This project is to update live positions and trade results of automated paper trading of futures contracts of Micro E-mini S&P 500 Index.
The strategies used are constructed using XJ-Strategist, a machine learning and AI system capable of constructing and validating sustainable trading strategies, as introduced and showcased here.
Position and returns data here are automatically updated daily just after the market close of CME directly from my trade server.
2024-12-20 21:57:00(UTC), MES H25.CME_Eq, Price = 599475, Position = 33
Start of plot is 2023-10-19
*** Due to issues with broker platform, positions and results during Aug 4-13, 2024 are based on back-filled data. ****
The following file contains dates (of trade opening), daily closing prices, positions at daily market close, daily returns, and cumulated returns.
The daily prices are of a continuous contract that is formed by rolling and stitching previous contracts with the current contract. The algorithm of forming the continuous contract can be found here.
A contract of Micro E-mini S&P 500 Index futures is valued at $5 times S&P 500 Index. At the trading platform of the brokerage, the quoted index of S&P 500 Index comes in as 100 times the actual S&P 500 Index, thus a contract is valued at $0.05 times the quoted index.
The maintenance margin requirement is 5%.
I use a maximum of $50,000 as margin to enter into trades. So the trade value (based on 5% margin) is equivalent to a maximum of $1 million.
As an example. if the quoted index is 400,000, then at maximum my position would be (50,000 / 0.05) / (400,000 * 0.05) = 50
"maximum" here means trades are not always using the full capital, but depend on an algorithm of position sizing at the time.
Performance in training period (2000-01-03 through 2009-02-23):
Performance on entire dataset, in-sample and out-sample, (2000-01-03 through 2024-02-29. Note the training section at the beginning of the data set):
