ACCFIN5217 Financial Econometrics Assessment Brief | University of Glasgow

QUESTION/ DESCRIPTION OF ACTIVITY 

This is an individual assignment.

You are provided with a dataset called USstocks.csv. This dataset contains monthly stock returns of 982 U.S. Companies listed in the S&P 500 index spanning the period January 1980 - December 2015. Variables contained in this dataset are described below: 

• permno - Unique number that identifies a company
• year, or month - Identifies the time. 
• company name - Name of the company
• ri - Monthly Stock Returns of the company in % units.
• rm - Monthly Returns of the S&P 500 index in % units, a proxy for market return.
• rf - Monthly returns of 3 month U.S. Treasury Bills in % units, a proxy of the risk-free return. 

Form 10 portfolios using the method described in Tutorial 4. This procedure involves running 5-year rolling regressions to estimate beta and then sorting these beta values to form portfolio and monthly portfolio returns. Portfolio returns are the arithmetic average of individual stock returns. Start with the years 1980-1984 and get portfolio returns for each month in the year 1985. Repeat this procedure to construct portfolio returns for the sample period 1985- 2015. 

Unlike the dataset used in Tutorial 4, here you have incomplete data. Every few years, some firms drop out due to delisting and new ones get enlisted. To adjust for this you need to impose an additional condition when forming portfolios - For a firm to be included in a portfolio, it should have at least 24 months of returns data. This approach is consistent with Black, Jensen and Scholes (1972). For the 31-year sample of 1985-2015 monthly portfolio returns, conduct both the time series regression test and the cross-sectional test, and tabulate your results. Plot and discuss estimated portfolio betas against (annualized) portfolio returns.

Based on your test results, is there enough evidence to reject the CAPM? Explain.

Word Limit 1200 words (+/- 10%)