This coursework involves:

1. Verifying stationarity of a given time-series using ACF test
2. Portfolio Optimization

• Volatility forecast using GARCH(1,1)

1. ARMA model

Consider any 8 stocks’ latest daily data for 1 year from FTSE100 (for example, use the data between Nov 2018 and Nov 2019).  

Tasks:

(A)

Let z_it represent the log return time series of a stock i, whereiÎ{1,!,4}. Find ACF and PACF for each of the z_it time series and verify the stationarity of 4 stocks of your choice from the selected 8 stocks. Discuss the results.

(20 Marks)

(B)

• Briefly explain mean-variance portfolio optimization                                     (10 Marks)
• Estimate the covariance matrix for the selected 8 companies’ stocks

(10 Marks)

• Plot R_P ands_P by creating portfolios using the selected 8 companies and the obtained

covariance matrix.                                                                                              (10 Marks)

(C)

Using the first 10 month data as in-sample data, estimate GARCH(1,1) parameters for the volatility forecast of any two stocks of your choice. Comment on the volatility forecasting efficiency of GARCH(1,1) model considering the final two month data as out-of-sample data.

You need to present the results from the above tasks in a report form (2500 words) for a fund manager who is interested in creating a portfolio using the selected 5 companies. In the report, you need to give a brief introduction on the methods and their limitations using the existing literature.

Report writing (references, format, academic writing style etc.)                                                                    (10 Marks)