One-Variable Rule
Published Last updated
Key Takeaway
A strategy refinement discipline requiring that only one element of a trading system is changed between testing phases, ensuring each modification's isolated impact on performance can be accurately measured.
Learn These First
What Is One-Variable Rule?
A strategy refinement discipline requiring that only one element of a trading system is changed between testing phases, ensuring each modification's isolated impact on performance can be accurately measured.
How One-Variable Rule Works
Frequently Asked Questions
What is the one-variable rule in trading strategy development?
The one-variable rule is a refinement discipline stating that only one element of a trading strategy may be changed between testing phases. It applies the logic of controlled experimentation to strategy development: if you change multiple components simultaneously and performance shifts, you cannot determine which change caused the outcome. By isolating each modification across its own complete test cycle, you build a traceable record of which changes improved performance, which were neutral, and which caused deterioration — making strategy refinement a systematic, evidence-based process rather than guesswork.
Why is changing multiple strategy variables at once a problem?
Changing multiple strategy variables simultaneously creates compound unknowns that make the results of the next test cycle analytically uninterpretable. If you alter your entry filter, tighten your stop methodology, and reduce your position size at the same time — then performance improves — you have no way to know whether one of those changes drove the improvement, two of them working together did, or all three were necessary. This ambiguity means you cannot replicate the improvement reliably, cannot reverse a harmful change selectively, and cannot build a coherent understanding of what your strategy actually requires to generate edge.
How do I decide which variable to change first when refining a strategy?
Prioritise the variable most directly linked to the performance problem identified in your journal review. If execution gap analysis shows that your exits are consistently the largest source of deviation — for example, closing trades before system signals due to open-profit anxiety — then the exit rule is the highest-priority variable for the next testing cycle. If compliance tracking shows the primary failure point is entry criteria ambiguity, address the entry specification first. The one-variable rule does not dictate which variable to change — it requires that whichever variable you identify as the priority is the only one changed in the next phase.
Common Misconceptions About One-Variable Rule
The one-variable rule slows strategy development too much to be practical.
The one-variable rule extends individual testing cycles but dramatically accelerates the development of a genuinely effective strategy by eliminating unattributable changes. Multi-variable changes appear to move faster — more modifications per phase — but produce results that cannot be interpreted, requiring further testing to untangle the effects. The one-variable approach builds a clean, documented modification history where each phase adds a clear, verified insight about the strategy. The cumulative knowledge built this way is directly actionable in a way that multi-variable testing results never are.
If the strategy is clearly broken, it is acceptable to change multiple variables at once to fix it.
A strategy that appears clearly broken still requires one-variable discipline during refinement, because the perceived cause of the breakdown may not be the actual cause. Traders under performance pressure frequently misdiagnose problems — attributing strategy failures to rule elements when the actual issue is execution compliance, or vice versa. Making sweeping multi-variable changes in response to this misdiagnosis compounds the diagnostic error and creates a fundamentally different strategy from an unresolved analytical starting point. One-variable refinement remains the only method that produces interpretable diagnostic results regardless of how severe the performance problem appears.
The one-variable rule only applies to quantitative or algorithmic trading strategies.
The one-variable rule applies to any rule-based strategy — quantitative, systematic discretionary, or hybrid — because the underlying problem it solves is universal: multi-variable changes produce unattributable outcomes in any testing context. Whether a strategy's rules are expressed in code or written plainly in a document, changing multiple elements simultaneously produces compound unknowns that prevent accurate performance attribution. Discretionary traders who do not follow this discipline are particularly vulnerable, as they already face challenges separating strategic performance from execution behaviour without introducing additional attribution ambiguity through simultaneous rule changes.