Data Backtesting
Last updated
Last updated
Lubi supports back-testing of transaction data in the past 12 months. You can customize the range of back-test data and the frequency of back-test data according to actual needs.
Such as: real-time data (millisecond level), second level, 10 second level, 30 second level.
This backtesting system supports combined strategy backtesting. Users can input the range of data parameters required for backtesting according to their own needs. The system can perform combined backtesting on all parameters and output all test results to facilitate users to find the most suitable for themselves. strategy.
After the data backtesting task is created, a cloud host will be dynamically created for actual measurement, and eventually, the meter will be billed according to the complexity of the data.
Actual test cost = cloud host running time * unit price per minute
The current test is 0.02 USDT per minute, and less than 1 minute is counted as 1 minute.
Example: end time - start time = 68 minutes, test fee = 68*0.02=1.36U.
In order to ensure the smooth progress of the backtest, please make sure that there is a full amount of service fees in the Lubi account.
1.Create a backtest job
Taking the test of ETH U-standard long order as an example, first create a trading currency pair ETH, select the denomination currency, select "Long" for the trading direction, and select the backtest time. and the frequency of backtesting, click the "Create" button to create a backtesting task.
2.Edit Strategy
After successfully creating a backtest task, you need to set the strategy parameters of the task backtest, and click the "Edit/Review" button.
There are two types of backtesting methods: single group backtesting and combined backtesting
Single group backtesting: The value setting is the same as the strategy setting in the normal trading process. The backtest will use the set single strategy to backtest the data within the set time range, and output the corresponding results.
Combined backtesting: By setting the lowest value and the highest value for each parameter, as well as increasing progressive parameters between the values, the system will automatically count all parameters within the value range for cross-backtesting.
For example, ETH, the number of grids is set, the minimum is 100, the maximum is 1000, and the increment is 100. Then, the system will perform numerical tests of 100, 200, 300, 400, 500, 600, 700, 800, 900, and 1000.
The grid take profit ratio is set, the minimum is 0.01, the maximum is 0.03, and the increment is 0.005, then the system will test 5 values of 0.01, 0.015, 0.02, 0.025, and 0.03.
If other parameters are set, there is only one value, then this combination will test 10*5=50 groups in total.
Because the combination test can be set for 8 parameters, the default maximum number of combinations that can be tested in the backtest system is 400 groups/round. When the number is exceeded, the system will issue a warning.
3.Start Backtest
After setting the corresponding backtest strategy, please go to the "Backtest Center - Backtest Jobs" page, find the currency pair you want to backtest, click the "Start Backtest" button, and the system will conduct the test.
The specific duration of the backtest is closely related to the number of backtest combinations and the range of backtest data, and the duration varies from several minutes to several hours.
4.Review
When the backtest status is displayed as "Successful", you can click the "View Results" button to view the specific results of this backtest.
The results include the number of positions opened, closed, maximum grid, minimum unrealized profit and loss, grid and account returns within the test range.
All transaction details can also be viewed at the same time.
Tip: The backtest data only represents the trading situation of the strategy in the past time, not the future. Contract trading is a high-risk transaction. It is recommended to set a reasonable strategy according to your own situation to prevent large losses.
