In CW1, you have been asked to prepare a report for the manager of a new hedge fund detailing at least two types of technical indicators. CW2 is mainly about improving such technical indicators and back testing your trading idea with the historical data. Using one of the following three asset classes, i.) Stocks which can include indices, ii.) Currencies or iii.) Commodities, you’ll be asked to test a short term investment strategy for a single or portfolio of at least 3 assets. In either case, use an initial capital of $10 million or equivalent in another currency.
1.[10 pts] Write an executive summary about backtesting with a discussion on potential drawbacks and importance of this approach for algorithmic trading.2.[20 pts] Explain your strategy and technical indicator in details. The coursework should encompass the selection and rationale behind the strategy. Please note that you are required to use of the indicators covered during the lectures.
3.[10 pts] Describe your daily dataset, which you should obtain from sources such Bloomberg, or Capital IQ, and explain any pre-processing you have done and how you have cleaned it.
4.[10 pts] Describe your steps how to backtest the strategy over a previous period of last 12 months.
5.[25 pts] Present your findings on backtesting which must include an estimate of trading costs. Note that you will be using past 12 months of daily data for backtesting but your strategy must also be tested on a forthcoming daily or weekly period (out-of-sample test) in April or May 2018. Demonstrate how you have tried to optimise the strategy for performance and why you believe that the strategy is not over-fitted.
6.[15 pts] Compare your chosen indicator with another similar indicator. Show how you have tried to minimise trading costs. Present your portfolio (or single asset) returns.
7.[10 pts] Give a detailed assessment of the performance of the strategy. Here, plot and show time periods where your strategy is profitable. What do you observe? Explain.
8.[Bonus — Harder — 10 points] Turn your back-testing into a real-time algo strategy which reads data from Yahoo API and makes decision in real-time.
The coursework is expected to be referenced against authoritative and/or academic papers.
However, in the analysis of the results (both for the backtest and the out-of-sample test), your own reflections upon what went well and what did not go well (and why) should be stated.
Word limit: 1,500 excluding tables, references and appendices
The report should include an executive summary at the start but a table of contents should not be included.
Harvard style referencing should be used.Word limit: 1500 excluding tables, references and appendices.
You should also include your Python code to demonstrate how you have carried out the key steps in your assignment.