Insider Trading Disclosure, Information Asymmetry, and Differential Earnings Relevance as Indicated By Trading Volume

Electronic copy of this paper is available at: http://ssrn.com/abstract=983365

INSIDER TRADING DISCLOSURE, INFORMATION ASYMMETRY, AND DIFFERENTIAL EARNINGS RELEVANCE

AS INDICATED BY TRADING VOLUME

Don Vickrey, Arizona State University West Taylor Foster III, New Mexico State University

Donn Vickrey, Gradient Analytics Carr Bettis, Gradient Analytics

ABSTRACT

We use buy and sell signals derived from insider trading disclosures to identify experimental-group cases in which decreases in earnings-related predisclosure information asymmetry have arisen through the use of the disclosures. For each experimental firm, each earnings-announce-ment date is preceded, within 90 days, by at least 2 specific buy or 2 specific sell signals, while none of these signals occurs within 90 days of any control group earnings-announcement date. Our null hypothesis is that mean differences in experimental and control group earnings-announcement-period trading volumes are greater than or equal to similar mean differences occurring over contiguous non-announcement periods. Rejection of our null, for various periods, implies less earnings relevance for our experimental group vis-à-vis our control group as a consequence of decreases in earnings-related predisclosure information asymmetry for the latter. This result is consistent with the view that changes in firms’ information environments are the most pervasive factor explaining secular changes in earnings relevance.

INTRODUCTION

In recent years, conflicting evidence has

been advanced concerning the issue of

whether earnings relevance has increased or decreased over time. For example, the evidence of Amir and Lev (1996), Collins et al. (1997), Francis and Schipper (1999), Lev and Zarowin (1999), Ely and Waymire (1999), Ryan and Zarowin (2003), and Dechow and Strand (2004) tends to imply decreasing earnings relevance. In contrast, the evidence of Brown et al. (1999), Kross and Kim (2000), Landsman and Maydew (2002), Francis et al. (2002a), Francis et al. (2002b), and Beaver et al. (2005) tends to imply the opposite in the context of primarily larger firms (see Francis et al. 2002b) and bankruptcy prediction (Beaver et al. 2005).

Various explanations have been provided for the apparent decreases/ increases in earnings relevance. Sinha and Watts (2001) argue that decreasing earnings relevance (however measured) is a consequence of increases in the amount of information available from both extra-firm and intra-firm sources. Ryan and Zarowin (2003) conclude that the primary factors affecting decreasing earnings relevance (as measured by the contemporaneous linear relationship between annual stock returns and earnings) are that (1) earnings increasingly reflect news with lags relative to stock prices and (2) earnings reflect current positive price changes/negative price changes less/more strongly over time. They also conclude that their results are primarily

Electronic copy of this paper is available at: http://ssrn.com/abstract=983365

attributable to accounting reasons (e.g., the availability of more relevant information, conservatism in setting accounting standards, and managerial reporting behavior) and not economic reasons. In contrast, Francis et al. (2002b) conclude that the increasing number and types of disclosures made concurrently with earnings announcements (especially detailed income statements) account for increasing earnings relevance (as measured by absolute market reactions).

Our view is that each of the factors discussed above, as well as others, are likely to bear on the issue of changing earnings relevance, but that the jury still is out on whether earnings relevance has decreased or increased secularly. Additionally, we believe, generally speaking, that the most pervasive factor explaining the apparent change(s) in earnings relevance can be characterized simply as the changing information environments of firms (i.e., the frequent changes in both the types of available information and the extents of distribution of the varying types). In concept, such changes frequently, but not always, can be described as leading to decreases or increases in predisclosure information asymmetry with respect to earnings.

Our study adds to the dialogue by investigating a factor that is consistent with the above characterizations—reductions in earnings-related information asymmetry occasioned by SEC-mandated insider-trading disclosures. We emphasize that both the number of SEC-mandated insider-trading disclosures and the availability of these disclosures have increased greatly over time (see, e.g., Vickrey et al. (2003) and the discussion in the third section of this article).

Thus, we study the relationships among SEC-mandated insider-trading disclosures, information asymmetry, earnings-announce-ment trading volume, and relative earnings relevance. We limit the term, insider, to corporate officers and directors who must

report their trades under Section 16 of the Securities and Exchange Act of 1934.

The SEC-mandated disclosures have the potential for reducing information asymmetry by communicating insider information about firms’ prospects (see below). In this regard, various studies support the models that predict that prior disclosures lead to lower levels of earnings- related predisclosure information asymmetry and ultimately to lower earnings-announcement-period trading volume. For example, Atiase and Bamber (1994), Lobo and Tung (1997), Utama and Cready (1997), and Bettis et al. 2002 yield evidence supporting the validity of the trading models of Kim and Verrecchia (1991a) and He and Wang (1995). These studies also support the models of Kim and Verrecchia (1991b), (1994), and (1997) under conditions where increases in information asymmetry are precluded.

Specifically, we control for firm size and earnings-announcement abnormal return and predict that our experimental group’s earnings-announcement trading volume is lower than that of our control group because of systematic reductions in experimental-group information asymmetry occasioned by the SEC-mandated disclosures. Our null hypothesis is that mean differences in experimental group and control group earnings-announcement-period trading vol-ume are greater than or equal to similar mean differences occurring over contiguous non-announcement periods.

Our findings imply rejection of our null. The full implications of its rejection are discussed in the final section. For now, we note first that our findings ultimately imply lower/greater relative earnings relevance for our experimental/control group. Second, they are consistent with the view that changes in firms’ information environments are the most pervasive factor explaining secular changes in earnings relevance. Note that our results are only consistent with this view since we do not demonstrate that temporal changes in the information

environment of our experimental group relative to that of our control group have followed from reductions in earnings-related information asymmetry occasioned by the SEC-mandated disclosures. We do not take this step because of the formidable difficulties involved in controlling temporally for both information environment and the insider-trading signals described below and, at the same time, matching on firm size and abnormal return.

The remainder of the article is organized as follows. The next section discusses additional pertinent research and explains our method for inferring reduced levels of information asymmetry. The third discusses remaining design and methodological issues. The fourth presents our results. The last section contains final observations.

ADDITIONAL PERTINENT RESEARCH, TRADING SIGNALS,

AND INFORMATION ASYMMETRY,

To date, four studies focus specifically on the relationship between trading volume and earnings relevance (Beaver 1968; Kiger 1972; Bamber 1986; Landsman and Maydew 2002). In this regard, volume effects are widely viewed as mirroring changes in individual expectations and related changes in portfolio holdings (see, e.g., Beaver 1968; Hakasson et al. 1982; Karpoff 1986; Varian 1986; Kim and Verrrecchia 1991a, 1991b, 1994, 1997; Dontoh and Ronen 1993; Kross et al. 1994). Given this view, we are justified in focusing on such effects as indicators of the comparative earnings relevance of our experimental and control groups given reductions in information asymmetry for the former.

Our study requires a medium for identifying firms such that reductions in information asymmetry are likely to have arisen through the use of the SEC-mandated disclosures. The identified firms, for a given year, constitute our experimental group for

that year. We use two types of "buy" signals and two types of “sell” signals to identify the experimental-group firms. The signals utilized are large-volume purchases, large volume sales, purchase clusters, and sales clusters. Lancer Analytics (2002) uses different names to characterize these signals.

We derive our signals, and their absence, using the Lancer Analytics (2002) Insider Trading Model. Lancer Analytics, which is a product line of Thompson Financial, provides, among other things, a weekly data feed of company rankings derived from insider trading data via its model. We are grateful to Lancer Analytics for allowing us to utilize their data base in conducting this study.

In this model, (a) a large-volume purchase (sale) is defined as a single purchase (sale) of more than 6,300 (17,500) shares and (b) a purchase (sale) cluster is defined as the purchase (sale) of shares by two or more insiders within a 2-week period with no insider selling (purchasing) within the same period. Our reasons for using these signals are as follows.

Insiders undoubtedly possess private information about the earnings prospects of their firms, and much research suggests they can use this information (legally or illegally) to earn abnormal returns by trading in their firms' securities (e.g., Lorie and Niederhoffer 1968; Pratt and Devere 1970; Jaffe 1974a, 1974b; Finnerty 1976a, 1976b; Penman 1985; Givoly and Palmon 1985; Seyhun 1986, 1992; 1998; Fowler and Rorke 1988; Rozeff and Zaman 1988; Lin and Howe 1990; Moss and Kohers 1990; Pope et al. 1990; Bettis and Coles 1997; Bettis et al. 1997; Noe 1999; Jeng et al. 1999; Lanonishok and Lee 2001; Ke et al. 2003; Frankel and Li 2004).

Various works also confirm that outsiders (i.e., non-insider investors and intermediaries) are potentially able to analyze buying or selling related insider trading data and to use the results in successfully garnering abnormal returns for themselves or their clients (e.g., Lorie and

Niederhoffer 1968; Pratt and Devere 1970; Jaffee 1974b; Givoly and Palmon 1985; Rozeff and Zaman 1988; Dorfman 1989; Pope et al. 1990; McCarroll 1992; Bettis 1992; Gabele 1993; and Bettis et al. 1997). Consistently, prior research by Lancer Analytics (2002) implies that our buy/sell signals are pertinent in this context (i.e., in identifying value-relevant and, thus, earnings-related SEC-mandated insider-trading disclosures). Thus, we use these signals to identify cases in which decreases in earnings-related experimental-group predisclosure information asymmetry are likely to have arisen through the use of the SEC disclosures.

We emphasize that the Lancer Analytics data, and similar data, are widely used by those seeking to garner abnormal returns for themselves or their clients. For example, approximately 100 portfolio managers currently obtain buy and sell signals from Lancer Analytics for a fee. Additionally, various firms (e.g., Thompson Financial, Muzea Consulting, Modern Portfolio Theory, Vicker’s Stock Research, and the Washington Service) specialize in either analyzing insider-trading activities or providing related data. For example, Thompson Financial and Vicker’s Stock Research provide information to different types of subscribers including money managers and analysts. According to Craig Columbus (Vice President of Investment Strategy at Thompson Financial), his firm has used both large-volume transaction and purchase/sale clusters since 1985 to identify significant insider activity.

REMAINING DESIGN AND METHODOLOGY ISSUES

Our overall test period is the 3-year

period 1999-2001. During this period, the hard-copy, insider-trading disclosures pertinent to this study were due in the SEC’s office by 10th of the month following the trade. Our study covers the last 3 full-years prior to the SECs 2002 requirement that

insider-trading disclosures must be filed within 2 days of the trades.

Our test period represents the end of an era that is characterized generally by increases in both the number of SEC-mandated insider-trading disclosures and the availability of these disclosures. For example, during our test period the SEC-mandated data were available from the SEC in paper form almost immediately upon receipt. As a rule, such data also were recorded electronically by an outside vendor (originally Primark Corporation and subsequently Thompson Financial after it acquired Primark) within approximately 3-5 business days of receipt by the SEC. The Thompson/Primark data then were made available via a variety of sources including private real-time data feeds from Thompson/Primark and delayed feeds from Yahoo!, MSN, ThomsonFN, and other web sites.

For each experimental-group firm from this period, each earnings-announcement date studied is preceded by at least 1 large-volume purchase and 1 purchase cluster or 1 large-volume sale and 1 sale cluster occurring within 90 days of this date. In contrast, for each control firm, none of these signals occurs with 90 days of any earnings-announcement date studied. We use this double-signal/no signal approach to ensure that the experimental firms are associated with lower levels of information asymmetry on announcement dates than the control firms. As indicated, we derive our signals, and their absence, using the Lancer Analytics Insider Trading Model. We identify our earnings announcement dates via Compustat. Given adjustments for data availability, data errors, and anomalous observations our signal-identification approach yields a sample of 2,548 experimental firms for the 3-year test period. For the years 1999-2001, it yields sub-samples of 545, 1,133, and 870 experimental firms, respectively.

Our experimental and control firms are matched each year of our 3-year test period

on two dimensions. First, we match on firm size—as measured by year-end market capitalization. We obtained all price and share data needed in calculating market capitalization from the Center for Research in Security Prices (CRSP) database. Matching on size is meant to lessen the unwanted volume effects of differences in two factors: (a) information environment and (c) risk. With respect to (a), the observed inverse relationships between size and abnormal trading volume (e.g., Karpoff 1987) and size and duration of trading-volume reaction (e.g., Bamber 1986, 1987) is likely to be a consequence of the differing information environments of smaller versus larger firms. That is, other things equal, the earnings announcements of smaller firms are likely to yield greater abnormal volume than those of larger firms because more preannouncement information, from both public and private sources, typically is available for the latter (e.g., Grant 1980; Atiase 1985, 1987; and Bamber 1986, 1987). Thus, by matching on size, we control for information environment and, thus, abstract from the observed relationships mentioned above (see, e.g., Seyhun 1986). In regard to (b), matching on size also controls for risk to some extent since size is a reasonable proxy for risk (e.g., Fama and French, 1992).

Second, for each year, we match on earnings-announcement cumulative abnor-mal return over a 4-day cumulation period (i.e., days -1, 0, +1, and +2, where 0 is the earnings announcement date). In this regard, our volume results are based on 1 to 7 day cumulation periods utilizing days -1 through +5. This procedure ensures a reasonable level of value-relevant parity with respect to each of our matched pairs for each year and cumulation period (see below). This conservative procedure is intended to allow us to focus more finely on the purely volume-related implications of earnings announcements (i.e., changes in individual expectations and related changes in portfolio holdings). We utilize the 4-day cumulation

period for matching because most abnormally high earnings-related trading volume is likely to have occurred during this period. In this regard, prior research (e.g., Morse 1981; Bamber 1987; Cready 1988) suggests that the bulk of abnormally high earnings-related trading volume occurs on days -1 and 0 and that such reactions persist for up to day +5.

Our abnormal return matching procedure was successful since matched-pairs t-tests yield no significant mean differences for any of the 7 cumulation periods. We were not as successful in matching on size.

That is, the procedures we used in identifying our matched pairs tended to emphasize accuracy in regard to matching on abnormal returns as discussed above. This observation follows because, in each year, we first identified size deciles for our firms. Subsequently, we matched on cumulative abnormal return as precisely as possible for each experimental firm given its size decile. This procedure ensures that our matched firms are from the same size deciles; however, it does not necessarily ensure the best size-related match for each experimental firm given its size decile. Thus, in relation to size, matched-pairs t tests yield a significant mean difference for our 2,548 pairs. However, this difference is attributable to only the 7th and 10th deciles. Our design tends to compensate for size-related mismatches since we test mean differences in experimental group and control group announcement-period trading volume against similar mean differences occurring over contiguous nonannounce-ment periods.

We use Eventus 7 (Cowan, 2002) and the CRSP database to calculate abnormal returns using the market model, discretely compounded returns, and the equally weighted market index over the 255 days (i.e., the Eventus default) ending 90 days before the earnings announcement date, thus, avoiding the periods over which our

signals are generated as well as our earnings release dates.

In testing our null hypothesis, we take the daily difference in the earnings-announcement trading volume measures of matched experimental and control group firms and cumulate these differences over periods of 1 to 7 days for the years 1999-2001. Subsequently, we test our hypothesis using matched-pairs t-tests (supplemented by matched pairs signed-rank tests). In performing our t-tests, we compare our cumulation-period means with analogous experimental/control group non-announce-ment-period means for cumulation periods within 7-day periods occurring both before and after days -1 to +5. We use these non-announcement periods because they should be largely free of earnings-announcement effects. All needed volume data were obtained from the CRSP database.

We use market adjusted volume (MAVit) in calculating our cumulation period values. Calculations similar to those for MAVit shown below are used in produc-ing the means for the non-announcement periods. MAVit is based on unadjusted volume for firm i on day t (UVit). UVit is the percent of firm i’s shares traded on day t (t = -1, 0, +1 ,…, +5). UVit is given by:

UVit = STit / SOit (1)

Where: STit = shares of firm i traded on day t X 100;

SOit = shares of firm i outstanding on day t; and

UVit − UVjt = a UV volume difference for day t for matched firms i and j.

MAVit, as specified in, and calculated via, Eventus, is based on the natural log transformation of UVit (see Ajinka and Jain 1989; Cready and Ramanan 1991; Campbell and Wasley 1996) and is given by:

MAVit = ln (.000255 + UVit) −

{[ ∑i = 1

N ln (.000255 + UVit)] / N] (2)

Where: N = the number of days in the calculation;

.000255 = a constant preventing taking ln 0 for a day with no trading; and

MAVit − MAVjt = a MAV volume difference for day t for matched firms i and j.

RESULTS

Table 1 contains our matched-pairs t-test and signed-rank test results for MAVit. The probability values in Table 1 imply rejection of our null hypothesis that mean differences in experimental group and control group earnings-announcement-period trading volumes are greater than or equal to similar mean differences occurring over contiguous non-announcement holding periods. Thus, we conclude that our experimental group’s earnings-announce-ment trading volume is lower than that of our control group because of systematic reductions in experimental-group information asymmetry occasioned by the SEC-mandated disclosures. We draw this conclusion because Table 1 reveals significant differences for all comparisons except those for day −1, the day prior to each earnings announcement. Per-year tables (not included for brevity), following the pattern of Table 1, yield precisely the same conclusion at no worse than the .05 level with, for example only 2 minor per-year/period t-test exceptions (significance levels of .0563 and .0755).

Table 2 presents results implying that our cumulation-period significant differen-ces are largely attributible to volume reactions occurring on days 0, +1, and +2. Similar per-year analyses yield essentially the same conclusion for each year. In other words, they imply that each of the years contributes to our overall results similarly.

We replicated all the tests mentioned above using UVit. While we view the UVit results primarily as benchmark values for judging the efficacy of the market-adjustment process implicit in MAVit, these results were largely consistent with those of MAVit. Thus, they also imply rejection of our null.

FINAL OBSERVATIONS

The FASB’s conceptual framework (especially FASB, 1978, 1980) emphasizes the use of accounting information in individuals’ consumption/investment deci-sions, allocating scarce resources efficiently, and promoting the efficient functioning of securities markets. Consistently, the objective of FASB (1980) is identifying the characteristics, or qualities, of accounting information that make it decision useful (FASB, 1980, para 1). According to the FASB (1980, para 1), these qualities distinguish more, from less, decision useful accounting information. In this regard, relevance and reliability are the FASB’s primary qualities, and each of these qualities is necessary for decision usefulness (e.g., FASB, 1980, paras 15, 33, 44, 133).

Relevance is defined by the FASB (1980, Glossary) as “the capacity of information to make a difference in a decision by helping the user to form predictions about the outcomes of past, present, and future events or to confirm or correct prior expectations.” This definition is pertinent to interpreting the results of our study.

We find relatively low earnings- announcement trading volume for our experimental firms vis-à-vis our control firms. This result implies that the earnings announcements for our experimental firms (i.e., firms with decreased earnings-related information asymmetry) possess a decreased capacity for making differences in the decisions (i.e., portfolio holdings) of market participants and, thus, are less relevant than those of our control firms (i.e., firms with

greater earnings-related predisclosure information asymmetry). Ultimately, our results imply greater comparative earnings relevance for our control firms (i.e., firms such that the information contents of their earnings releases have not been preempted, at least to some extent, by earnings-related decreases in information asymmetry).

From a more general perspective, our findings are consistent with view that the most pervasive factor affecting secular changes in earnings relevance is the changing information environments of firms (but see the caveat in the introductory section). In the context of our study, the pertinent difference in information environments is the presence of SEC-mandated disclosures leading to Lancer Analytics buy and sell signals for our experimental firms, but not for our control firms. Our findings also are consistent with the idea that decreases in earnings-related predisclosure information asymmetry are primary factors in explaining changes (presumably decreases) in earnings relevance.

Three areas of future research are apparent from our results. First, in August of 2002, the SEC began requiring insiders to report their insider activities within 2 days of trading dates. Thus, studies could be designed to see if our results extend to periods beyond 2001. We have begun such a study.

Second, as mentioned, several studies present evidence supporting the validity of the models of Kim and Verrecchia (1991a) and He and Wang (1995) and to a lesser extent those of Kim and Verrecchia (1991b, 1994, 1997). In this regard, there is another important view related to information asymmetry and trading volume. This view follows primarily from the research of Diamond and Verrecchia (1991), Baiman and Verrecchia (1996), Healy et al. (1999), Brennan and Subrahmanyam (1995, 1996), Lang and Lundholm (1996), Leutz and Verrecchia (2000), Bushee and Noe (2000),

Verrecchia (2001, pp. 164-173), Healy and Palepu (2001), and Core (2001).

In sum, the later set of articles suggests that decreases in earnings-related predisclosure information asymmetry are associated with increases in institutional ownership, analyst following, trading by informed investors, and ultimately increased earnings-announcement trading volume. Thus, this view associates decreased information asymmetry and increased trading volume. In theory, the reductions in information asymmetry are significant because they increase share liquidity and reduce the cost of capital (e.g., Lang and Lundholm, 1996; Healy and Palepu, 2001; Core, 2001; Leutz and Verrecchia, 2000; and Verrecchia, 2001, pp. 164-173). In any case, our study suggests that a fruitful line of research is attempting to determine whether decreases in information asymmetry are associated with increased trading volume for appropriately selected and controlled samples of firms.

Third, as implied earlier, the models of Kim and Verrecchia 1991b, 1994, and 1997 encompass conditions under which new information can increase information asymmetry. Thus, studies could be designed to determine if these models have descriptive validity under conditions where one might expect new information to produce such increases.

_____________________________________________________________________________________________ TABLE 1

Test of Mean Differences—Market Adjusted Volume (MAVit)

Experimental and Control Groups ______________________________________________________________________________ Panel A: Severn-Day Non-announcement Period Preceding Cumulation Periods Years = 1999 to 2001; Number of Matched Firms = 2,548 One-Tailed Probabilities (p) Mean Differences ≥ Zero Matched Matched Cumulation Mean Standard Percent Pairs Pairs Signed- Period Difference Deviation Positive t-Test__ Rank Test 1-day 0.0359 1.1498 51.30 < .1152 < .1500 2-day 0.1643 1.6933 53.73 < .0001 < .0001 3-day 0.2704 2.1085 54.83 < .0001 < .0001 4-day 0.3252 2.5624 54.04 < .0001 < .0001 5-day 0.3665 3.0453 53.81 < .0001 < .0001 6-day 0.4034 3.5360 52.28 < .0001 < .0001 7-day 0.4373 4.0525 52.39 < .0001 < .0001 ______________________________________________________________________________ Panel B: Seven-Day Non-announcement-Period Following Cumulation Periods Years = 1999 to 2001; Number of Matched Firms = 2,548 One-Tailed Probabilities (p) Mean Differences ≥ Zero Matched Matched Cumulation Mean Standard Percent Pairs Pairs Signed- Period Difference Deviation Positive t-Test__ Rank Test 1-day 0.0359 1.1498 51.30 < .5205 < .1500 2-day 0.1643 1.6933 53.73 < .0001 < .0001 3-day 0.2704 2.1085 54.83 < .0001 < .0001 4-day 0.3252 2.5624 54.04 < .0001 < .0001 5-day 0.3665 3.0453 53.81 < .0001 < .0001 6-day 0.4034 3.5360 52.28 < .0001 < .0001 7-day 0.4373 4.0525 52.39 < .0001 < .0001 ______________________________________________________________________________ Notes: Mean differences are determined by non-announcement period values less cumulation-period values. _____________________________________________________________________________________________

_____________________________________________________________________________________________

TABLE 2

Tests of Per-Day Mean Differences—Market Adjusted Volume (MAVit) Experimental and Control Groups

______________________________________________________________________________ Panel A: Severn-Day Non-announcement Period Preceding Cumulation Periods Years = 1999 to 2001; Number of Matched Firms = 2,548 One-Tailed Probabilities (p) Mean Differences ≥ Zero Matched Matched Mean Standard Percent Pairs Pairs Signed- Days Difference Deviation Positive t-Test__ Rank Test −8 less −1 0.0359 1.1498 51.29 < .1152 < .1500 −7 less 0 0.1284 1.1096 53.73 < .0001 < .0001 −6 less +1 0.1062 1.0452 53.77 < .0001 < .0001 −5 less +2 0.0547 1.1517 51.69 < .0165 < .0277 −4 less +3 0.0414 1.2087 51.37 < .0841 < .1512 −3 less +4 0.0369 1.2393 50.27 < .1335 < .2803 −2 less +5 0.0339 1.2597 50.39 < .1745 < .5309 ______________________________________________________________________________ Panel B: Seven-Day Non-announcement-Period Following Cumulation Periods Years = 1999 to 2001; Number of Matched Firms = 2,548 One-Tailed Probabilities (p) Mean Differences ≥ Zero Matched Matched Mean Standard Percent Pairs Pairs Signed- Days Difference Deviation Positive t-Test__ Rank Test +6 less −1 0.0152 1.1920 49.45 < .5205 < .5097 +7 less 0 0.1084 1.1871 52.90 < .0001 < .0001 +8 less +1 0.0914 1.1198 53.49 < .0001 < .0002 +9 less +2 0.0413 1.1751 52.59 < .0765 < .0455 +10 less +3 0.0135 1.1829 50.51 < .5648 < .5147 +11 less +4 0.0313 1.2980 49.65 < .2240 < .4495 +12 less +5 0.0440 1.2324 52.12 < .0719 < .0565 ______________________________________________________________________________ Notes: Mean differences are determined by non-announcement period values less cumulation-period values.

REFERENCES

Ajinkya, B. and P. Jain. 1989. The behavior

of daily stock market trading volume. Journal of Accounting and Economics 11 (4): 331-360.

Amir, E. and B. Lev. 1996. Value-relevance of nonfinancial information: The wireless communication industry. Journal of Accounting and Economics 19 (1): 3-30.

Atiase, R. 1985. Predisclosure information, firm capitalization, and security price behavior around earnings announcements. Journal of Accounting Research 23 (2): 21-36.

_____. 1987. Market implications of pre-disclosure information: Size and exchange effects. Journal of Accounting Research 23 (2): 168-176.

_____, and L. Bamber. 1994. Trading volume reactions to annual accounting earnings announcements: The incremental role of pre-disclosure information asymmetry. Journal of Accounting and Economics 17 (2): 309-329.

Baiman, S. and R. Verrecchia. 1996. The relation among capital markets, financial disclosure, production efficiency, and insider trading. Journal of Accounting Research 34 (2): 1-22.

Bamber, L. 1986. The information content of annual earnings releases: A trading volume approach. Journal of Accounting Research 24 (2): 40-56.

_____. 1987. Unexpected earnings, firm size and trading volume around quarterly earnings announcements. The Accounting Review 62 (3): 510-532.

Beaver, W. 1968. The information content of annual earnings announcements. Journal of Accounting Research 6 (Supplement): 67-92.

_____, M. McNichols, and J. Rhie. 2005. Have financial statements become less informative? Evidence from the ability of financial ratios to predict bankruptcy.

Review of Accounting Studies 10 (2): 93-122.

Bettis, C. 1992. Revisiting the Profitability of Insider and Outsider Trading. Ph.D. dissertation: Indiana University.

_____, and J. Coles. 1997. New evidence on insider trading around merger announcements. Working Paper: Arizona State University.

_____, D. Vickrey, and D. W. Vickrey. 1997. Mimickers of corporate insiders who make large volume trades. Financial Analysts Journal 53 (3): 57-67.

_____, D. Vickrey, and D. W. Vickrey. 2002. Volume effects of predisclosure infor-mation asymmetry. Journal of Accounting and Finance Research 10 (Summer II): 93-103.

Brennan, M. and A. Subrahmanyam. 1995. Investment analysis and price formation in securities markets. Journal of Financial Economics 38 (3): 361-381.

_____ and _____. 1996. Market microstructure and asset pricing: On the compensation for illiquidity in stock returns. Journal of Financial Economics 41 ( 3): 441-464.

Brown. S, K. Lo., and T. Lys. 1999. Use of R2 in accounting research: Measuring changes in value relevance over the last four decades. Journal of Accounting and Economics 28 (3): 83-115.

Bushee, B. and C. Noe. 2000. Corporate disclosure practices, institutional investors, and stock return volatility. Journal of Accounting Research 39 (Supplement): 171-202.

Campbell, C. and C. Wasley. 1996. Measuring abnormal daily trading volume for samples of NYSE/ASE and NASDAQ securities using parametric and nonparametric test statistics. Review of Quantitative Accounting and Finance 6: 309-326.

Collins, D., E. Maydew, and I. Weiss. 1997. Changes in the value-relevance of earnings and book value over the last 40 years.

Journal of Accounting and Economics 24 (4): 39-64.

Core, J. 2001. A review of the empirical disclosure literature: Discussion. Journal of Accounting and Economics 31 (3): 441-456.

Cowan, A. 2002. Eventus 7 User’s Guide (revised edition). Cowan Research, L.C.

Cready, W. 1988. Information value and investor wealth: The case of earnings announcements. Journal of Accounting Research 26 (2): 1-27.

_____ and R. Ramanan. 1991. The power of tests employing log-transformed volume in detecting abnormal trading. Journal of Accounting and Economics 14 (2): 203-214.

Dechow, P and C. Strand. 2004. Earnings quality. Research Foundation of the CFA Institute.

Diamond, D. and R. Verrecchia. 1991. Disclosure, liquidity, and the cost of capital. The Journal of Finance 46 (3): 1325-1359.

Dontoh, A. and J. Ronen. 1993. Information content of accounting announcements. The Accounting Review 68 (3): 857-869.

Dorfman, J. 1989. Some investors see red in SEC's plan to modify insider reporting. Wall Street Journal (February 22): C1.

Ely, K. and G. Waymire. 1999. Accounting standard-setting organizations and earnings relevance: Longitudinal evidence from NYSE common stocks, 1927-1993. Journal of Accounting Research 37 (3): 293-318.

Fama, E. and K. French. 1992. The cross-section of expected stock returns. Journal of Finance 47 (2): 127-165.

Financial Accounting Standards Board. 1978. Statement of Financial Accounting Concepts No. 1: Objectives of Financial Reporting by Business Enterprises. Stamford.

_____. 1980. Statement of Financial Ac-counting Concepts No. 2: Qualitative

Characteristics of Accounting Inform-ation. Stamford

Finnerty, J. 1976a. Insider activity and inside information: A multivariate analysis. Journal of Financial and Quantitative Analysis. 11 (2): 205-216.

_____. 1976b. Insiders and market efficiency. Journal of Finance 31 (3): 1141-1148.

Fowler, D., and C. Rorke. 1988. Insider trading profits on the Toronto stock exchange. Canadian Journal of Admin-istrative Sciences 66 (2): 13-25.

Francis, J. and K. Shipper. 1999. Have financial statements lost their relevance? Journal of Accounting Research 37 (3): 319-352.

Francis, J., K. Schipper, and L. Vincent. 2002a. Earnings announcements and competing information. Journal of Accounting and Economics 33 (3): 313-342.

Francis, J. K. Schipper, and L. Vincent. 2002b. Expanded relevance and the increased usefulness of earnings announcements. The Accounting Review 77 (3): 515-546.

Frankel, R. and X. Li. 2004. Characteristics of a firm's information environment and the information asymmetry between insiders and outsiders. Journal of Accounting and Economics 37 (2): 229-259.

Gabele, B. 1993. The inside story: What key players are buying and selling. Barron’s 73 (August 23): 16-19.

Givoly, D., and D. Palmon. 1985. Insider trading and the exploitation of inside information: Some empirical evidence. Journal of Business 58 (1): 68-87.

Grant, E. 1980. Market implications of differential amounts of interim infor-mation. Journal of Accounting Research 18 (2): 255-268.

Hakansson, J., Kunkel, and J. Ohlson. 1982. Sufficient and necessary conditions for information to have value in pure exchange. The Journal of Finance 37 (4): 1169-1181.

He, H., and J. Wang. 1995. Differential information and dynamic behavior of stock trading volume. The Review of Financial Studies 8 (4): 919-972.

Healy, P., A. Hutton, and K. Palepu. 1999. Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research 16 (4): 485-520.

_____ and K. Palepu. 2001. Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accoun-ting and Economics 31 (3): 405-440.

Jaffe, J. 1974a. The effect of regulation changes on insider trading. Bell Journal of Economics and Management Science 5 (2): 93-121.

Jaffe, J. 1974b. Special information and insider trading. Journal of Business 47 (3): 410-428.

Jeng, L., A. Metrick, and R. Zeckhauser. 1999. The profits to insider trading: A per-formance-evaluation perspective. Working Paper: Harvard University.

Karpoff, J. 1986. A theory of trading volume. Journal of Finance 41 (12): 1069-1087.

_____. 1987. The relation between price changes and trading volume: A survey. Journal of Financial and Quantitative Analysis 22 (2): 109-126.

Ke, B., S. Huddart, and K. Petroni. 2003. What insiders know about future earnings and how they use it: Evidence from insider trades. Journal of Accounting and Economics 35 (3): 315-346.

Kiger, J. 1972. An empirical investigation of NYSE volume and price reactions to the announcement of quarterly earnings. Journal of Accounting Research 10 (2): 113-128.

Kim, O., and R. Verrecchia. 1991a. Trading volume and price reactions to public announcements. Journal of Accounting Research 29 (3): 302-321.

_____, and _____. 1991b. Market reactions to anticipated announcements. Journal of Financial Economics 30 (2): 273-309.

_____, and _____. 1994. Market liquidity and volume around earnings announcements. Journal of Accounting and Economics 17 (1): 41-67.

_____, and _____. 1997. Preannouncement and event-period private information. Journal of Accounting and Economics 24 (4): 395-419.

Kross, W., J. G. Ha, and F. Heflin. 1994. A test of risk clientele effects via an examination of trading volume response to earnings announcements. Journal of Accounting and Economics 18 (2): 67-87.

Kross , W. and M. Kim. 2000. Differences between market responses to earnings announcements in the 1990s vs the 1960s. Working Paper: Purdue Univeristy.

Lancer Analytics. 2002. Lancer Analytics: Insider Trading Model Documentation.

Landsman, W. and E. Maydew. 2002. Has the information content of quarterly earnings announcements declined in the past three decades? Journal of Accounting Research 40 (2): 707-808.

Lang, M. and R. Lundholm. 1996. Cross-sectional determinants of analysts ratings of corporate disclosures. Journal of Accounting Research 31 (3): 246-271.

Lakonishok, J., and I. Lee. 2001. Are insider trades informative? The Review of Finan-cial Studies 14 (1): 79-111.

Leutz, C. and R. Verrecchia. 2000. The economic consequences of increased disclosure. Journal of Accounting Research 38 (Supplement): 91-124.

Lev, B. and P. Zarowin. 1999. The boundaries of financial reporting and how to extend them. Journal of Accounting Research 37 (3): 353-385.

Lin, J., and J. Howe, 1990. Insider trading in OTC markets. Journal of Finance 45 (4): 1273-1284.

Lobo, G., and S. Tung. 1997. Relation between predisclosure information asym-

metry and trading volume reaction around quarterly earnings announcements. Jour-nal of Business Finance and Accounting 24 (2): 851-867.

Lorie, J. and V. Niederhoffer. 1968. Predict-ive and statistical properties of insider trading. Journal of Law and Economics 11 (2): 35-51.

McCarroll, T. 1992. Trading on the inside edge. Time 139 (2): 47-49.

Morse, D. 1981. Price and trading volume reaction surrounding earnings announce-ments: a closer examination. Journal of Accounting Research 19 (3): 374-383.

Moss, J., and T. Kohers, T. 1990. Abnormal gains for insiders trading prior to unexpected corporate earnings and divi-dend announcements. Akron Business Review 21 (2): 63-74.

Noe, C. 1999. Voluntary disclosures and in-sider transactions. Journal of Accounting and Economics 27 (2): 305-327.

Penman, S. 1985. A comparison of infor-mation content of insider trading and management earnings forecasts. Journal of Financial and Quantitative Analysis 20 (2): 1-18.

Pope, A., B. Morris, and C. Peel. 1990. Insider trading: Some evidence on market efficiency and directors' share dealings in Great Britain. Journal of Business Finance and Accounting 17 (2): 359-380.

Pratt, S. and C. W. Devere. 1970. Relationship between insider trading and rates of return for NYSE common stocks, 1960-1966 in J. Lorie and R. Breadley eds, Modern Developments in Investment Management. Praeger.

Rozeff, M., and M. Zaman. 1988. Market efficiency and insider trading: New evi-dence. Journal of Business 61 (1): 25-44.

Ryan, S. and P. Zarowin. 2003. Why has the contemporaneous linear returns-earnings relation declined? The Accounting Review 78 (2): 523-553.

Seyhun, N. 1986. Insiders' profits, cost of trading and market efficiency. Journal of Financial Economics 16 (2): 189-212.

_____. 1992. The effectiveness of insider-trading sanctions. Journal of Law and Economics 35 (2): 149-182.

_____. 1998. Investment intelligence: From insider trading. MIT Press.

Sinha, N. and J. Watts. 2001. Economic consequences of the declining relevance of financial reports. Journal of Accounting Research 39 (December): 663-681.

Utama, S., and W. Cready. 1997. Institutional ownership, differential pre-disclosure precision and trading volume at announcement dates. Journal of Account-ing and Economics 24 (4): 129-150.

Varian, R. 1986. Differences of opinion in financial markets. working paper, University of Michigan.

Verrecchia, R. 2001. Essays on disclosure. Journal of Accounting and Economics 32 (4): 97-180.

Vickrey, D., C. Bettis, D. W. Vickrey, and Taylor Foster. 2003. Implications of predisclosure information asymmetry for ethical managerial disclosure. Journal of Accounting and Finance Research 11 (4): 8-20.