1.Estimate the following regression: ln () = 0 + 1 + 2 + 3ln () + 4() + by OLS and report results in the usual format. Are the slope coefficient individually statistically significant?

2. Interpret in words the estimate for 3.

3. If a firms debt-to-capital ratio increases by 1.5 and, simultaneously, its earning per share increase by 4.75, what is the predicted change in actual returns (i.e. not in the log-returns!), holding ln(netinc) and ln(salary) fixed?

4. Compute the 90%, 95%, and 99% confidence intervals for the intercept. What do you conclude with respect to the following hypothesis: If everything else were equal to zero, the predicted (base) log return would be -0.05?

5. Evaluate the null hypothesis that 25 times the effect of the firm's debt-to-capital ratio is of the same magnitude as the negative effect of log net income. That is, test the following null hypothesis: 0: 251 = 3 Test the null hypothesis against the one-sided left-tailed alternative, at a 5% significance level.

6. Evaluate the null hypothesis that the effect of is zero while at the same time (i.e. jointly) the effect of ln() is zero. Let the significance level be 0.1.

7. Evaluate the EMH. That is, test whether the independent variables are jointly significant at a 1%, 5%, and 10% significance level. Calculate the test statistic as well as the critical values (or the pvalues) manually.

8. Now assume that assumption MLR.6 does not actually hold, but observe that the sample size of your data set is large. Re-evaluate the hypothesis test from part 7. above, relying on the correct probability distribution.

9. Based on your test outcome in part 8., what can you conclude with respect to the EMH?