{x} | Trading.live

The steps involved in back-testing are not standardized and can be adjusted based on the trader's available resources and the unique features of their trading model. It is important to note that back-testing can only evaluate the quantitative aspects of a trading model. If there are significant subjective analysis components in the trading model, the effectiveness of back-testing will be greatly diminished.
The first step in back-testing is to determine the size of the data source and select appropriate time frames based on the characteristics of the trading model. For instance, if a trading model generates only 20 trading signals in a year, it may require 5 years of data for an adequate sample size. However, if the model produces 50 signals in six months, back-testing with two years of historical data may suffice. Generally, brokers provide data for free if the data cycle is not particularly small. However, if high-frequency trading requires tick-level data, it may be a different story.
Secondly, in the initial stages of back-testing, a time-series approach can be employed to test the trading model. For instance, one can assess the performance of trading model A on January 1, 2015, and then test its performance on January 2, 2015, and so forth, until/>/>