How to test a trading strategy based on past data

Comments: 0Views: 47


Staff member

In brief

Think you have a great idea about the market but don't know how to test your idea without risking a loss? Learning how to test trading ideas against past data is an important job for a systematic trader.
The basic premise of testing against past data is that strategies that have worked in the past can work in the future. But how can you do this yourself? And how should you evaluate the results? Let's go through a simple testing process.


Testing is an important factor in developing your own charting and trading strategy. This process is accomplished by reconstructing past transactions using a system based on past data. Testing results based on past data will give you an overview of whether an investment strategy is working.
Before we go any further, Binance Futures is the right place for you to test your own strategies based on past data. If you would like to access past data from the platform, please fill out this registration form .

What is testing based on past data?​

First, if you want to dig deeper into past data-driven testing, read our article What is Past Data-Based Testing? .
In short, the main purpose of testing against past data is to show you whether your trading ideas are a good fit. You use past market data to see how the strategy is performing. If the strategy seems to have potential, it can also work in a real trading environment.

What to do before conducting testing based on past data​

Before we start with a test example based on past data, there are a few points you need to define. You need to determine what type of trader you are. Are you a hobbyist or a systematic trader ?
Voluntary trading is decision-based trading - traders will use their own judgment to know when to enter and exit the market. This is a relatively loose strategy with no fixed rules whereby most decisions depend on the trader's assessment of the available conditions. Therefore, testing based on past data will be less suitable for discretionary trading since the strategy is not strictly defined.

Of course, this doesn't mean that if you're a hobbyist, you shouldn't be testing against past data or paper trades. This just means that the results may not be as reliable as in the other case.
Systematic trading fits our theme better. Systematic traders rely on a trading system that will identify and tell them exactly when to enter and exit the market. Although they have full control over the strategy, the market entry and exit signals will be determined by the strategy. A simple systematic strategy would look like this:
  • When A and B occur simultaneously, enter the trade.
  • When X happens after that, exit the trade.
Some traders prefer this approach. This can eliminate emotional trading decisions and provide a reasonable level of assurance that the trading system is profitable. Of course, there are no guarantees.

This is why you need to have very specific rules about when to enter or exit positions in your system. If the strategy is not clearly defined, the results will also be inconsistent. Therefore, this type of trading is more popular with algorithmic trading.

You can also purchase historical data-driven testing software if you want to automate testing based on past data. You can enter your own data and the software will test against the past data for you. However, in this example, we will choose a manual testing strategy based on past data. Manual testing against past data takes a little more work, but it's free.

Please login or register to comment here.

Members online

No members online now.

Forum statistics

Latest member