Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Get this document free

02-

VIEWS: 17 PAGES: 17

									                                           The Timing of Directors‟ Sharedealings and Company Valuation 15




         The Timing of Directors‟ Sharedealings and Company
                            Valuation

                                           Jorge Belaire-Franch*
                                        University of Valencia, Spain

                                               Philip Hamill++
                                       University of Glasgow, Scotland

                                             Philip McIlkenny+*
                                       University of Glasgow, Scotland
                                        University of Ottawa, Canada

                                              Kwaku K. Opong+
                                       University of Glasgow, Scotland


Abstract Directors are assumed to have more information about their company‟s prospects than any other
market participant. Thus it would appear that directors are well placed to profit from insider trading. This
suggests, therefore, that company directors may time their dealing activities to coincide with periods which
will be most profitable to them. Directors will therefore buy when they perceive the firm‟s share price to be
undervalued and sell when the price is overvalued. This study seeks to provide evidence on the ability of
company directors to time their purchasing (selling) activities to coincide with periods when they perceive the
value of their firms as relatively low (high) compared to their valuation on the basis of their superior
information. In the presence of strong form market efficiency, the timing ability of directors share dealings
will be redundant. The sample comprises those firms that constitute the FTSE 250. Whilst not conclusive, the
results of the study appear to indicate that directors do have some timing ability in their share dealings.
However, no significant positive abnormal returns are earned following directors‟ share dealing
announcements.

Keywords Insider Trading, Market Efficiency, Timing ability, Directors‟ share dealings




Introduction

    Directors are assumed to have more information about their company‟s prospects than any
other market participant. Thus it would appear that directors are well placed to profit from insider
trading. The financial press (Nuki 1994) suggests that a profitable trading strategy is possible for
market participants who follow directors‟ trading pattern. This is only possible if directors could
time their share dealing activities due to their superior knowledge about the prospects of their
firms. Thus, this paper seeks to provide some evidence or otherwise of directors timing ability in
their share dealing activities. The next two sections discuss the rationale for the study and the
regulation of directors' share dealing activities respectively. This is followed by a discussion of the
extant literature on insider trading. The methodology employed in the study is also described in the
next section. This is followed by a discussion of the results and conclusions respectively.
16 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31


Do Directors Time Their Share Dealing Activities?

    A major rationale for this study is whether directors time their share dealing activities to
coincide with periods when the value of the firm is at its highest when they sell and lowest when
they buy. If this is the case, then, it would appear that it could be profitable to mimic what
directors do. A central point in mimicking what directors do is the assumption that directors know
whether their companies are over- or under-valued. This assumption is based on the notion that
directors are much more informed about the prospects of their companies and therefore can trade
on the basis of insider information1 by buying undervalued shares or selling overvalued shares.
The timing ability of directors is not made explicit in previous studies.

    The UK regulator, The Financial Services Authority (FSA) got tough new powers in 2001 to
act against directors who breach stock exchange rules on share dealing. The move will extend the
financial watchdog's remit to cover individuals who have broken the rules but then leave the
company before any investigation into their behaviour is completed. Under the rules the FSA has
the power to levy fines on former directors as well as companies and directors who remain in post.
Prior to 2001, the FSA could only censure companies and directors who had broken the rules. The
change comes against a background of high-profile cases where institutional investors have
questioned directors' share sales and insisted on board changes. In November 2000, Hartmut
Lademacher was forced out as non-executive director at Sema, the Anglo-French software
company, after selling 1.8m shares for £24m in the run-up to the interim results in September
2000.

    Malcolm Walker, the chairman of Iceland caused a furore about his sale of 4m shares in mid-
December 2000, five weeks ahead of a profit warning, and was forced to resign on 31 st January
2001 from the company he founded 30 years ago. Iceland insisted that Mr Walker had done
nothing wrong and said that the sale was in preparation for his long-standing plans to move to a
non-executive role in March. Iceland board pointed out that the sale was approved by the board
and would not have been authorised if it knew that it was about to publish information that would
affect the company's share price. This appears to indicate an incredible ignorance about what is
happening in the company. The sale sparked a row because the average price Mr Walker received
was 339p, a five year price high for Iceland. The shares slumped by about 40% just before Mr
Walker resigned. Mr Walker left by agreement that he will not be receiving compensation for loss
of office.

    Due to information asymmetry that exists between directors and investors, knowing what
directors are doing is important since it can indirectly reveal what they know. Information about
directors' trades is widely available in the financial press and is followed with interest by investors.
In September 1997, Chrysler Corporation had to issue a public statement following the sale by
directors of unusually large quantities of shares that the sale was not to be interpreted as a bearish
signal. The issue as to whether directors can time their trades is an empirical one and the current
study attempts to add to the scant but growing literature in this area. Two hypotheses are tested in
this study. First, if directors, by virtue of superior information can time their trades, we should
expect that directors‟ purchases should coincide with periods when prices are low and sales will be
during periods when prices are at their highest. The second hypothesis tested in this study is that if
directors trade in their securities, it should make those securities much more liquid and lead to an
increase in trading volume around the period when they deal because some investors will attempt
to mimic what they do.

1
    Insider trading is illegal under the law. Directors can only trade outside a closed period.
                                        The Timing of Directors‟ Sharedealings and Company Valuation 17


UK Regulation of Directors’ Sharedealing Activities

    Sections 198-220 of the Companies Act 1985 contain provisions for securing the disclosure
and registration of substantial individual interests in share capital that carry unrestricted voting
rights. The relevant sections require notification to the company of any acquisition or divestiture
of three percent in writing within two business days. The relevant sections also require notification
of family and corporate interests as well as members acting in a concert party. Section 142 of the
Financial Services Act (1986) and the Financial Services and Markets Act (2000) empowers the
Stock Exchange itself to draw up listing rules for the admission of securities to listing on the
Exchange. There are provisions in the listing rules that deals with directors continuing obligations
among which is a model code on directors‟ share dealings activities. The purpose of the Code is to
ensure that directors, certain employees and persons connected with them do not abuse, or even
place themselves under suspicion of abusing, price-sensitive information that they may have or be
thought to have. This is especially important in periods leading up to an announcement of
company financial reports. The principal restrictions on dealings cover the following areas:

(a) A director must not deal in his company's securities on considerations of a short term nature
    or at any time when he is in possession of unpublished price-sensitive information in relation
    to those securities or during a 'close period' i.e. the period of two months before the
    announcement of the company's results for a financial period (or one month in the case of
    quarterly reports) or, if shorter, the period from the end of the relevant financial period to the
    time of the announcement.

(b) A director needs clearance from his or her chairperson or other director designated for the
    purpose before they can deal in the securities of their company.

(c) In exceptional circumstances, clearance may be given for a director to sell (but never to
    purchase) securities when he would otherwise be prohibited from doing so only because the
    proposed sale would fall within a close period. Note however that the company's obligation to
    notify the dealing must state the exceptional circumstances which were pertaining at the time
    of the dealing which may lead to personal circumstances being disclosed publicly.

    There are special rules where a director has funds under discretionary management or an
interest under a trust or acts as a trustee, and also in the case of single company personal plans and
other saving schemes.

    It should be noted that compliance with the Model Code does not of itself ensure compliance
with the insider dealing legislation in the Criminal Justice Act (1993). The act stipulates that it is a
criminal offence to deal or encourage another person to deal or disclose inside information unless
no dealing is expected. Even though directors may have the incentive to trade on inside
information, in the presence of market efficiency, such an attempt may not be worthwhile. Also, It
can be argued that a company that gets a reputation for 'leakiness' (ie. where inside deals in its
securities are regularly suspected) may find it difficult or expensive to raise new capital. The Stock
Exchange attempts to control insider trading and has published a document providing guidance on
the dissemination of price-sensitive information, with the intention that all listed companies follow
the same strict guidelines. The managerial market may also penalize managers who betray their
fiduciary duty to shareholders by trading on inside information.
18 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31


Literature Review

    Prior studies on insider trading in the U.K. seem to suggest that insiders can earn significant
positive abnormal returns by trading their own firm‟s securities (see King and Roell,1999; Pope,
Morris and Peel, 1990; Gregory, Matatko, Tonks and Purkis, 1994; Gregory, Matatko and Tonks ,
1997), Ireland (Hamill, McIlkenny and Watson, 1999) and the USA (Jaffe, 1974; Finnerty, 1976,
Givoly and Palmon, 1985; Seyhun, 1986; Lin and Howe, 1990; Jeng, Metrick and Zeckhauser
1999). The seminal paper by Jaffe (1974) provides evidence that insiders‟ purchases (sales) take
place after abnormal share price decreases (increases). Thus insiders are able to time their trades.
Nunn, Madden and Gombola (1983) hypothesised the existence of an information hierarchy. They
find that the performance of trades by chairmen and directors outperforms those made by other
corporate insiders such as officers and substantial shareholders. Seyhun (1986) and Lin and Howe
(1990) corroborate these findings.

    Results from studies examining market reaction to directors‟ trades in UK companies provide
conflicting evidence (King and Roell,1988; Pope, Morris and Peel, 1990; Gregory, Matatko,
Tonks and Purkis, 1994). Gregory, Matatko and Tonks(1997) seek to reconcile the differences in
the evidence by using the most comprehensive data set to date. They report that significant
abnormal returns can be earned if the appropriate trading strategy, based on the directors‟ trades, is
followed. They also emphasise the importance of controlling for size in event studies where
cumulative abnormal returns are being investigated over a long post-event window and the
proportion of small companies in the sample is high. In using the FTSE 250 companies we hope to
avoid the size effect detected by Gregory et al (1997). Whilst significant abnormal returns can be
earned, these returns take no account of risk. Modigliani and Modigliani (1997) suggest that the
resulting figures from current measures of risk adjusted performance [Sharpe ratio, Sharpe (1966),
Jensen‟s alpha, Jensen (1968) and Treynor ratio, Tryenor (1968)] are difficult to interpret. Thus
they propose an alternative risk-adjusted performance (RAP) measure. RAP, they suggest, makes
the comparison in the performance of any managed portfolio against that of a relevant unmanaged
„market‟ portfolio only after properly adjusting the portfolio return for risk.

    Jeng, Metrick and Zeckhauser (1999) analyse the returns over a one year period to a value-
weighted rolling purchase portfolio and a sale portfolio. Their approach, they argue, is free of the
statistical difficulties that plague event studies on long-horizon returns. They find that the purchase
portfolio only earns abnormal returns and that there is no difference between the abnormal returns
to insider trades in small firms and the abnormal returns to insider trades in large firms. McIlkenny
and Opong (2000) contend that directors purchase or sell with a longer time horizon in mind than
one year. Thus by replicating the trading activities of directors for a five year period, they create a
portfolio that captures this time horizon. Their results suggest that an investor can obtain an annual
7.4% return from tracking directors‟ trading activity compared to the FTSE 250 index return of
9.9%. However after adjusting for risk, using the method proposed by Modigliani and Modigliani,
the directors‟ portfolio return declines to 4.96%. A major contribution of this study is that it
examines directors timing ability which previous studies appear to have overlooked. The focus of
previous studies have been on whether directors can earn superior returns from their trading
activities.
                                           The Timing of Directors‟ Sharedealings and Company Valuation 19


Data

    The sample of firms in this study comprises those firms that constitute the London Stock
Exchange‟s FTSE 250 that have data on directors‟ share dealings from June 1994 to June 1999. Data
is available in Extel on directors‟ share dealings from January 1994 which puts a limit on the starting
date for the current study. To be included in the final sample, an event has to satisfy two sets of
criteria. First, data must be available in Extel and second, price and/or volume data must be
available in Datastream/Primark. The first set is used to identify open market transactions by
directors (purchase or sale of shares in their own companies) and the second set is used to ensure
that the firm‟s weekly share price is available from Datastream.

    Only open market transactions by directors are used as the proxy for the timing of directors‟
share dealings. All other share transactions by directors are excluded from the study, examples of
which include; exercise of options on shares, share bonus entitlement, etc, (see Hamill, McIlkenny
and Watson, (1999) for a detailed explanation of their exclusion). A total of 7956 open market
transactions by directors took place during the five year period encompassed by this study.




Methodology

    The sample for the study is made up of firms listed on the London Stock Exchange‟s FTSE
250 from June 1994 to June 1999 that have data on directors‟ share dealings during the period
mentioned above. Each firm is analysed in two time periods, namely, an estimation period made
up of 129 trading days prior to the beginning of the test period. The test period is made up 21
trading days prior to the announcement day and 21 days subsequent to the announcement day

   Normal daily returns were generated using the standard market model [see Fama (1965)].
   The market model is given by:

          R i,t =  +  ( R m,t )+  i,t                                                  (1)

    where,
    εit is the excess returns accruing to shareholders in firm i on day t relative to the day of the
announcement of the directors‟ share dealing; Rit is the daily returns to shareholders adjusted for
dividends and other capital changes; Rmt is the daily returns on the Financial Times All Share
Index; α and β are parameter estimates.

    The market model parameters in the estimation period are estimated using the 129 daily price
observations (from day -150 to day -22) before the announcement of the directors‟ share dealing.
Fisher (1966) first pointed out the problems that are caused by asynchronous prices in the
calculation of returns. The importance of this problem becomes amplified with a shorter
differencing interval and infrequently traded securities. The systematic risk estimated from the
market model in (1) is subject to overestimation (underestimation) for frequently (infrequently)
traded stocks. The Scholes and Williams (1977) procedure for overcoming the problem of non-
synchronous trading regresses the firm return on the contemporaneous, lagged and lead market
return, thus producing a contemporaneous         0,
                                                  j
                                                       lagged  1 and lead  1 . The beta required for
                                                                j             j
equation (3) is then estimated as,
20 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31



                   (  1   0   1)
           =          j      j     j
                                                                                      (2)
                           1  2 1

   where ρ is the autocorrelation coefficient of the market index return.


   The daily excess returns were computed as:

          ei,t = Ri,t - (  +  ( Rm,t ))                                             (3)

   where ei,t , Ri,t, α , β and Rm,t are as defined previously.

    The daily excess returns were averaged across the observations according to

                    1 N
          AR t =        e                                                            (4)
                    N i=1 i,t

   Daily averages of excess returns are calculated for each of the days from day -21 to day +21
and day +150. These averaged daily excess returns were tested for significance according to

                   ARt
          t AR =                                                                       (5)
                   SE it

    where SEit = [var (ARt)]1/2 with var estimated over the 129 days, -150 to -22. In addition,
cumulative average excess returns (CARs) are calculated over various holding periods from day K
to L:

                         L
          CARK,L =           ARt                                                     (6)
                       t= K


    where K and L are the beginning and ending day of the holding periods respectively.
   The significance of CARs are tested using the following [see Rubac (1982), Bonnier and
Bruner(1989)]:

           t CARK,L = CARK,L /S( CARK,L )                                             (7)

    where S(CARK,L) = [T( var(ARt)) + 2(T-1)cov(ARt,ARt-1)]2 , with var and cov estimated over
the 129 days -150 to -22 and T = L-K + 1. Coutts, Mills and Roberts (1995) provide a robust test
of cumulated excess returns in the test period in equation (4). The Coutts et al test (see Appendix
A) remedies the deficiencies in the market model and accounts for serial correlation and non-
normality in abnormal returns in event studies. A test based on Coutts et al is also conducted.
                                        The Timing of Directors‟ Sharedealings and Company Valuation 21


Trading Volume Analysis

    Due to missing volume data, the number of firms used in the volume analysis totalled 111. The
methodology adopted for the analysis of trading volume in the period when directors trade
assumes that the volume of shares traded on a particular day equals to the average of shares traded
in the estimation period. At the moment, there is no theoretical model available to explain trading
volume behaviour. The expected volume of shares traded in a particular day is given by:


           E (V i,t )  V                                                             (7)

   where
                                                                  
    Vi,t is the percentage of firm i's shares traded in day t and V  is the average of firm i‟s shares
traded based on 129 daily observations in the estimation period from day -150 to -22.

   Abnormal trading volume is therefore given by


            i,t  V i,t  V                                                          (8)

   Abnormal volume is estimated over test period days -21 to +150.             Significance tests are
conducted using the following equation:

                           t
           t         =                                                                (9)
                i,t       S
                            

   where t  e = [var (  )]1/2 with var(  ) estimated over the 129 days, -150 to -22. In addition,
equations (5) and (6) are used to compute cumulative excess trading volume for the test period.




Discussion of Results

    The results of the behaviour of abnormal returns around the period of directors‟ share dealings
are reported in Table 1. Column 1 in table 1 indicates the days relative to the purchase or sale of
equities by company directors. Column 2 represents directors‟ purchases and column 4 represents
directors‟ sales. Columns 3 and 5 represent the t-statistics for the values in columns 2 and 4
respectively. It appears from Table 1 that the price effects of directors‟ purchases is more
pronounced than sales. Whereas directors‟ sales are not associated with any significant price
activity in the test period examined, a significant positive impact is observed on the day following
directors‟ equity purchases. However, all significant values after day +1 are negative.
22 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31


                    Table 1. Abnormal Returns Around Directors‟ Equity Dealings

                        Directors‟                                 Directors‟
  Day                   Purchases               t-statistics        Sales              t-statistic

   -21                  -0.0008                 -1.3229             0.0010              1.0300
   -20                  -0.0008                 -1.3520             0.0001              0.0541
   -19                   0.0000                  0.0798             0.0001              0.0724
   -18                  -0.0007                 -1.1253             0.0001              0.1257
   -17                   0.0000                  0.0728             0.0003              0.3173
   -16                  -0.0006                 -0.9474            -0.0000             -0.0397
   -15                  -0.0004                 -0.6005             0.0015              1.5558
   -14                  -0.0008                 -1.3200             0.0016              1.6630
   -13                   0.0004                  0.6675             0.0008              0.7779
   -12                  -0.0000                 -0.0571             0.0012              1.2307
   -11                  -0.0008                 -1.3580             0.0017              1.7869
   -10                  -0.0014                 -2.2666 **          0.0014              1.3959
    -9                   0.0011                  1.8044            -0.0012             -1.2035
    -8                   0.0005                  0.8278             0.0006              0.6243
    -7                  -0.0003                 -0.5121             0.0009              0.9037
    -6                   0.0009                  1.4474             0.0003              0.2834
    -5                  -0.0020                 -3.2688             0.0014              1.4676
    -4                  -0.0007                 -1.2171             0.0018              1.8638
    -3                   0.0005                  0.7731             0.0004              0.3711
    -2                   0.0001                  0.1860             0.0016              1.6687
    -1                   0.0001                  0.1331            -0.0000             -0.0159
     0                   0.0006                  0.9090            -0.0011             -1.1735
     1                   0.0019                  3.0233**          -0.0018             -1.8269
     2                  -0.0014                 -2.2421            -0.0016             -1.6924
     3                  -0.0000                 -0.0392            -0.0016             -1.6631
     4                  -0.0000                 -0.0005            -0.0007             -0.7041
     5                   0.0002                  0.2675            -0.0006             -0.6032
     6                   0.0002                  0.3140             0.0002              0.1621
     7                  -0.0003                 -0.5279             0.0003              0.3214
     8                  -0.0005                 -0.7869            -0.0017             -1.7623
     9                   0.0000                  0.0197             0.0000              0.0382
    10                  -0.0016                 -2.6327 **          0.0001              0.0565
    11                  -0.0001                 -0.1025             0.0008              0.8066
    12                   0.0003                  0.4293             0.0000              0.0216
    13                   0.0007                  1.0873            -0.0014             -1.4439
    14                   0.0012                  1.9773            -0.0010             -1.0699
    15                   0.0002                  0.2663            -0.0004             -0.3666
    16                   0.0009                  1.4460            -0.0003             -0.2855
    17                   0.0000                  0.0168            -0.0013             -1.3173
    18                  -0.0004                 -0.6548            -0.0015             -1.5633
    19                   0.0006                  1.0173            -0.0005             -0.5487
    20                   0.0005                  0.7529             0.0008              0.7965
    21                   0.0002                  0.3355             0.0004              0.4429

Notes: ** Significant at 0.01 level; * Significant at 0.05 level

    To test the hypothesis regarding directors‟ timing ability in their dealing activities, we plot the
behaviour of cumulative abnormal returns following directors‟ purchase and sales of equities.
Figure 1 shows the behaviour of cumulative abnormal returns following directors‟ purchase of
equities in their own firms. The graph shows that, on the average, directors purchase when prices
appear to be low. Figure 2 shows the behaviour of cumulative abnormal returns in the period
around directors‟ sales of equities and this is very revealing. The graph shows that, on the
average, directors sell when prices appear to be highest in the test period. The two graphs provide
evidence that directors appear to have some timing ability regarding their share dealings. The
                                            The Timing of Directors‟ Sharedealings and Company Valuation 23


results support the seminal paper by Jaffe (1974) who provides evidence that insiders‟ purchases
(sales) take place after abnormal share price decreases (increases).

    Tables 2 and 3 provide results of cumulative abnormal returns for various holding periods in
the test period examined. The results show that none of the values for the holding periods reported
is statistically significant. The test based on Coutts et al (1995) and reported in Tables 2 and 3
support the results based on equations (5) and (6).



    Table 2. Cumulative Abnormal Returns Around Announcement of Directors‟ Equity Purchases Date

                                                                             Cumulative
                              Cumulative
           Day                                        t-statistics       Abnormal Returns         p-value
                            Abnormal Returns
                                                                            (Coutts et al)
    -21           -17              0.00155                0.4437              -0.00448            0.05729
    -16           -12              0.00503                1.4388              -0.00200            0.39970
    -11            -7              0.00340                0.9728              -0.00081            0.73507
     -6            -3              0.00386                1.2605              -0.00488            0.02017
     -2            2              -0.00295               -0.8432              -0.00081            0.73178
     -1            1              -0.00292               -1.1401               0.00178            0.16075
     0             0              -0.00114               -1.1735              -0.00114            -1.1735
     3             7              -0.00241               -0.6897               0.00169            0.23904
     8             12             -0.00081               -0.2328              -0.00156            0.50899
     13            17             -0.00435               -1.2434               0.00476            0.02264
     18            21             -0.00085               -0.2759               0.00155            0.22875
Notes: None of the values is statistically significant; Day is the day relative to the announcement of directors‟
      equity purchase/sale




 Table 3. Cumulative Abnormal Returns Around Announcement of Directors‟ Equity Sales Date

                                                                             Cumulative
                              Cumulative
           Day                                        t-statistics       Abnormal Returns         p-value
                            Abnormal Returns
                                                                            (Coutts et al)
     -21          -17             0.00155                 0.4437               0.00516            0.03758
     -16          -12             0.00503                 1.4388               0.00621            0.01582
     -11           -7             0.00340                 0.9728              -0.00032            0.91174
      -6           -3             0.00386                 1.2605               0.00566            0.01275
      -2           2             -0.00295                -0.8432              -0.00387            0.19137
      -1           1             -0.00292                -1.1401              -0.00320            0.14638
      0            0             -0.00114                -1.1735              -0.00114            -1.1735
      3            7             -0.00241                -0.6897              -0.00191            0.51093
      8            12            -0.00081                -0.2328              -0.00201            0.48645
      13           17            -0.00435                -1.2434              -0.00168            0.56176
      18           21            -0.00085                -0.2759              -0.00063            0.80545
Notes: None of the values is statistically significant; Day is the day relative to the announcement of directors‟
equity purchase/sale
24 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31


           Table 4. Mean Abnormal Trading Volume Around Directors‟ Equity Dealings

                                   Directors‟                        Directors‟
           Day                     Purchases                           Sales

           -21                     -0.052139                          0.172593**
           -20                     -0.050155                          0.130925**
           -19                     -0.023182                         -0.038391
           -18                     -0.002390                          0.032815
           -17                     -0.024776                          0.083712*
           -16                     -0.019944                          0.088995*
           -15                     -0.028530                          0.015520
           -14                     -0.025640                         -0.021717
           -13                     -0.021620                          0.042441
           -12                     -0.045296                          0.008809
           -11                      0.001676                         -0.019848
           -10                     -0.029247                         -0.046189
           -9                      -0.053222                          0.041487
           -8                      -0.034844                          0.051725
           -7                      -0.010253                          0.058274
           -6                       0.081895**                        0.015511
           -5                       0.065156*                        -0.003118
           -4                      -0.031085                          0.242107**
           -3                      -0.008935                          0.099894**
           -2                       0.024315                          0.026313
           -1                       0.172661**                        0.486997**
            0                       0.023234                          0.128562**
            1                       0.042299                          0.037790
            2                       0.023481                         -0.030506
            3                      -0.021178                          0.005597
            4                      -0.001896                         -0.008874
            5                      -0.014375                         -0.008235
            6                      -0.004737                         -0.017876
            7                       0.004499                          0.034528
            8                      -0.030691                         -0.005214
            9                       0.001413                          0.052122
           10                      -0.047573                          0.001174
           11                      -0.009552                         -0.002310
           12                      -0.017140                          0.032794
           13                       0.029082                          0.006398
           14                      -0.041964                          0.007505
           15                       0.035582                         -0.029083
           16                      -0.028833                          0.044153
           17                       0.075392**                        0.003277
           18                      -0.044701                         -0.004792
           19                      -0.042931                          0.006977
           20                      -0.038804                         -0.061639
           21                      -0.029332                          0.060038


  Notes: ** Significant at 0.01 level; * Significant at 0.05 level
                         The Timing of Directors‟ Sharedealings and Company Valuation 25




               Days Relative to Directors Equity Sales

Figure 1. Cumulative Abnormal Returns for Directors‟ Equity Purchases




           Days Relative to Announcement of Equity Sales

      Figure 2. Cumulative Abnormal Returns for Directors sales
26 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31




                                Days Relative to Equity Purchase

   Figure 3. Mean Abnormal Trading Volume Around Directors‟ Equity Purchase Date




                          Days Relative to Directors‟ Equity Purchase

                   Figure 4. Cumulative Abnormal Trading Volume Around
                               Directors‟ Equity Purchase Date
                              The Timing of Directors‟ Sharedealings and Company Valuation 27




                    Days Relative to Directors‟ Equity Sales

     Figure 5. Abnormal Trading Volume Around Directors‟ Equity Sales Date




                    Days Relative to Directors‟ Equity Sales

Figure 6. Cumulative Abnormal Trading Volume Around Directors‟ Equity Sales Date
28 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31



    The results of the behaviour of trading volume are reported in Table 4 and Figures 3 –6. Table
4 indicates that significant above average trading takes place on the day prior to directors‟ equity
purchases. For directors‟ equity sales, significant above average trading takes place in the equity
of those firms on the day prior to the sale as well as on the day that directors sell equities in their
firms. A plot of abnormal trading volume is reported in Figures 3 and 5 for directors‟ purchases
and sales respectively. The plot clearly shows that abnormal trading volume is highest on the day
prior to and the day that directors‟ purchase or sales take place. If the liquidity hypothesis is valid
for directors‟ equity dealings, there should be an immediate increase in trading volume following
the periods. While the results support the view that directors possess some timing ability in their
share dealing activities, it is not clear from this study that mimicking what company directors do
could be profitable since no significant positive abnormal returns are earned following directors‟
share dealings announcements.




Conclusions

    Directors are assumed to have more information about their company‟s prospects than any
other market participant. Thus it would appear that directors are well placed to profit from insider
trading. This suggests, therefore, that company directors may time their dealing activities to
coincide with periods which will be most profitable to them. Directors will therefore buy when
they perceive the firm‟s share price to be undervalued and sell when the price is overvalued. This
study seeks to provide evidence on the ability of company directors to time their purchasing
(selling) activities to coincide with periods when they perceive the value of their firms as relatively
low (high) compared to their valuation on the basis of their superior information. In the presence
of strong form market efficiency, the timing ability of directors share dealings will be redundant.
The sample comprises those firms that constitute the FTSE 250. Whilst not conclusive, the results
of the study appear to indicate that directors do have some timing ability in their share dealings. It
appears that directors share purchases (sales) usually follow abnormal share price decreases
(increases). However, no significant positive abnormal returns are earned following directors‟
share dealing announcements.
                                       The Timing of Directors‟ Sharedealings and Company Valuation 29




References

Bonnier, K. and Bruner, R.F. (1989) An analysis of stock price reaction to management change in
           distressed firms, Journal of Accounting and Economics, 11, 95-106.
Coutts, J A, Mills, T. C. and Roberts, J. (1995). Testing cumulative prediction errors in event study
           methodology, Journal of Forecasting, 14, 107-115.
Fama, E. (1965), „The Behaviour of Stock Market Prices‟, Journal of Business, (January), pp. 34-
           105
Finnerty, J.E. (1976), Insiders and Market Efficiency, Journal of Finance, Vol.31, pp.1141-8.
Fisher, L., (1966), Some new market indexes, Journal of Business, Vol. 39, No. 1, Pt. 2(January).
Givoly, D. and Palmon, D., (1985), Insider Trading and The Exploitation of Inside Information:
           Some Empirical Evidence, Journal of Business, Vol.58, pp.69-87.
Gregory, A., Matatko, J. and Tonks, I. (1997). Detecting Information from Directors Trades:
           Signal Definition and Variable Size Effects, Journal of Business Finance and Accounting
           17(3), pp. 309-342.
Gregory, A., Matatko, J., Tonks, I. and Purkis, R. (1994), UK Directors‟ Trading: The Impact of
           Dealing in Smaller Firms‟, Economic Journal, Vol.104, pp.37-53.
Hamill, P., McIlkenny, P. and Watson I. (1999), Market Reaction to Directors‟ Share Dealings:
           Some Irish Evidence, Irish Accounting Review, 6(1), pp.59-78.
Jaffe, J. (1974). Special Information and Insider Trading, Journal of Business 47,
Jeng L.A., Metrick A., and Zeckhauser R.J. (1999) , The Profits to Insider Trading: A
           Performance-Evaluation Perspective, Working Paper, Wharton School, University of
           Pennsylvania,
Jensen, M.C., (1968), The Performance of Mutual Funds in the Period 1945-64, Journal of
           Finance, May.
King, M. and Roell, A. (1988), Insider Trading, Economic Policy, Vol.7, pp.163-93.
Lin, J.C. and Howe, J.S., (1990), Insider Trading in the OTC Market, Journal of Finance, 45,
           pp.1273-84.
Nuki (1994), Follow Insiders to Best Shares, The Sunday Times, 25 Sept.
Nunn, K.P., Madden, G. P. and Gombola, M. H., (1983). “Are Some Insiders More 'Inside' than
           Others?” Journal of Portfolio Management, 9, 18-22.
McIlkenny, P. and Opong, K K (2000), Portfolio Returns From Tracking U.K. Directors‟ Share
           Dealings, Working Paper.
Modigliani, F. and Modigliani, L. (1997), Risk-Adjusted Performance How to Measure and Why,
           The Journal of Portfolio Management, Winter, pp.45-54.
Pope, P.F., Morris, R.C. and Peel, D.A. (1990), Insider Trading: Some Evidence on Market
           Efficiency and Directors‟ Share Dealings in Great Britain, Journal of Business Finance
           and Accounting, Vol. 17, No.3 (Summer), pp.359-80.
Ruback, R. (1982). „The Effect of Discretionary Price Control Decisions on Equity Values‟,
           Journal of Financial Economics, Vol. 10, 1982, pp.46-60.
Scholes, M. and Williams, J. (1977). Estimating Betas from Nonsynchronous Data, Journal of
           Financial Economics, 5, pp.309-327.
Seyhun, H.N. (1986), Insiders‟ Profits, Costs of Trading and Market efficiency, Journal of
           Financial Economics, Vol.16, pp.189-212.
Sharpe, W.F., (1966), Mutual Fund Performance, Journal of Business, Supplement on Security
           Prices, 39.
Tryenor , J.L., (1968), How to Rate Management Investment Funds, Harvard Business Review, 43.
30 Belaire-Franch, Hamill, McIlkenny and Opong / Journal of Accounting and Finance 3 (2004) 15~31




Appendix A

   A.1. ZD – test

    Coutts, Mills and Roberts (1995) provide an alternative specification to Patell's methodology,
which they demonstrate does not have the correct standard error. However, errata in Coutts et al's
methodology makes it difficult to replicate their ZD test. The methodology presented here
remedies these errors. Figure below outlines the notation used for the estimation and event period,
primarily for the purposes of pedagogy.

Time 1                                                T               T+1                       e                     T+m



                Estimation period                                                           Event Period

   We have the single index market model written in vector form as:

          y it  xit  i  u it                                                                            ( A.1.1)


   where xit = (1, xit ) and  i = ( i ,  i) or, in matrix form as

          yi  Xi  i  ui                                                                                  (A.1.2)
          where yi  yi1,......... yiT  ,
                                  ..,                      ui  u i1,......, iT  , and
                                                                             u
                                        1,......., 
           Xi  xi1,......,xiT   
                                                     1
                                                       
                                        xi1,.....,xiT 

   The estimated model is used to forecast m future observations yi *  ( yi,T 1,...,yi,T  m) using
the matrix of future observations;


                     
                    i,           xi,        
                                            
           Xi *  x*T 1,......, *T m  
                                                           1
                                                    1,......,        
                                                                     
                                             xi,T 1,......,xi,T  m
                                                                    


   and the OLS estimator  i  Xi Xi  1 Xi y i .
                           ˆ
   The vector of prediction errors is then ui *  (u i,T 1,......,ˆ i,T  m), obtained from:
                                                      ˆ            u

                              ˆ
           ui *  y i* X i *  i                                                                          (A.1.3)


   where yi * is the return on the firm over the test period, X i * is a typical m  2 OLS matrix of
                                            ˆ
market returns over the test period and  i is the vector of OLS estimated parameters. The
cumulative sum of forecast errors over the event window (T+m1, T+m2) is:
                                                     The Timing of Directors‟ Sharedealings and Company Valuation 31


             T m2
                 u i  Cui*
                  ˆ                                                                                    (A.1.4)
             T  m1

    where 'C' is an appropriately designed 1  n selection vector, as opposed to a m × 1 selection
                                                                                 ˆ
factor in the original article, which has the elements taking the value unity if ui is contained in
the event window and zero if it is not.

   The covariance matrix is given by:

                               
           Di  Xi Xi T 1Qi Xi Xi T
                                       1                                                            ( A.1.5)


   where Qi is an estimate of E Xi ui ui Xi T                      which can be approximated by:

                                                                                      
                     T                       p T
           ˆ      1                   1  
           Qi  T      x it xit u it  T
                                ˆ2                       xit xi,t  s  xi,t  s xit u it u i,t  s
                                                                                      ˆ ˆ              (A.1.6)
                    t 1                   s 1 t  s 1


    for p chosen to be approximately T 1/ 3 ; p should be increased until the truncations become
         ˆ
trivial. Qi is thus an estimate of the average of the variances of xit uit plus a term that takes into
account the covariances between xit uit and xi,t s ui,t s , the number of covariances being
truncated at s=p through the assumption of mixing (see, Mills, 1993, Chapter 5). In these
circumstances we have:

                              
           E  Cui * u i * C    i CC  CXi * Di Xi * C / T  V D i
                              
                                     2                                                                 (A.1.7)
                              


                                                                           
  and using this expression, an adjusted statistic Z D  N CAR m1, m 2 V D1 / 2 ~ N (0,1) , where

   V D i 1V D i
         N           . Equation A.1.7 does not include an erroneous X term, which is in the original
  formula for V D i in Coutts et al.

								
To top