8342937137

MEANS OF PAYMENT AND BIDDING FIRMS' STOCK RETURNS IN I I1E UK

sample contains hiddcr firms engaged in success-ful iakcovers for the period January 1985 tojuly 1988, for which daily share returns were calculated for eleven days surrounding the announcement

day.

4

The announcement day is tłu* day the bidder makes its first offer for the target. The takeover ofFer is considered successful if the bidding firm accjiiires the required interest in the target firms common equitv (i.e. equal to or morę than 30 percent). The offers studied were not preceded by a mergcr altempi.

The sample was identified by a search of vari-ous financial publications such as Acquisition M&nthly, hwestors Chronicie, Business Research Index, Fair and Trade Publications, Stock Exchange Yearbook and financial newpapers.

The selection for the sam pled firms was sub-jected to the following criteria:

(a)    The takeover offer is considered successful if it isdeclared unconditional as to acęeptances.

(b)    Bidders are listed on the London Stock Exchange and drawn only from alpha and some beta stocks to mitigate the non-synchronous trading problem.

(c)    To filier the confounding efTects ofexogenous events on bidder returns during the eleven days surrounding the event. the bidders sampled did not experience any other major corporate evcnt al the time of the announcement of the offer. Major corporate evcntsare defined as death orappointment of key executives, announcement of financial reportsor new investment program mes l or the elcvcn days surrounding the announcement.

(d)    Daily share prices for each firm were available for at least 58 days before and 58 days after the announcement day.

The overall sample was divided into stock cxchange offers, if the takeover was fmanced wholly through an exchange of common stocks, offers fmanced w holly by cash and combination offers fi-nanced by cash and shares.

The finał sample contained 90 bidder firms, composed of thirty common stock offers, thirty cash offers and thirty combination offers where stocks represented between 28 and 70 percent of the transaction.

Methodology

I he research hypotheses examined are tested by applyingan event-type methodology similar to that described in Fama, Fisher, Jensen and Roli (1969)

and I ravlos (1987). The ordinary - least squares coefficients of the market model regression were estimated over the period n = -58 to n =-6 and n =

+6 to n = +58 relative to the day of the of the initial

✓

announcement. Daily abnormal common stock

/

returns were calculated for each firm i over the interval n =—5 to n = +5. For a sample of N firms, the daily average abnormal returns (AR) for each day n are computed as follows:

AR„ = Xj\,(Rin - a, - j8,Rmn) (l)

n = day -5,0,...., +5 N = the number of firms in the sample

Where R is the return for common stock of firm

in

i, is return for the market approximated by the Ali Share Market Index on day n, a., B. are ordinary-

^y    i i    *

least scjuares estimates of the market model parameters (adjusted for non-synchronous trading problem using Dimsons (1979) aggregated coefficient approach).

The daily returns were calculated from stock prices after adjustment for stock splits and divi-dends. The returns for each firm were further adjusted for any significant first-orderauto-corrcla-tion using Cochrane-Orcutfs (1949) quasi-first-ditTcrence approach before calculating the abnormal returns.

Average cumulative abnormal returns ((lARj j _{I 2}) were deriyed by summing the AR’s over the eleven days surrounding the announcement. The expected values of AR _j and CARp_|T9 are zero in the absence of abnormal performance. The test statistics of ARn* and CAR_TIT9 are based on the average standardised abnormal return respectively, where

CAR.

riT2

(Rin -    - /JjRmn)

Sin

T2

m = Tl

The estimation period is from n =-58 to n =-6 and n = +6 to n = +58 and the analysis period is from n = -5 to n = +5.

The standardising factor S_jn mitigates the problem of non-constant yariance of the residuals in the estimation and analysis period, and computed as follows:

Sin =

S² (1 + — +

(Rinn - Rm)

L Xk = i (Rmk - Rm y

W i

(4)

PF.RTANIK.AVOL. 14NO.I, 1991

95