable with disposals exceeding acquisitions in every year across each trader and firm size, with insider traders in medium sized companies slightly more active than other inside traders in other companies.

Table 2 Stocks transactions, volume and value broken down by firm size and type of trader, January 2000–March 2007. Directors

Officers

Others

Small Acquisitions Volume acquired (millions of shares) Value acquired ($million) Average value acquired as proportion of market value per year Disposals Volume disposed (millions of shares) Value disposed ($million’) Average value disposed of as proportion of market value per year Total market value of firms 2000–March 2007

50,144 483 3519 0.346% 68,760 718 8142 0.801% 1,016,450

21,946 120 879 0.086% 34,061 136 1950 0.192%

11,713 270 2167 0.213% 13,204 613 4801 0.472%

Medium Acquisitions Volume acquired (millions of shares) Value acquired ($million) Average value acquired as proportion of market value per year

48,280 724 7800 0.195%

40,704 337 3890 0.097%

18,971 437 3861 0.096%

Disposals Volume disposed (millions of shares) Value disposed ($million) Average value disposed of as proportion of market value per year Total market value of firms 2000–March 2007

149,396 2363 45,422 1.133% 4,009,351.5

87,577 397 9561 0.238%

19,775 1228 20,988 0.523%

Large Acquisitions Volume acquired (millions of shares) Value acquired ($million) Average value acquired as proportion of market value per year Disposals Volume disposed (millions of shares) Value disposed ($million) Average value disposed of as proportion of market value per year Total market value of firms 2000–March 2007

67,757 1528 28,630 0.034% 247,352 5343 164,312 0.196% 83,767,906.5

83,285 1138 22,749 0.027% 211,669 1361 54,608 0.065%

5702 301 7148 0.009% 34,854 4028 113,676 0.136%

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

259

Table 3 Stocks Ratio of average value of acquisitions and disposals per firm per year to average firm size of active firms. Directors

Small Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006 Medium Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006 Large Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006

Total value $m

Officers Ave value per active firm $m

Ratio ave value to ave firm size

Total value $m

Others Ave value per active firm $m

Ratio ave value to ave firm size

Total value $m

Ave value per active firm $m

Ratio ave value to ave firm size

Ave size of active firms $m

No. of active firms

100 100 100 101

1270 1372 1388 1356

387 1167 1070 716

0.30 0.85 0.77 0.53

0.30% 0.85% 0.77% 0.52%

104 220 241 264

0.08 0.16 0.17 0.19

0.08% 0.16% 0.17% 0.19%

333 387 675 656

0.26 0.28 0.49 0.48

0.26% 0.28% 0.49% 0.48%

100 100 100 101

1270 1372 1388 1356

885 2576 2209 2103

0.70 1.88 1.59 1.55

0.70% 1.89% 1.59% 1.53%

216 508 514 616

0.17 0.37 0.37 0.45

0.17% 0.37% 0.37% 0.45%

851 1109 1765 999

0.67 0.81 1.27 0.74

0.67% 0.81% 1.27% 0.73%

540 537 539 539

931 999 1022 1019

1569 2387 1828 1673

1.69 2.39 1.79 1.64

0.31% 0.44% 0.33% 0.30%

460 950 1086 1121

0.49 0.95 1.06 1.10

0.09% 0.18% 0.20% 0.20%

742 1378 662 823

0.80 1.38 0.65 0.81

0.15% 0.26% 0.12% 0.15%

540 537 539 539

931 999 1022 1019

6388 9961 17,113 10,021

6.86 9.97 16.75 9.83

1.27% 1.86% 3.11% 1.83%

1187 2188 2648 2882

1.28 2.19 2.59 2.83

0.24% 0.41% 0.48% 0.52%

6339 5962 4418 3434

6.81 5.97 4.32 3.37

1.26% 1.11% 0.80% 0.63%

10,406 10,298 10,174 10,183

1082 1115 1135 1134

3422 6402 7293 8777

3.16 5.74 6.43 7.74

0.03% 0.06% 0.06% 0.08%

2659 5701 6266 6212

2.46 5.11 5.52 5.48

0.02% 0.05% 0.05% 0.05%

1480 1955 493 2238

1.37 1.75 0.43 1.97

0.01% 0.02% 0.00% 0.02%

10,406 10,298 10,174 10,183

1082 1115 1135 1134

33,433 45,487 42,900 35,074

30.90 40.80 37.80 30.93

0.30% 0.40% 0.37% 0.30%

6336 13,190 15,873 15,289

5.86 11.83 13.99 13.48

0.06% 0.11% 0.14% 0.13%

28,901 41,629 22,655 17,257

26.71 37.34 19.96 15.22

0.26% 0.36% 0.20% 0.15%

Table 4 shows volume trading by insider traders over the period 2003–2006. It shows that insider traders in small, medium and large firms, traded in over 561 million, 1.3 billion and about 3.3 billion of their companies' shares per annum, respectively. The proportion of firms where insiders were either acquiring or selling ranged from 80%–97.5% and 70%–96.7% respectively. Table 5 shows the individual traders with at least one sell or buy transaction. It also shows that the number of traders has increased over the sample period across all traders and firm sizes. An interesting finding is that while the number of buyers is larger than sellers the volume of disposals is larger than acquisitions. Officers followed by Directors tend to dominate the trading in their companies' shares. On average officers has the largest number of traders, over 6470 officers per year reaching to over 11,736 officers trading in their companies' shares in 2006 in large companies compared to only 7366 traders in directors group in large firms. Traders in other groups are relatively small averaging just over 500 traders per year. Table 6 shows the number of firms with active insider trading, number of transactions, volume and value of shares acquired and disposed per year and firm sizes. The results suggest that over 92% of firms are active, i.e. there had insider trading. It also shows that there has been an upward trend in average transactions, volume and the value of the shares traded with disposals dominating acquisitions which could suggest that shares were considered over-valued by other insiders across all firms.

4.2. Results for All firms Table 7 reports our estimated results from the above equations, using NNI, NSI and NVI as the measures of insider activity. Panel A reports the results of the simple predictive regression, while Panel B reports the results from the regression that includes a variety of publicly available predictive regressors. Taking Panel A first, and looking across all categories of insider we can see a positive relationship between insider activity and future returns. This result makes intuitive sense, where net purchases are positive, insiders believe the future value of the firm will increase, moreover, if the market observes this positive net purchase then they will also increase their demand for the particular company's equity. Similarly, if net purchases are negative then insiders, and the market if it can observe the insider activity, will believe the future value of the firm will be lower, hence negative future

260

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Table 4 Stock acquisitions and disposals.

Directors Small 2003 2004 2005 2006 Medium 2003 2004 2005 2006 Large 2003 2004 2005 2006 Officers Small 2003 2004 2005 2006 Medium 2003 2004 2005 2006 Large 2003 2004 2005 2006 Others Small 2003 2004 2005 2006 Medium 2003 2004 2005 2006 Large 2003 2004 2005 2006 Totals Small 2003 2004 2005 2006 Medium 2003 2004 2005 2006 Large 2003 2004 2005 2006

% firms where insiders were acquiring

% firms where insiders were disposing

713 940 954 932

68.66% 82.07% 84.73% 84.07%

56.14% 68.51% 68.73% 68.73%

689 864 859 883

626 804 797 810

74.01% 86.49% 84.05% 86.65%

67.24% 80.48% 77.98% 79.49%

15.54% 23.80% 28.22% 49.92%

882 1026 1034 1032

787 966 985 991

81.52% 92.02% 91.10% 91.01%

72.74% 86.64% 86.78% 87.39%

17.95 38.46 33.65 38.97

71.30% 83.53% 97.93% 93.19%

564 850 869 881

496 699 700 715

44.41% 61.95% 62.61% 64.97%

39.06% 50.95% 50.43% 52.73%

45.47 86.79 96.82 89.66

59.28 102.94 107.07 105.95

76.71% 84.31% 90.43% 84.63%

616 798 792 813

612 762 795 828

66.17% 79.88% 77.50% 79.78%

65.74% 76.28% 77.79% 81.26%

1082 1115 1135 1134

157.09 294.61 308.03 302.41

180.25 353.95 387.59 351.43

87.15% 83.23% 79.47% 86.05%

873 1001 1027 1008

903 1016 1026 1036

80.68% 89.78% 90.48% 88.89%

83.46% 91.12% 90.40% 91.36%

1270 1372 1388 1356

51.39 55.12 75.61 70.28

135.80 152.01 215.77 98.30

37.84% 36.26% 35.04% 71.50%

137 206 218 204

174 252 208 190

10.79% 15.01% 15.71% 15.04%

13.70% 18.37% 14.99% 14.01%

931 999 1022 1019

47.82 123.64 80.34 151.72

394.86 337.14 237.56 156.99

12.11% 36.67% 33.82% 96.64%

95 127 122 129

159 186 163 147

10.20% 12.71% 11.94% 12.66%

17.08% 18.62% 15.95% 14.43%

1082 1115 1135 1134

56.86 78.62 18.11 109.52

983.39 1452.64 1007.86 458.33

5.78% 5.41% 1.80% 23.90%

98 120 116 96

147 188 166 153

9.06% 10.76% 10.22% 8.47%

13.59% 16.86% 14.63% 13.49%

1270 1372 1388 1356

112.26 233.46 257.01 217.55

249.33 410.05 478.23 286.85

45.02% 56.93% 53.74% 75.84%

1011 1238 1274 1248

884 1118 1116 1098

79.61% 90.23% 91.79% 92.04%

69.61% 81.49% 80.40% 80.97%

931 999 1022 1019

237.02 424.79 381.12 379.18

774.71 968.67 1370.95 683.17

30.60% 43.85% 27.80% 55.50%

808 947 951 970

763 903 933 934

86.79% 94.79% 93.05% 95.19%

81.95% 90.39% 91.29% 91.66%

1082 1115 1135 1134

421.77 724.48 739.00 853.22

2501.11 3282.22 2858.31 1693.85

16.86% 22.07% 25.85% 50.37%

1016 1087 1104 1099

990 1078 1087 1094

93.90% 97.49% 97.27% 96.91%

91.50% 96.68% 95.77% 96.47%

Active firms

Volume acquired (millions of shares)

1270 1372 1388 1356

48.08 146.21 148.45 110.95

931 999 1022 1019

Volume disposed of (millions of shares)

Ratio acquisitions to disposals

Firms where insiders were acquiring

95.58 219.58 228.81 149.58

50.30% 66.59% 64.88% 74.17%

872 1126 1176 1140

143.73 214.37 203.96 137.80

320.58 528.59 1026.32 420.22

44.83% 40.55% 19.87% 32.79%

1082 1115 1135 1134

207.83 351.25 412.85 441.29

1337.47 1475.63 1462.85 884.08

1270 1372 1388 1356

12.80 32.13 32.95 36.32

931 999 1022 1019

Firms where insiders were disposing

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

261

Table 5 Stocks Individual buyers and sellers. Directors

Small 2003 2004 2005 2006 Medium 2003 2004 2005 2006 Large 2003 2004 2005 2006

Officers

Others

Totals Sellers

Ratio purchasers to sellers

3988 6568 6909 6799

2760 4179 3972 4043

144% 157% 174% 168%

48% 62% 72% 70%

4142 6658 6725 6843

3743 5370 5391 5682

111% 124% 125% 120%

58% 62% 72% 53%

6816 10,296 10,337 9940

6316 9083 9337 9665

108% 113% 111% 103%

Purchasers

Sellers

Ratio purchasers to sellers

Purchasers

Sellers

Ratio purchasers to sellers

Purchasers

Sellers

Ratio purchasers to sellers

2570 4162 4389 4271

1425 2152 2058 2078

180% 193% 213% 206%

1240 2092 2206 2240

1055 1645 1609 1698

118% 127% 137% 132%

178 314 314 288

280 382 305 267

64% 82% 103% 108%

2137 3465 3479 3574

1597 2245 2171 2223

134% 154% 160% 161%

1844 2977 3044 3066

1811 2775 2939 3169

102% 107% 104% 97%

161 216 202 203

335 350 281 290

2987 4532 4479 4398

2050 2956 2909 2968

146% 153% 154% 148%

3643 5514 5603 5367

3945 5726 6073 6369

92% 96% 92% 84%

186 250 255 175

321 401 355 328

Purchasers

Note: purchasers: individual traders with at least one acquisition; sellers: individual traders with at least one disposal. The same individual may be both purchaser and seller.

returns. Looking more specifically, we can see that the insider activity of directors and officers appear to have predictive power for future returns, with a positive and significant coefficient on each of the measures of insider behaviour. Panel B reports the results using the various insider net purchasing ratios including a selection of alternative variables. This allows us to examine whether there is information content from insider activity over and above that contained within publicly available information. Moreover, given the large body of evidence that states such variables are important, then, it could be argued that the first regression suffers from omitted variable bias. These results, however, largely support those in Panel A, suggesting predictive power from the activities of directors and officers and with coefficients of similar magnitude. However, these results now suggest that there is also predictive power rise from the insider activities of other groups (such as company lawyers). Of further interest we can see that the coefficient magnitudes for directors and officers are of similar magnitude in the regressions in Panel B, although, in Panel A this suggests that the insider behaviour of officers has a greater impact on future returns. Furthermore, the results in Panel B suggest that the impact of other groups is even stronger, with a large impact, although it should be borne in mind that this is not significant in the first regression. These results suggest that the categories of insiders that have predictive power for returns include both company presidents, vice-presidents, chief executive officers and senior management. Whereas the actions of large shareholders and other groups with vested interests, such as company lawyers do not have such predictive power. Thus, predictive power lies with those who have more information and thus a direct impact on company strategy and behaviour. In sum, these results suggest that first, insider trading activity does have predictive power over and above sources of publicly available information which questions the strong-form market efficiency. Second, if the market is tracking the behaviour of insiders, then the activities of directors and senior management are the most important in terms of mimicking their net purchasing behaviour, although there is some evidence that the actions of the other group also has predictive power and therefore of use to the market.

4.3. Results by firm size Panel C of Table 7 reports the results of the predictive regressions estimated according to firm size, where size is determined by the criteria whereby small firms have a market value of less than $250 million, medium size firms have a market value of between $250 million and $1 billion and large firms have a market value of greater than $1 billion. The regression includes all the predictive variables as in Panel B and Eq. (2); however, they are not reported but are available. Here we can observe a different pattern of behaviour than that taken from the above analysis where all firms are included. Taking the insider actions of directors first, we can see that across the three definitions of insider behaviour, related to the number of transactions, volume of transactions and the value of transactions, only for medium size firms do all measures suggest that directors insider dealings have predictive power for future returns. With regard to small and large firms, the number of directors' transactions has no predictive power, however, both the volume of shares transacted and the value transacted does have predictive power, indicating that it is not whether directors trade but the amounts they trade that is important. With regard to officers, which was significant for all firms, here we can see that the insider trading only has a significant impact on the future returns of small firms, although there is a marginal effect on medium and large firms according to the value

262

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Table 6 Stocks (average) transactions, volume and value acquired and disposed per year and firm size. Trans-actions

Small Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006 Medium Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006 Large Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006 Totals Acquired 2003 2004 2005 2006 Disposed 2003 2004 2005 2006

Volume (million shares)

Value ($m)

Active firms

Ave transactions per active firm

Ave volume per active firm (million shares)

Ave value per active firm ($m)

11,198 20,456 22,963 24,098

112.26 233.46 257.01 217.55

823.69 1773.50 1985.95 1636.00

1270 1372 1388 1356

8.82 14.91 16.54 17.77

0.09 0.17 0.19 0.16

0.65 1.29 1.43 1.21

15,791 30,380 29,977 32,422

249.33 410.05 478.23 286.85

1951.99 4192.35 4487.53 3718.32

1270 1372 1388 1356

12.43 22.14 21.60 23.91

0.20 0.30 0.34 0.21

1.54 3.06 3.23 2.74

15,221 28,588 31,082 27,523

237.02 424.93 381.12 379.18

2771.77 4716.27 3576.52 3617.17

931 999 1022 1019

16.35 28.62 30.41 27.01

0.25 0.43 0.37 0.37

2.98 4.72 3.50 3.55

35,785 59,439 63,066 82,552

774.71 968.70 1370.95 683.17

13914.17 18111.66 24178.94 16337.08

931 999 1022 1019

38.44 59.50 61.71 81.01

0.83 0.97 1.34 0.67

14.95 18.13 23.66 16.03

20,732 42,542 42,218 41,229

421.77 726.04 739.00 853.22

7560.21 14081.21 14052.73 17227.01

1082 1115 1135 1134

19.16 38.15 37.20 36.36

0.39 0.65 0.65 0.75

6.99 12.63 12.38 15.19

55,426 100,636 130,108 160,776

2501.11 3283.76 2858.31 1693.85

68670.13 100386.04 81428.50 67619.33

1082 1115 1135 1134

51.23 90.26 114.63 141.78

2.31 2.95 2.52 1.49

63.47 90.03 71.74 59.63

47,151 91,586 96,263 92,850

771.05 1384.42 1377.13 1449.95

11155.67 20570.99 19615.20 22480.18

3283 3486 3545 3509

14.36 26.27 27.15 26.46

0.23 0.40 0.39 0.41

3.40 5.90 5.53 6.41

107,002 190,455 223,151 275,750

3525.15 4662.52 4707.49 2663.86

84536.30 122690.05 110094.97 87674.73

3283 3486 3545 3509

32.59 54.63 62.95 78.58

1.07 1.34 1.33 0.76

25.75 35.20 31.06 24.99

measure of insider activity. Perhaps it could be argued that senior management can have a more immediate impact on the future performance of smaller firms. With regard to large shareholders we see no significant impact and indeed the coefficient is of the wrong sign. However, with the other insider group there is some evidence to suggest that the value of this group trades does have positive and significant predictive power for future returns. Overall, these results suggest that the predictive power of insider activities differs not only according to the type of insider as the above analysis suggested, but also according to the size of the company. Moreover, with evidence supporting the view that it is not whether insider trades that is important as much as the volume or value of the trading. Furthermore, taking a broad view, these results suggest that while directors have predictive power across all firm sizes, officers only have predictive power for small firms. 4.4. Separating buy and sell The above results have focussed upon examining the impact of insiders' actions irrespective of whether their actions are a buy or sell on returns. In this section, we briefly examine whether there is a difference in the impact on returns of insider behaviour if

Table 7 Predictive Regression. Directors NNI

Officers NSI

Panel δ DY PE PB β DE MR IR

Other

NVI

NNI

NSI

NVI

NNI

NSI

NVI

NNI

NSI

NVI

0.30 (4.56)

0.35 (3.95)

0.36 (4.08)

0.43 (4.76)

0.35 (1.02)

0.33 (0.98)

0.36 (1.05)

0.27 (1.18)

0.28 (1.19)

0.34 (1.46)

0.21 − 0.10 − 0.02 − 0.01 0.90 0.01 0.03 − 0.26

0.24 − 0.10 − 0.02 − 0.01 0.91 0.01 0.03 − 0.26

0.32 − 0.11 − 0.02 − 0.01 0.93 0.01 0.03 − 0.26

− 0.10 − 0.33 − 0.01 − 0.11 1.26 0.01 0.12 − 0.66

− 0.07 − 0.33 − 0.01 − 0.11 1.26 0.01 0.12 − 0.66

− 0.05 − 0.33 −0.01 − 0.11 1.26 0.01 0.12 − 0.66

0.44 (1.81) − 0.15 (− 1.2) − 0.01 (− 1.9) − 0.08 (− 2.4) 1.79 (3.20) 0.01 (1.34) − 0.10 (− 1.1) − 0.41 (− 3.2)

0.54 − 0.15 − 0.01 − 0.08 1.81 0.01 − 0.10 − 0.41

0.59 − 0.16 − 0.01 − 0.08 1.86 0.01 − 0.10 − 0.41

B. Predictive regression with other variables 0.22 (3.50) 0.24 (3.79) 0.27 (4.47) − 0.09 (− 2.9) − 0.09 (− 2.9) − 0.10 (− 3.1) − 0.01 (0.00) − 0.01 (0.00) − 0.01 (0.00) − 0.01 (− 1.4) − 0.01 (− 1.4) − 0.01 (− 1.4) 0.86 (6.10) 0.86 (6.11) 0.88 (6.23) 0.03 (0.85) 0.03 (0.85) 0.01 (0.80) 0.05 (2.12) 0.05 (2.13) 0.05 (2.16) − 0.28 (− 8.3) − 0.28 (− 8.3) − 0.28 (− 8.2)

Panel C. Predictive regression with other variables by size SM 0.21 (1.71) 0.23 (1.88) 0.26 (2.14) ME 0.26 (2.05) 0.24 (1.91) 0.27 (2.13) LG 0.14 (1.75) 0.20 (2.37) 0.23 (2.89)

(2.78) (− 3.0) (− 3.3) (− 0.9) (5.96) (0.30) (1.18) (− 7.4)

0.40 (2.37) 0.23 (1.46) 0.03 (0.33)

(3.08) (− 3.1) (− 3.3) (− 1.0) (5.98) (0.33) (1.18) (− 7.4)

0.40 (2.34) 0.27 (1.66) 0.08 (0.78)

(4.29) (− 3.3) (− 3.3) (− 0.9) (6.14) (0.24) (1.24) (− 7.3)

0.51 (3.08) 0.29 (1.83) 0.18 (1.78)

(− 0.3) (− 1.7) (− 2.0) (− 1.9) (1.26) (1.81) (0.76) (− 3.0)

− 0.12 (− 0.2) − 0.49 (− 0.9) − 0.06 (− 0.1)

(− 0.2) (− 1.7) (− 2.0) (− 1.9) (1.25) (1.81) (0.76) (− 3.0)

− 0.10 (− 0.2) − 0.45 (− 0.8) − 0.04 (− 0.1)

(− 0.1) (− 1.7) (− 2.0) (− 1.9) (1.26) (1.80) (0.76) (− 3.0)

− 0.11 (− 0.2) − 0.43 (− 0.8) 0.06 (0.12)

0.68 (1.47) 0.07 (0.15) 0.51 (1.55)

(2.23) (− 1.2) (− 1.9) (− 2.4) (3.24) (1.35) (− 1.1) (− 3.2)

0.76 (1.64) 0.30 (0.60) 0.56 (1.69)

(2.42) (− 1.3) (− 1.9) (− 2.4) (3.31) (1.34) (− 1.1) (− 3.2)

0.87 (1.90) 0.24 (0.48) 0.60 (1.78)

Notes: Basic equation given by: r tþ1 ¼ α þ δIT t þ

k X

γi xi;t þ εtþ1

i

where the xi,t terms refer to explanatory variables, such as the dividend yield, the price-earnings ratio, the price-to-book ratio, the companies' beta, the equity to debt ratio and the movement of short-term interest rates. Again, the key point of interest remains the significance of the δ coefficient. Numbers in () are t statistics and: IT = measure of insider trading. DY = dividend yield. PE = price/earnings ratio. PB = price/book ratio. Β = companies beta. DE = debt/equity ratio. MR = market return (S&P 500). IR = 3-month interest rate. SM = small companies. ME = medium companies. LG = large companies.

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Panel A. Predictive regression δ 0.23 (3.42) 0.25 (3.70)

Large shareholders

263

264

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Table 8 Buy/Sell Signals. Director δ δ

Buy — NNI > 0 0.75 (8.68) Sell — NNI b 0 0.52 (2.87)

Officer

Large Shareholders

Other

0.57 (2.39)

− 3.73 (− 2.12)

0.09 (0.11)

0.44 (2.27)

1.37 (0.81)

− 1.04 (− 1.15)

See Table 7 for general equation specification.

the insiders are buying or selling. In order to do this we follow Lakonishok and Lee (2001), and separate NNI into positive and negative values. The effect of doing this and estimating the Eq. (1) is reported in Table 8, we only report the results for the insider parameter, δ, for brevity. The results reported here remain broadly consistent with the previous sets of results. In particular, when NNI is positive, that is, insiders are buying more than they sell, then there is a positive and significant relationship with returns for both directors and Officers. Thus, if these two groups of insiders are buying then future returns increase. Of interest, there is a negative and significant relationship between returns and the net buying behaviour of large shareholders, which seems contradictory. The coefficient for the other group is statistically insignificant. With respect to when NNI is negative, again, the results for directors and officers are intuitive. That is, a positive and significant coefficient indicates when these groups of insider engage in net selling then future returns fall. The coefficients for large shareholders and other are insignificant. Of further note, the coefficient, or response of stock returns to a change in insider trading, is larger for buys than for sells (especially for directors). This is, also consistent with the view that buy signals carry more information with regard to future firm performance, whereas sells may arise for reasons other than the future performance of the firm.

4.5. Why do insiders trade? In this last section, we look to see if any of the variables are important in determining whether insiders decide to trade or not. In particular, we estimate a regression similar to that above but instead of having returns as the dependent variable, we have the number of transactions undertaken by an insider. Moreover, we also include as explanatory variables lags of other insiders' number of transactions in order to determine whether insiders all trade together and follow each other's behaviour. The results from these regressions are reported in Table 9. From this table it can be immediately seen that while there is a lot of interaction between different types of insider groups, there is no evidence that insiders' actions are determined by general market factors, such as past market and individual firm returns, individual firm ratios and interest rates. More specifically, in the equation for directors, the lag of both officers and large shareholders is significant, although the lag of the other group is not significant. For officers, the lag of directors is significant, but not large shareholders or other. For large shareholders, the lag of directors is significant but not the lag of officers or other. Finally, for other, the lag of directors is significant but not for officers, while the lag on large shareholders is significant, however, the coefficient is of the wrong sign, indicating an increase in net purchase transactions by large shareholders leads to a fall in net purchase transactions by the other group. Overall, the results from this section suggest that insiders tend to trade at the same time, with the actions of directors affecting the trading behaviour of all other insider types, while both officers and large shareholders have an effect on some other groups of insider.

Table 9 What causes insiders to trade.

Returns DY PE PB β DE MR IR Directors Officers Large shareholders Other

Directors

Officers

Large shareholders

Other

− 0.01 (− 1.78) 0.08 (1.76) − 0.01 (− 1.14) − 0.02 (− 0.88) − 0.33 (− 1.45) 0.01 (1.51) 0.04 (1.44) 0.03 (0.65) – 0.23 (2.70) 0.19 (2.44) 0.05 (0.69)

− 0.01 (− 1.10) − 0.06 (− 1.30) − 0.01 (− 1.67) 0.01 (0.38) − 0.01 (− 0.08) − 0.01 (− 0.35) 0.03 (1.02) − 0.03 (− 0.70) 0.29 (3.81) – 0.02 (0.54) 0.01 (0.11)

− 0.01 (− 1.57) 0.01 (0.08) − 0.01 (− 1.01) − 0.01 (− 1.16) 0.38 (1.82) − 0.01 (− 0.11) 0.02 (0.72) 0.01 (0.34) 0.18 (2.07) 0.10 (0.99) – 0.16 (0.85)

− 0.01 (− 1.50) − 0.05 (− 1.11) 0.01 (0.90) − 0.01 (− 0.78) − 0.27 (− 1.56) − 0.01 (− 0.95) − 0.01 (− 0.38) − 0.05 (− 1.46) 0.18 (2.54) − 0.10 (− 1.15) − 0.55 (− 2.94) –

See Table 7 for general equation specification.

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

265

5. Summary and conclusion This paper seeks to analyse whether insiders have information that could lead to profitable trading. Using a new dataset of insider activities, we examine whether insider trading has information content for stock returns. In particular, we seek to contribute to the literature in four ways. First, we use an extended dataset covering both ball and bear markets and economic expansions and contractions. Second, we split the management team into directors and senior management to examine for a hierarchy of information and interaction. Third, we see whether the results differ by firm size. Fourth, we examine whether insiders' trading behaviours are affected by other insider groups or past returns. Our results suggest the following conclusions. First, looking across all firms and insider behaviour as determined by three alternate measures of insider activity (number of transactions, volume of transactions and the value of transactions), our estimated results are generally supportive of a positive relationship between insider activity and future returns. This is consistent with our intuitive belief, where net purchases are positive (negative), such that insiders believe that the future value of the firm will increase, then future returns are also positive (negative). Second, consistent with the hypothesis that those engaged in decision-making will have the greatest signalling effect, significant predictive power for future returns arises from the actions of directors and officers only. These results are robust to the inclusion of alternative, publicly available information, such as the dividend yield, price-earnings ratio and a variety of other relevant variables. Third, having established that directors and officers have significant explanatory power across firms and insider trades, we break down our analysis by firm size and buy/sell. Here the results suggest that for all firm sizes directors still have predictive power, while for officers the predictive effect of insider activity only occurs for small firms. This latter result is consistent with previously reported results, although the former is different and suggests that the actions of executives can act as a signal regardless of firm size. With regard to the influence of net buy and sell behaviour, our results demonstrate that both directors and officers have the correct sign and are statistically significant. That is, a positive coefficient indicating that net buys lead to increased future returns and net sell leads to decreased future returns. Although, the response of returns is great for a buy, suggesting the signal from buys is stronger than the signal from sells. This is consistent with the view that while sells may signal the insiders' expectations of future performance, sells may also occur for other reasons, such as liquidity. Fourth, we consider is what factors affect the decision of insiders to trade. To this end, we estimate a regression where the dependent variable is the number of transactions undertaken by an insider, with all remaining variables, including a lag of other insiders' number of transactions, as explanatory variables. These results suggest that insider tends to trade at the same time, with the actions of directors affecting the trading behaviour of other insider types, while both officers and large shareholders have an effect on some other groups of insider. In sum, our results support the view that director and officer actions do have positive predictive power for future returns, and that these insiders and any outsider who was able to mimic their behaviour, could make positive trading returns. In addition, while director actions have predictive power for firm of all sizes, officers only have predictive power for small firms. The signal emanating from buys is stronger than the signal emanating from sells. Finally, the trading actions of directors, and to a lesser extent, officers have significant effects on the trading behaviour of other groups of insiders. Acknowledgement We are most grateful to the anonymous referees for their most helpful comments and suggestions. References Bettis, C., Vickrey, D., & Vickrey, D. W. (1997). Mimickers of corporate insiders who make large volume trades. Financial Analysts Journal, 53, 57–66. Carter, M. L., Mansi, S. A., & Reeb, D. M. (2003). Quasi-private information and insider trading. Financial Analysts Journal, 59, 60–68. Chakravarty, S., & McConnell, J. J. (1999). Does insider trading really move stock prices? Journal of Financial and Quantitative Analysis, 34, 191–209. Cheng, L. T. W., Davidson, W. N., & Leung, T. Y. (2011). Insider trading returns and dividend signals. International Review of Economics and Finance, 20, 421–429. Chowdhury, M., Howe, J. S., & Lin, J. -C. (1993). The relation between aggregate insider transactions and stock market returns. Journal of Financial and Quantitative Analysis, 28, 431–437. Conrad, J., & Kaul, G. (1993). Long-term market overreaction or biases in computed returns. Journal of Finance, 48, 39–63. Eckbo, B. E., & Smith, D. C. (1998). The conditional performance of insider trades. Journal of Finance, 53, 467–498. Fidrmuc, J., Goergen, M., & Renneboog, L. (2006). Insider trading, news releases and ownership concentration. Journal of Finance, 61, 2931–2973. Finnerty, J. (1976). Insiders and market efficiency. Journal of Finance, 31, 1141–1148. Givoly, D., & Palmon, D. (1985). Insider trading and the exploitation of inside information: Some empirical evidence. Journal of Business, 58, 69–87. Hotson, L., Singh, H., & Singh, N. (2008). The information content of directors' trades: Empirical analysis of the Australian market. Investment Management and Financial Innovations, 5, 122–133. Ikenberry, D., Lakonishok, J., & Vermaelen, T. (1995). Market underreaction to open market share repurchases. Journal of Financial Economics, 39, 181–208. Iqbal, Z., & Shetty, S. (2002). An investigation of causality between insider transactions and stock returns. The Quarterly Review of Economics and Finance, 42, 41–57. Jaffe, J. (1974). Special information and insider trading. Journal of Business, 47, 410–428. Jeng, L., Metrick, A., & Zeckhauser, R. (2003). Estimating the returns to insider trading: A performance-evaluation perspective. The Review of Economics and Statistics, 85, 453–471. John, K., & Lang, L. (1991). Strategic insider trading around dividend announcements: Theory and evidence. Journal of Finance, 46, 1361–1389. Kaniel, R., Saar, G., & Titman, S. (2008). Individual investor trading and stock returns. Journal of Finance, 63, 273–310. Karpoff, J. M., & Lee, D. (1991). Insider trading before new issue announcements. Financial Management, 20, 18–26. Kumar, A. (2009). Who gambles in the stock market? Journal of Finance, 64, 1889–1933. Kyle, A. S. (1985). Continuous auctions and insider trading. Econometrica, 53, 1315–1336.

266

M. Tavakoli et al. / International Review of Economics and Finance 22 (2012) 254–266

Lakonishok, J., & Lee, I. (2001). Are insider trades informative? Review of Financial Studies, 14, 79–111. Lamba, S. A., & Khan, W. A. (1999). Exchange listings and delistings: The role of insider information and insider trading. Journal of Financial Research, 22, 131–146. Lee, D. S., Mikkelson, W. H., & Partch, M. M. (1992). Managers' trading around stock repurchases. Journal of Finance, 47, 1947–1961. Lin, J., & Howe, J. (1990). Insider trading in the OTC market. Journal of Finance, 45, 1273–1284. Lorie, H. H., & Niederhoffer, V. (1968). Predictive and statistical properties of insider trading. Journal of Law and Economics, 11, 35–51. Ma, Y., Sun, H. -L., & Tang, A. P. (2009). Do insiders have inside tracks: An examination of Wall Street Journal's Inside Track columns? International Review of Economics and Finance, 18, 520–530. Penman, S. H. (1982). Insider trading and the dissemination of firms' forecast information. Journal of Business, 55, 479–504. Rozeff, M. S., & Zaman, M. A. (1988). Market efficiency and insider trading: New evidence. Journal of Business, 61, 25–44. Rozeff, M. S., & Zaman, M. A. (1998). Overreaction and insider trading: Evidence from growth and value portfolios. Journal of Finance, 53, 701–716. Seyhun, H. N. (1986). Insiders' profits, costs of trading, and market efficiency. Journal of Financial Economics, 16, 189–212. Seyhun, H. N. (1988). The information content of aggregate insider trading. Journal of Business, 61, 1–24. Seyhun, H. N. (1990). Do bidder managers knowingly pay too much? Journal of Business, 63, 439–464. Seyhun, H. N. (1992). Why does aggregate insider trading predict future stock return? Quarterly Journal of Economics, 107, 1303–1331. Seyhun, H. N., & Bradley, M. (1997). Corporate bankruptcy and insider trading. Journal of Business, 70, 189–216. Yur-Austin, J. (1998). Can insiders bail themselves out before private renegotiation? Review of Financial Economics, 7, 197–211.