Financial Solutions for Business Bloomberg
Task 1
According to Frazzini and Pedersen (2014) “Betting against beta”:
High-beta assets are generally over-priced and low-beta assets are under-priced. A trading strategy to capture this opportunity is to short high-beta stocks in the meantime long low-beta stocks. The strategy can be self-financed and involves holding a zero-beta portfolio. Constructing this portfolio involves two steps: 1. Holds a portfolio of low-beta assets, leveraged to a beta of one. 2. Shorts a portfolio of high-beta assets, de-leveraged to a beta of one. The combined portfolio should have a beta of zero and yield superior returns.
Based on stock price data from FTSE index, construct a zero-beta portfolio consisting of high and low beta stocks. You are required to:
- Demonstrate implementation of the model in R. This includes showing high beta as well as low beta stocks in the portfolio; explaining rules used for stock selection; building accounting functions that can automatically report daily VaR and various balance sheet items relate to financial reporting.
- Evaluate performance of the portfolio between 1st January 2021 and 30th June 2021. Performance should be evaluated on a monetary basis as well as on a return basis, assuming the initial portfolio value is £1,000,000. Performance measures like Sharpe ratios, VaR should also be discussed. Performance of the portfolio should also be compared against FTSE100 during the same period.
Task 2
One of the most well-known statistical arbitrage strategies is pairs trading, which involve investing in “correlated” stock pairs. To profit from the trading strategy, matching (or “correlated”) pairs needs to be found and closely monitored for correlation divergence. Once a divergence is detected, a long position should be entered into on the underperforming stock, at the same time a short position should also be entered on the overperforming stock. When the pair’s performance convergent, both positions should be closed. Based on stock price data from FTSE index, construct a pair trading strategy. You are required to:
- Demonstrate implementation of the strategy in R. The strategy should be semi-automatic with minimum parameter input; it should contain at least 20 matching pairs. The implementation should show rules of identifying pairs, as well as pairs that are found in the data.
- Backtest performance of the strategy between 1st January 2021 to 30th June 2021. Performance should be evaluated on a monetary basis as well as on a return basis, assuming the initial capital is 1,000,000. Performance of the portfolio should also be compared against FTSE100 during the same period.
This is an individual coursework; you must work on it on your own. You can submit the R code directly in the submission box. There is no word limit on the coursework, as long as your code complete the 2 required tasks. The coursework account for 70% of the module’s grade.
Rubric
Some rubric
| Criteria | Ratings | Pts | ||
| This criterion is linked to a learning outcomeCode quality
Model implementation in R needs to be transparent, verifiable and replicable. |
|
30 pts | ||
| This criterion is linked to a learning outcomeZero beta model
Understanding of the zero beta model, these include: identifying high and low beta stock; leveraging to create zero beta portfolio; calculating return and risk measures; reporting profit, VaR and other accounting information; comparing performance against benchmark; etc. |
|
20 pts | ||
| This criterion is linked to a learning outcomePairs trading strategy
Understanding of the pairs trading strategy, these include: identifying matching pairs; trading pairs based on pre-defined rules; calculating returns and risk of the strategy; evaluating performance against benchmark index; etc. |
|
50 pts | ||
| Total points: 100 | ||||
| Some rubric | ||||
