The 7 most expensive trading mistakes (and how to spot them)

Seven mistakes come up again and again in retail trade logs. They are not exotic, and they are not the things newsletters warn you about — they are mundane, repeatable, and each one leaves a specific signature you can spot in your own journal. What follows is the signature for each, the fix, and the cut you run to find out what it is costing you.

They are listed in no particular order of expense. Which of the seven costs you most is a question about your book, not a general fact, and the "what to measure" line under each one is how you answer it.

Mistake 1: Overtrading after a win

Symptom in the journal: you took 3 trades on Monday morning, won them all, and then took 11 more trades that afternoon. The 11 afternoon trades net-bled most of the morning's profit. Look for clusters of post-winning-streak frequency spikes.

Why it happens: a win triggers dopamine. Dopamine triggers the next trade. The 12th trade of the day is statistically far worse than the 3rd, because you've already taken the high-conviction setups; the marginal trade is the bottom-of-the-barrel setup.

What to measure instead of guessing the cost: group your closed trades by trades-per-day and compare expectancy in each bucket against your overall baseline. If the days where you took twice your median count have a worse expectancy than the days you didn't, that gap is the cost, in your own numbers. No published figure substitutes for it — your count threshold and your give-back are specific to you.

Fix: hard daily trade cap. "I will not take more than 5 trades per day." Tooling enforced; not willpower-dependent.

Mistake 2: Revenge-trading after a loss

Symptom in the journal: a -$400 loss at 10:15 followed by a $1,200 position opening at 10:18. The follow-up trade is 3x your average size, opened within minutes of the loss, often on a thinner setup. (Background on the underlying psychology in the tilt primer.)

Why it happens: the loss feels asymmetric. The brain wants the money back NOW. The trader sizes up to make the recovery feel emotionally proportionate. The math doesn't care about emotional proportion.

What to measure: isolate the trades whose entry came after a realised loss and compare their win rate and expectancy to your baseline. Sequence matters and so does the ordering you use — the comparison has to be built on realisation order, not entry order, or a scale-in gets counted as four revenge trades when the whole position stopped out at once.

Fix: 15-minute lockout after any loss exceeding 1% of account. Force a walk. Re-engage from neutral state.

Mistake 3: Size inflation after a winning streak

Symptom in the journal: position size on Friday is 2x position size on Monday because the week has been green. Track size as percent of account; look for the upward drift after winning sequences.

Why it happens: confidence compounds. You start the week 2 contracts; you win 4 days; on Friday you take 4 contracts because "the setup is the same and I've been hitting." The setup IS the same. Your psychology is not. (For the bias mechanic underlying this, see the loss-aversion primer.)

What to measure: take your top quartile of trades by notional size and compare its win rate and expectancy to the other three quartiles. The reading you are looking for is whether your biggest positions are your best ones. If the largest quarter underperforms the rest, conviction and size have come apart in your book, and the size column is where it shows.

Fix: fixed-percent-of-account sizing. Computed weekly, locked Monday morning, doesn't move.

Mistake 4: Fading the trend in the last hour

Symptom in the journal: short trades opened 2:30-3:30 ET on up days (or longs on down days) with sub-30% win rates. Filter by time-of-day bucket; look at counter-trend trades in the close window.

Why it happens: "this rally is exhausted, it has to reverse." Mean-reversion bias in a trending tape. The market is in a trend because flow is one-directional; the close hour often accelerates the trend, not reverses it.

What to measure: cut your history by hour of day, in market time rather than your browser's, and look at the last trading hour against your baseline. Then narrow to counter-trend entries inside that hour. Two cuts, both of which need a real sample in the bucket before they say anything.

Fix: rule against counter-trend trades 2:30 PM ET onward. Or no trades at all 2:30 PM onward; come back tomorrow.

Mistake 5: Ignoring a stop

Symptom in the journal: the trade's actual stop hit price is wider than the planned stop. Compare plan vs actual on each trade; look for the trades where actual exit was past the planned stop.

Why it happens: the stop is at -$200, the market touches -$200, the trader thinks "it'll come back," doesn't sell, the market keeps going, the trader sells at -$600. The original analysis was fine; the stop discipline failed.

What to measure: this is the one mistake on the list with an exact cost rather than an estimated one. For each trade you held past your planned stop, the damage is your realised loss minus the loss the planned stop would have taken. Sum that column. It requires a planned stop recorded before entry, which is the whole reason to record one.

Fix: bracket orders at the broker. Stop submitted at entry, attached to the position. Not in your head — in the broker's system.

Mistake 6: No pre-trade plan

Symptom in the journal: blank or generic "setup" tag, no documented entry trigger, no documented stop level. The post-trade notes describe what happened, not what was planned.

Why it happens: the trade was an impulse. You saw a chart, you felt the urge, you clicked buy. The journal entry is reverse-engineered after the fact.

What to measure: split your trades into the ones with a written pre-trade plan and the ones without, and compare win rate and expectancy. This is the cleanest A/B in a trading journal because you control the assignment, and it is also the one most traders refuse to run, because the answer tends to be unflattering.

Fix: pre-trade plan field required before trade opens. Keyboard-shortcut driven form: symbol, side, entry, stop, target, setup. Done in ~15 seconds. Now the trade has an articulable thesis or it doesn't get opened.

Mistake 7: Trading outside your defined hours

Symptom in the journal: trades at 4:15 AM ET, 11:30 PM ET, 1:00 AM Sunday. Filter by time-of-day; look for outliers far from your normal trading window.

Why it happens: you're up late, you're bored, the futures market is open, you take a position. There was no setup. There was no plan. There was a notification on your phone.

What to measure: tag the hours you have decided are yours to trade, then compare everything outside them against your baseline. Check stop coverage in the same cut — the off-hours trades that hurt are usually the ones entered from a phone with no stop attached, and that is a field you can count rather than a feeling.

Fix: explicit trading hours rule. Tooling that flags off-hours trades in the journal so the pattern is visible. Eventually: phone trading app deleted.

How TFQ's mistake-checker surfaces these

TradeFlow Quantum has a /mistakes page that runs most of the cuts above against your own history, and it is worth being exact about what it does and does not do. It slices your trades nine ways — hour of day in market time, day of week, setup, side, psych score, the trades straight after a loss, the trades straight after a win, the top quartile by notional size, and symbol — and flags a slice when its win rate is at least 15 percentage points below your overall baseline, or when its expectancy is negative. It needs 20 trades before it reports anything at all, 5 closed trades in a slice before that slice is eligible, 8 for a symbol, and it shows the ten strongest patterns by severity.

Mistakes 5 and 6 are not on that page, because they are per-trade comparisons rather than slices. They come from the pre-trade plan: if you record a plan before entry, the trade is checked against it afterwards for three specific drifts — held past the planned stop, exited a winner before the planned target, and realised a loss worse than the risk cap you committed to. Those findings appear on the trade's own page and in the scheduled digest email.

One number on the /mistakes page deserves a caveat rather than a headline. It shows the net P&L of every flagged slice added together, and that is not the cost of your mistakes — it is the total P&L of those slices, which includes the winners inside them and does not isolate how much of the loss the behaviour caused rather than the setup. Read it as a size indicator for where to look, not as a bill.

The point isn't to make you feel bad. The point is to make the patterns visible. Most traders are stunned the first time they see their own mistake-cost breakdown — the patterns are real and the dollar amounts are large. The fix is almost always tooling-level, not willpower-level.

Honest disqualifier

If you're already disciplined enough to follow your plan 95%+ of the time, the mistake-checker won't show you much. It's built for the rest of us — the traders who know what we should do and don't reliably do it. The system exists to compensate for the gap.

The /mistakes page reads whatever trades you have imported, so a broker connection on day one of the trial gives it a history to work with immediately. $15/mo. 7-day free trial.