A trading strategy describes the logic for finding and exploiting an opportunity; a trading system turns that logic into a repeatable operating process. The terms are often used interchangeably, but separating them makes testing and execution clearer.
A strategy may say, for example, that a certain market behaviour creates an opportunity. A system must go further: it must define when the idea is eligible, how a position is entered, how much can be risked, how the position is managed, how it exits and how the evidence is reviewed.
A strategy answers why and when
The strategy is the market hypothesis. It explains the behaviour the trader expects to exploit and the conditions under which that behaviour should matter. That logic should connect to a defensible trading edge, not merely to a preferred indicator or chart pattern.
A strategy can be discretionary or highly rule-based. What matters is that its central claim is understandable enough to test against evidence.
A system answers how the strategy is operated
The system adds the operating rules. It defines the setup, entry conditions, exit conditions, position sizing, risk limits and permitted trade-management actions.
It should also define how trades are recorded and how the strategy will be reassessed. Without that layer, two traders can claim to use the same strategy while producing very different decisions.
The trading plan documents the system
A trading plan is the written document that captures the system and its operating boundaries. The plan may include more than one strategy, but each strategy should still have a clear role and evidence base.
This distinction matters because a trader can improve documentation without improving the strategy, or improve the market hypothesis without fixing inconsistent execution.
Risk rules are part of the system, not the edge
Position sizing, stop rules and exposure limits determine how capital experiences a strategy. They do not create the underlying edge. A weak strategy does not become strong simply because it uses careful risk management, although disciplined risk control may reduce the damage from being wrong.
Likewise, a strategy with favourable historical statistics can still be unusable if its execution costs, leverage or drawdown are incompatible with the capital objective.
Testing should preserve the distinction
When a system is backtested, the researcher should know which elements belong to the market hypothesis and which belong to implementation. Otherwise, changing entry filters, stops, sizing and exit rules simultaneously can make it difficult to know what actually created the historical result.
That problem becomes more serious when many combinations are tried and only the best one is retained.
Discretion does not remove the need for a system
Discretionary trading can still be systematic at the process level. A trader may make qualitative judgments about context while using fixed risk limits, review rules and eligibility criteria.
The useful question is not whether every decision can be coded. It is whether decisions can be described clearly enough to distinguish intentional judgment from unplanned improvisation.
Common strategy-versus-system mistakes
- calling a single entry trigger a complete strategy;
- calling risk management the trading edge;
- changing several system components at once and attributing the result to one idea;
- using “discretion” as a label for undocumented exceptions;
- testing a system without separating market logic from implementation effects.
Strategy and system inside the MFXG framework
The Trading Systems framework separates the market idea from the machinery used to execute it. That separation makes both easier to test, review and improve.