Prueba

UNIVERSITAT POMPEU FABRA
EMPIRICAL CORPORATE FINANCE

Angel Samaniego A. angel.samaniego@upf.edu

EVENT STUDY

20 June 2006

Empirical Corporate Finance.

Answer 1
a) Event definition: The event of interest of this analysis was to study the effect on stock prices front the announcement of equity offering. The event window was 1 week before the announcement and 2 weeks after theannouncement (16 days). b) Selection criteria: The analysis was over 5 firms listed in the Spanish stock exchange. The announcement will be in 5 different years and different industry representation to don’t have potentials biases with possible simultaneous announcement effect and have most of the industry representation. We didn’t take into account the firm market capitalization to reflect possibledifferent between small firms and value firms.
Table 1: Firms selection.
Telefonica Event day Event window 26-Mar-98 16 Banco Guipozcoano 22-Apr-99 16 Aguas de Barcelona 2-Jul-92 16 ERCROS 17-Nov-95 16 ACERINOX 28-Sep-94 16

c) Normal and abnormal returns: For each firm j and event date τ we have (from class notes):

ε*jt = R jt − E R jt X t

[

]

where ε*jt , R jt and E R jt X t arethe abnormal, actual and normal returns, respectively. € For modeling the normal return we used the market model where X t is the Spanish market

[

]

€ return (Ibex ).
We are assuming a stable linear relation between the market€ return and security return (iid).
Table 2: Abnormal returns.
Event study Telefonica Day 5 Day 4 Day 3 Day 2 Day 1 Event day Day 1 Day 2 Day 3 Day 4 Day 5 Day 6Day 7 Day 8 Day 9 Day 10 CAR 0.385% -1.194% -0.002% 0.626% 2.112% -1.080% -1.544% 2.944% -1.241% 0.247% 0.151% 1.036% 0.077% -0.117% -0.539% -2.014% -0.154% Abnormal returns Banco Aguas de ERCROS ACERINOX Guipozcoano Barcelona -0.908% -3.746% -0.100% 0.489% 1.487% -0.404% -4.252% -0.179% -2.199% 0.057% 1.153% 0.355% 2.409% -2.936% 1.148% -0.035% -0.234% -1.965% -0.500% -0.015% -0.870% 0.336% 1.427%-3.328% -1.043% 1.668% 2.981% 0.464% -0.435% -0.613% 0.535% 1.776% 0.486% 0.283% -4.887% -0.817% 0.918% -1.150% -3.504% 0.335% -0.413% -0.084% -2.058% 0.129% 0.861% -0.830% -17.240% -0.346% -0.240% -0.155% -3.614% -0.260% -0.276% -2.085% 4.164% -1.016% 0.002% -0.648% 7.285% -0.535% -0.538% 6.210% 0.104% -0.445% -0.995% -6.062% -17.358% -3.430%

2

Empirical Corporate Finance. d) Estimationprocedure: The estimation window was 1 year or 250 days prior to the event window. e) Testing procedure: The parametric tests proposed in the literature remain behind on the assumption that individual firm’s abnormal returns are normally distributed. We used the cumulative abnormal return (CAR) to accommodate multiple sampling intervals within the event window. The purpose of standardization is toensure that each abnormal return will have the same variance. By dividing CAR by its standard deviation, each residual has an estimated variance of 1.
ˆ SCAR j (τ1, τ 2 ) = ˆ CAR j (τ1, τ 2 ) ˆ σ j (τ1, τ 2 ) # days

ˆ The σ j  is an estimate of standard deviation of the average abnormal returns over the event
€ window (τ1, τ 2 ) . The#days is the number of days into the event window.

€

ˆ Under the null hypothesis the distribution of SCAR j (τ1, τ 2 ) is student t with n-2, degrees of € freedom, where n is the number of days in the event window. € To test the event study is the same process with the different of taking average, for N events

we have
CA R j (τ1, τ 2 ) = 1 N ˆ ∑ CAR j (τ1,τ 2 ) N j=1

1 N ˆ ˆj σ 2 (τ1,τ 2 ) = ∑σ 2 (τ1, τ 2 ) N j=1 CA R j (τ1, τ 2 )

(

ˆ σ 2 (τ1, τ 2 )

)

1

2

≈ N (0,1)

€ f) Empirical results: In graph 1 shows the CAR for the 5 firms and it clear the undervaluation

of ERCROS and Aguas de Barcelona.

3

Empirical Corporate Finance.

Graph 1: Cumulate Abnormal Returns
0.050 0.000 -0.050 -0.100

CAR

-0.150 -0.200 -0.250 -0.300 -0.350

da y

3...