Part of The Trading Journal System series. This uses the tagged setup data built up over the past several Sunday reviews.
A hot streak is not an edge
Five winning trades in a row feels like proof you’ve figured something out. It might be. It might also be a normal winning streak inside a system with no real edge at all — the kind of thing that happens by chance to plenty of traders every month. The only way to tell the difference between “I found something” and “I got lucky” is enough trades to run the math, which is exactly what a weekly Sunday review can’t give you on its own. That’s what the monthly audit is for.
Where the Sunday review is about catching a pattern early and making a quick adjustment, the monthly audit is about stepping back far enough to see whether your actual numbers, over a real sample size, support a bigger decision — cutting a setup, sizing up on a proven one, or tightening a rule that keeps costing you.
Why 20 trades is the minimum
Twenty isn’t an arbitrary round number — it’s roughly the point where a string of wins or losses stops looking like it could just be noise. Below that, a 70% win rate over eight trades could easily be six lucky trades and two unlucky ones reversed. Investopedia’s guide on building a trading plan recommends testing a system through a meaningful run of trades before drawing conclusions for exactly this reason — small samples lie convincingly.
If a single setup tag has fewer than 20 trades logged this month, don’t make a permanent decision about it yet. Flag it for another month of data and audit the setups that do have enough volume.
The expectancy formula, in plain terms
Expectancy tells you, on average, how many dollars (or R-multiples) you can expect to make per trade for a given setup. The formula:
Expectancy = (Win Rate × Average Win) − (Loss Rate × Average Loss)
Say your “pullback-to-VWAP” tag shows 20 trades this month: a 45% win rate, an average win of 1.8R, and an average loss of 0.9R.
Expectancy = (0.45 × 1.8) − (0.55 × 0.9) = 0.81 − 0.495 = +0.315R per trade
That’s a real, positive edge — even with a win rate under 50%. This is the same math covered in the risk management series: a lower win rate with a favorable win/loss ratio can still be a profitable system, and a high win rate with a poor ratio can quietly lose money. You can’t tell which one you’re running without doing the calculation.
Now compare a second tag — “failed-breakout reversal,” 22 trades, 65% win rate, but an average win of only 0.6R against an average loss of 1.1R.
Expectancy = (0.65 × 0.6) − (0.35 × 1.1) = 0.39 − 0.385 = +0.005R per trade
That setup is winning almost twice as often but is essentially break-even after costs. Without the audit, the higher win rate would feel like the better setup. The math says otherwise.
Running the audit in four steps
1. Pull every trade from the past month, grouped by setup tag. This is only possible because you’ve been tagging consistently — another reason tagging isn’t optional.
2. Calculate win rate, average win, and average loss (in R) for each tag with 20+ trades. Skip any tag below that threshold for now.
3. Run the expectancy formula for each qualifying tag. Rank them from highest to lowest.
4. Choose exactly one structural change based on the ranking. Not five. One. Options include: stop taking the lowest-expectancy setup entirely, increase size modestly on the highest-expectancy setup (within your existing risk rules), or tighten entry criteria on a setup that’s marginally positive but inconsistent.
What a real rule change looks like
Following the example above, a trader running this audit has a clear, data-backed decision: the “failed-breakout reversal” setup is functionally break-even despite feeling productive because of the high win rate. The rule change for next month: stop taking that setup, or raise its confirmation bar significantly, and route that screen time toward more pullback-to-VWAP setups instead. That’s a specific, testable change — not a vague resolution to “trade better.”
Guard against overreacting to one bad month
The audit is powerful, but it can also mislead you if you treat one month as gospel. A setup with a long track record of positive expectancy that has one rough month isn’t automatically broken — check whether this month’s underperformance came from plan adherence (were you actually following the rules?) or from the setup itself. A proven setup traded poorly for a month should produce a discipline fix, not a setup cut.
The monthly menu review
A restaurant doesn’t rewrite the whole menu after one slow Tuesday. At month-end, the owner pulls the full ticket data and makes real structural decisions: which dishes actually made money, which specials looked busy but barely covered cost. That’s a monthly audit for a trading account: a large enough sample, reviewed honestly, that produces a real decision instead of a gut reaction to the last few nights.
Takeaway: Once a setup has 20+ tagged trades, run the expectancy formula before deciding whether to keep, cut, or scale it. Let the math override the feeling, and make exactly one structural change per audit.
Next up: Journaling Tools Compared: Spreadsheet vs. Edgewonk vs. TraderSync vs. Paper — which tool actually makes this whole system sustainable.
