This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Econ 104 - Problem Set 5 Lorenzo Braccini * November 3, 2011 Question 1 a) As sufficient condition for the coefficients to be unbiased (jointly) we can simply take, in this particular case, the usual assumption on the conditional mean of the errors, i.e.: E [ U i | X i ] = 0 b) The effect of a 1% change in PSML from 5% to 6% is given by: Y =- . 02 (0 . 06- . 05) + 0 . 001 (0 . 06 2- . 05 2 ) =- . 0001989 Hence the described change in PSML is estimated to decrease the profit of an Investment Bank by 0 . 01989%. Question 2 - S&amp;W 7.7(e) We are testing the null hypothesis that the coefficients for BDR and Age are jointly statistically equal to zero. Hence our F-statistic is ap- proximately distributed as a F q, distribution where the number of restriction, q , is equal to two. The critical value for rejecting the null hypothesis at a 10% level is, in this case, given by 2 . 30 (Table 4, S&amp;W pag. 753). Therefore, since F = 0 . 08 &lt; 2 . 3, the null hypothesis is ac- * [email protected] 1 cepted and the coefficients for BDR and Age cannot be said statistically...
View Full Document
- Spring '12
- Regression Analysis, Null hypothesis, Yi, X1i