SAIF Futures Quantitative Course I - Course Project - Futures Arbitrage
The system aims to back test Futures arbitrage strategy for China commodity futures trading. Basic philosophy is to identify the price gap between main contract and sub-main contract.
- Data extraction and cleaning
- Extract historical market data for all products. (e.g. 橡胶，螺纹钢，铜，股指期货IC，IH，IF... etc)
- The granularity of the market data should be at least as finer as minute-wise.
- Market data cleaning must be done prior to algorithm implementation so that we'll have main contract market data align with that of the sub-main contract.
- Overall back testing cycle should be from the data of past 6 months.
- Data processing algorithm
- Calculate the price gap between main contract and sub-main contract for each discrete point (close price for a certain minute) , denote the gap as priceDelta
- Calculate MA of priceDelta for past 20 units (minutes), denote as MA20
- Calculate SD of priceDelta for past 20 units (minutes), denote as SD
- Generate a band with upper as SDx2, middle as MA20, and lower as SDx-2
- Arbitrage algorithm
- Order shall be placed for current main contract
- Open buy when price hits the lower bound, and close sell when it regresses to middle.
- Open sell when price hits the upper bound and buy close when it regresses to middle.
- Back testing statistics
- Based on the behavior described in 3, filter out those don't follow the strategy which leaves the ones applicable for arbitraging.
- Calculate the statistics, net value, sharpe ... etc(TBD)
- Generate the report (TBD)