Turning an Idea Into a Testable Strategy
A strategy is not a feeling about the market. It is a set of rules precise enough that two people would trade it identically.
A trading idea is a feeling about the market. A strategy is a set of rules precise enough that two different people, given the same data, would take exactly the same trades. Almost all the work of system trading happens in the gap between those two things.
Until your idea survives being written down unambiguously, it cannot be tested, and until it is tested you do not have a strategy — you have an opinion.
Unambiguity is the requirement
Every condition must be answerable with yes or no on any given bar. 'Enter when the trend is strong' is not a rule. 'Enter when the 50-period moving average is above the 200-period and price closes above the 50-period' is.
If you find yourself needing to look at the chart to decide whether a signal occurred, the rule is not finished. That interpretive gap is where discretion creeps in, and discretion cannot be backtested.
Parameters: named, bounded, and few
Every number in your rules is a parameter. Give each one a name and a plausible range. A moving average period is a parameter; so is a stop multiple and a session filter.
The count matters enormously. Every additional parameter multiplies the number of combinations you can try, which multiplies the chance that the best combination you find is fitted to noise rather than to a real effect. Three or four parameters is workable; fifteen is not.
Write it before you look
The most common methodological error is developing rules while looking at the same data you will later test them on. You are not discovering a pattern; you are describing the chart in front of you and then congratulating yourself when the description matches.
Write the rules down first, from logic and observation of a different period, then test. If the test fails, revise the logic, not the parameters — and test again on data you still have not used for development.
Define the exits before the entries
Beginners obsess over entries and treat exits as an afterthought. The exit determines your average win, your average loss, and therefore your expectancy. The entry mostly determines how often you get to find out.
Specify the stop, the target, the time-based exit if any, and what happens to a position when the signal reverses. All of it, in writing, before a single test runs.
Build
Create and backtest your systematic trading strategies without coding.
Key takeaways
- Every rule must be answerable yes or no without interpreting the chart
- Keep parameters few — each one multiplies the ways you can fit noise
- Write the rules before looking at the data you intend to test on
- Exits determine expectancy more than entries do; define them first
- Every entry, exit, and filter must be unambiguous
- Parameters must be named, bounded, and few in number
- If a rule needs interpretation, it cannot be tested
- Write the rules down before you look at more charts
Fitting rules to the chart currently on screen, then calling the result a strategy.
Reading is not verification
Take the concept you just read and turn it into explicit rules, then test it. That is the only way to know whether it actually works.
Build a strategy