Committed to sharing quantitative strategies, training videos, Python, algorithm research and other related content.
Brief essay
Recently, some small partners have reported that they will not iterate strategies, and hope that a strategy can be directly installed and used in a phase. In fact, the SF series in the past can be sold for real, there is no future function and stealing price, no BUG signal WYSIWYG. It's just that everyone's risk tolerance and operation style are different. Try to use it after modifying the basis of the original strategy. If you are not familiar with the risks and principles of the strategy, and do not know how to operate, we do not recommend you to use it directly;
The SF13 strategy in this issue is one of the three strategies that small accounts are currently running. This strategy is mainly for trading on several varieties with small contract values, not a traditional trading strategy with multiple varieties and multiple cycles with the same parameters (small funds only Don't play CTA, you can't afford it, understand?). Of course, considerable restraint is maintained in parameter tuning. Within the effective range, the risk control parameters are fixed, and the tuning samples are separated from the test samples. Finally, the plain parameters for comprehensive performance are selected.
The principle of the SF13 strategy can be traced back to 2018. After 6 iterations,
A version was iteratively used for small accounts with small funds, and it was put on the new small account on April 20 for a firm sale.
About 11 trading days or so, the net value is close to 1.10. Among them, the contribution of the SF13 strategy is in the thread and PTA. At present, this strategy only trades the thread and PTA (small account capital limit), and the other two strategies run palm oil and methanol.
Let's talk about the characteristics of this strategy:
In principle, this strategy needs to run 6 varieties, namely apple, thread, thermal coal, PTA, PP, egg, and the maximum use margin is 35400. The varieties here are basically 3-4% of the initial stop loss, because the funds are small, so strict risk control is required. The core parameters have only one cycle parameter and one coefficient, and all the normal constants in the code do not participate in tuning.
Margin/Max Drawdown = 2.5;
Annualized return / maximum drawdown = 1.88;
Use cycle: 5 minutes cycle;
Trading Frequency: Midline Strategy
Related breed performance:
apple:
Thread:
egg:
PER:
Thermal coal:
PP：
In order to examine the performance distortion caused by code and data discrepancies, the strategy has been transferred to Mandarin 8, Pyramid, MC, Rebar as an example.
Rebar Wenhua (beginning in 10 years):
Rebar MC (starting from 10 years):
Rebar Pyramid (beginning in 2016):
Rebar TB (starting in 10 years):
After inspection, there is basically no big difference between transcompiling to other platforms, and the generality of the strategy principle is very good.
illustrate
This strategy has a filtering mechanism, so there are not many transactions, mainly to do some main band market. There will also be many cases of loss and profit taking in many shocks. It is a very common programmatic strategy that focuses on long-term positive expectations. For so many years, it is the survival rule in this market to find the long-term light warehouse. If you have any questions, please communicate in the group. The squirrel community focuses on quantitative trading and research.
Source code collection: TB (Transaction Trailblazer), Mandarin 8, MC, Pyramid
This strategy is only used for learning and communication, and investors are personally responsible for the profit and loss of real trading.