There are some problems in backtesting, such as price slip, over-optimization of parameters, price theft and so on. For example, when backtesting is based on historical data, the closing price is fixed, and your strategy can be bought at 100%. In real transactions, the closing price is the latest price and is dynamic. For example, gold card 10 5-day moving average buy 1 hand. Because on this day, it is possible that the signal of the morning and afternoon of the 5-day moving average of Jinka 10 disappears, while you buy according to the strategy in the morning and the signal disappears in the afternoon, which is the difference between the firm offer and the back test. There are many examples that need to be solved one by one, and it is too easy to formulate a set of strategies with high return test data but no practical use.