A Clean Loser That Was Still Correctly Traded
This session lost money and contains no mistakes. That combination is uncomfortable enough that most people, reviewing it later, will invent a mistake in order to explain the loss. The invention is the actual error, because it produces a rule change that the evidence does not support and that will then apply to every session afterwards.
The Setup Passed Every Filter
The calendar was clear. The range formed at an ordinary height, neither compressed nor stretched, and its shape was balanced with both edges tested more than once. Every question the plan asks before a trade returned an acceptable answer, in the same order and with the same checks as on a session that works out.
This is worth dwelling on because it is the entire point. The filters are not there to select winners, which is not something filters can do. They are there to exclude the situations where the arithmetic is against you before the trade begins. A session passing all of them can still lose, and if it could not, the filter would be a prediction rather than a filter.
The Entry Filled and Went Nowhere
The resting order triggered on a break of the upper edge, filled slightly beyond the trigger, and price moved a short way in the intended direction. Then it stalled.
There was no dramatic reversal and no obvious moment at which the trade was clearly wrong. Price simply stopped extending, drifted sideways above the level, then a little below it, then back inside the range. The absence of a dramatic signal is characteristic. Most losing breakouts do not announce themselves, they just fail to continue, and the failure is only visible in retrospect.
The Decision Not Taken
The interesting part of this session is a decision that was available and was not made. Once price returned inside the range, the position was clearly not doing what it had been entered to do, and closing it there would have produced a smaller loss than the stop eventually did.
The plan did not contain that action, so it was not taken. That is defensible and it is not obviously optimal. An exit rule that closes a position when price re enters the range is a perfectly legitimate design, and had it been in the plan it would have been correct to use it here. What is not legitimate is inventing it mid trade, because the version invented mid trade is the version that fires whenever a position becomes uncomfortable, which is most of the time and includes a great many trades that go on to work.
The correct home for that idea is the review. Test it against the whole record, adopt or reject it as a rule, and then apply it to every session rather than only to the ones where hindsight shows it would have helped.
The Stop Executed
Price continued through the range and reached the opposite edge, where the stop was waiting. It filled a little worse than the level, for the same structural reason entries do, and the realised loss came out slightly larger than the planned figure.
That gap between planned and realised loss is worth recording rather than shrugging off. It is small on any single trade and it is systematic, which means it belongs in the expectancy calculation. A strategy evaluated on planned risk rather than realised risk is being evaluated at a discount it does not receive in practice, and the discount is applied to every trade in the record.
Grading the Trade Instead of the Outcome
Reviewed against the specification, every step here was correct. The session qualified, the range qualified, the entry was placed as written, the size followed from the stop distance, and the exit was the one that had been planned. The grade is clean and the money is gone, and those two facts coexist without contradicting each other.
The value of grading this way is that it makes a losing run diagnosable later. A month of clean losses is a sample from a distribution and calls for patience. A month of losses on trades that were early, oversized or exited by hand is not information about the strategy at all and calls for something else entirely. Without the grading, both months look identical on the equity curve, and the response to them is usually the same undirected panic.