You can work alone or in a group of up to 3 members. Each group needs to turn in a report describing and backtesting a strategy. You are free to choose any strategy. This year, I ask that your strategy trades at least 20 different stocks, or spans at least 3 different asset classes (e.g. stock, bond, derivatives, currency, cryptocurrency).

The grade will be based on quality of the report. My general suggestion for writing your report is to always keep your reader in mind when writing. A good report draws the attention of the reader, and prepares an answer for each potential question. For example, you can start with a motivation for your strategy, i.e. why do you consider this particular trading signal in the first place. Then you describe your strategy thoroughly enough to allow your reader to replicate your strategy. When reporting the performance of the strategy, try to create graphs in addition to the tables, e.g. a graph of its cumulative return together with that of the stock index, or a graph of drawdown. It is also a good idea to add some robustness analysis, e.g. how does the performance change when you adjust the parameters such as # of stocks in each portfolio. Finally, it is helpful if you can compare your results with the existing work in the finance academic literature. These are meant to be suggestions, not requirements.

I would like to share some of my observations from the past years: (1) Many students used value strategy (i.e. based on P/B, P/D etc). Note that value strategy doesn't work very well in the past ten years. If you'd like to backtest value strategy, please use a longer sample period. (2) Time series momentum is arguably the easiest strategy for the group project. However, please limit your testing asset in the future markets. (3) I encourage you to consider currency carry trade, which is also easy to implement. (4) A good source of data is Bloomberg. If you'd like to download a large panel of data, try Bloomberg Excel Add-in.