Essay on Data and Methodology

Submitted By minnaface
Words: 1200
Pages: 5

Data and Methodology
We are assessing the market reaction for the event of an acquisition, for GKN Plc and Vitec Group Plc and (the acquiring companies), separately and together. GKN Plc announced on Monday 18th July 2011 that it had acquired the entire issued share capital of Stromag Holding GmbH (Stromag). Vitec Group plc (Vitec) announced on 30 September 2011 that it had agreed to acquire the entire share capital of Haigh-Farr Inc.

We chose the two firms from the Investegate website. The historical data, which includes the announcement date, the stock prices of the companies before and after the event, and the corresponding indexes, were obtained from Bloomberg. The recently published paper was discovered via the Factiva research tool. We used the FTSE 100 for the index.

Event period
The event period determines the number of days over which we measure the possible abnormal return caused by the event. The event window is set to 81 trading days, including the event day; 40 days before and 40 days after it.

Estimation period

The estimation period estimates the expected return of the stocks. In our event study, it is set to 200 trading days when the weekends are excluded, 41 days before the event.

An event study investigates the effect of a particular event on the stock price of a firm (the dependent variable). The key assumption of this event study methodology is that the markets are semi-strong form efficient. The methodology we used studies the changes in stock price beyond expectation (abnormal returns), over a period of time; the event period. A model based on the estimation period is used for the estimation of the required rate of return for the firm’s stock and then we obtain the errors of such model, which are abnormal returns, i.e. returns beyond what would be predicted from market movements alone.
1) Calculating abnormal returns: Using the market (index) model [or the market adjusted return model (α=0 & β=1) or the mean adjusted return model (α=R & β=0)] Rt=α+βRmt+ εt t=1,2,…, T,
Rt = Return on the stock at day t
Rmt= Return on the market index at day t α= Intercept of the regression for the stock β = Beta of the stock εt = Unsystematic/unexplained component of the regression
T= 200 trading days, 41 days before event

We obtain the abnormal return (ARt) for the stock at time t, given by:
ARt = Rt-Rt=Rt-( α+βRmt) NEED HAT

We then tested the significance of the average abnormal returns and cumulative average abnormal returns using the t-test. This standard statistic relies on the assumption that the firm’s abnormal returns are normally distributed.
e.g. Day 0:
H0=AR0=0 No significant abnormal return
H0=AR0≠0 Significant abnormal return t=AR0/SD(AR) t=AR0/SD(AR) t=AR0/SD(AR)≈N(0,1)
SD (AR)=√[∑-41, t= -240 (Art –1/200∑-41, t= -240 Art)^2]/199

2) Obtaining cumulative abnormal return over t1 to t2: Isolating the part of a stock movement that is attributable to the effects of the event is difficult as there may be information leakage and/or delayed reaction. We therefore summed all abnormal returns over the event period to obtain the CAR.
CARt1, t2=∑t=t1, t2 Art
e.g. Days-40 to +40:
H0:CAR-40,+40=0 No significant cumulative abnormal return
H0:CAR-40, +40≠0 Significant cumulative abnormal return t=CAR-40, +40/√81SD(AR)≈N(0,1)
3) For both firms together, N firms, calculate average abnormal returns, AARt and cumulative average abnormal returns,CAARt. We obtained the AAR and CAAR and continued as before, with AARt to replace ARt, and CAARt to replace CARt.
AARt=(∑i=1, N ARit)/N
CAARt1, t2= ∑t2, t=t1AARt
By using these figures calculated, we can determine whether there is an abnormal stock price effect associated with an event, and thus determine the significance of the event.
The limitations of the analysis

There are 5 main limitations in our analysis of the correlation between acquisition announcement and the stock price.

1. The fist limitation is that