Why might univariate factor tests not be sufficient to determine the factors in a fundamental model? Write the homework in a word file and attach the code.

R programming –finance

1. Based on Lab3a exercises, select three most significant factors from single variable regressions, and run a regression of returns on these three variables. Discuss the joint significance of these variables.

2. Based on Lab3b exercises, one can study the regression output and determine the significance of each factor. The following questions might help you in analyzing the regression output:

• (a) Is marketcap significant? Can you explain the sign of the coefficient?
• (b) Is price2cashflow significant? Can you explain the sign of the coefficient?
• (c) Is price2earnings significant? Can you explain the sign of the coefficient?
• (d) Is momentum significant? Can you explain the sign of the coefficient?

3. From the Lab3b data set, one can include a few additional factor exposures. For example, one might decide to include price2dividend instead of price2earnings, or roe instead of price2cashflow.

Estimate the fundamental factor model with log size, momentum, log price-to-book ratio based on fixed effects model with a. time fixed effects and b. time and firm fixed effects. Compare the estimates with the one we obtained above.

Note: # required library for running the function of felm() to fit a linear model with multiple group fixed effects

• install.packages(“lfe”)
• library(lfe)

4. Why might univariate factor tests not be sufficient to determine the factors in a fundamental model? Write the homework in a word file and attach the code.