OptiNod Academy
Revenge trading and overtrading — Extra trades after a loss can turn expectancy negative
After a loss, entry standards may loosen and trade count rise. Per-trade costs can then turn even a positive-expectancy system negative in practice.
Right after a loss, entry standards can slip and trading frequency rises. The cost attached to each extra trade reduces positive expectancy, and the trade record shows by how much.
The cost of overtrading appears in a 2000 study by Brad Barber and Terrance Odean. They examined six years of trading in roughly 66,000 individual accounts at a large discount broker. The most active 20% of households earned about 11.4% in annual net returns, well below the roughly 17.9% market index return over the same period. Gross returns differed less from those of less active investors; trading costs explained much of the net gap. Revenge trading is overtrading concentrated right after a loss. Pressure to recover makes entry standards looser and packs more trades into a short interval.
It is often described only as an emotional reaction to losing. That frames the remedy as “calm down,” which may fail again after the next loss. The cost of revenge trading is not measured by the intensity of the emotion. It is the sum of per-trade costs on the extra trades, visible in the record.
Net expectancy per trade equals gross expectancy minus round-trip cost. Express the cost in R by adding round-trip fees and slippage, then dividing by the planned stop distance. Trades added after loosening entry standards may have gross expectancy near zero or below it, but they incur the same round-trip cost as validated entries. More trades can therefore push account expectancy negative in proportion to those costs.
Pressure to recover after a loss loosens entry standards
Revenge trading begins with risk seeking in a loss. Immediately after realizing a loss, pressure to recover it grows. That pressure reinforces the risk seeking discussed in loss aversion. The pain of the loss makes even a small chance of getting back to break-even feel worth pursuing, so the trader enters again quickly.
The re-entry is different from an entry that waited for the signal. A validated system enters only when its specified conditions are met. Under recovery pressure, the next trade may be taken before those conditions appear, on the belief that “this one will get it back.” Lower standards mean the trade does not have the edge measured for the validated setup. An entry meeting only half the criteria can have gross expectancy near zero or negative.
The pressure may not stop with one trade. If the loose re-entry also loses, the urge to recover grows and the interval before another entry shrinks. Loss creates pressure, pressure lowers standards, and lower standards produce more trades. This cycle can multiply frequency over a short period.
Each trade's costs reduce expectancy by a known amount
More trades reduce expectancy through costs. Every entry and exit incurs fees and slippage regardless of whether the trade wins. These trading costs can be larger live than a backtest assumed and contribute to the execution gap.
Converting them into R makes their size clear. Suppose taker fees on perpetual futures are 0.05% each way, or 0.10% round trip, and slippage is 0.03% each way, or 0.06% round trip. Total cost is about 0.16%. If the stop is 2% away, 1R is 2%, so round-trip cost is 0.16 ÷ 2.0 ≈ 0.08R. Opening and closing every trade costs 0.08R regardless of its outcome.
That 0.08R is already accounted for in a validated trade. An entry with +0.20R gross expectancy has +0.12R net after costs. The problem is extra trades taken under relaxed standards. A trade with zero gross expectancy becomes −0.08R net. Each additional trade contributes that negative amount.
Extra trades can make a positive-expectancy system lose money
Consider a week with ten validated entries, each with +0.12R net expectancy. Expected profit is +1.2R. Now add 20 revenge trades after losses. Even generously assuming zero gross expectancy for each, round-trip cost of 0.08R makes each −0.08R net, totaling −1.6R. The weekly expected result becomes +1.2R − 1.6R = −0.4R.
The validated entry rule still has positive expectancy. The ten compliant trades still contribute +1.2R. Costs on the 20 additions exceed that, turning the week negative. Assuming their gross expectancy was zero is generous; trades taken with weaker criteria may perform worse.
This arithmetic is the core of revenge trading. The negative account result comes from the trades added after losses. Trade count can be governed by rules, so the cause can be addressed.
Measure revenge trading through re-entries after losses
Blaming emotion alone obscures where to intervene. Isolate trades placed shortly after losses and calculate three numbers.
First is the post-loss re-entry rate. For each trade closed at a loss, mark whether a new trade opened within a chosen window, such as 30 minutes or five bars. Calculate average R for just those re-entries and compare it with the average for all trades. A clearly lower or negative value identifies costly revenge trading.
Second is weekly R versus trade count. Record the number of trades and net R for each week. If high-frequency weeks repeatedly have lower R, the relationship points to overtrading.
Third is frequency after consecutive losses. Compare trades per hour after two or three losses in a row with the ordinary rate. An increase with streak length shows recovery pressure becoming extra trades.
This surge is common in sharp declines. Bitcoin fell about 39% from near $97,924 on January 14, 2026 to $60,000 on February 6, over a little more than three weeks. A fast decline can bring losses close together; each can increase the pressure and shorten the wait before re-entry. The market decline is public, but only your own records show whether your trade count multiplied during it.
Waiting periods and limits make trade count a rule
Separate trade frequency from emotion by deciding it in advance. Since post-loss pressure increases entries, place limits on behavior after a loss and on the whole day's trading.
- Wait after a loss: After closing at a loss, stop new entries for a set period. Enter again only when a new valid signal appears.
- Set a daily trade maximum: Fix the most entries allowed each day and stop when the limit is reached.
- Set a daily loss maximum: End trading for the day when cumulative losses hit a predefined limit, such as −2R.
- Keep entry criteria fixed: Do not lower conditions after losing. If they are not met, do not trade.
- Record the sequence: Note the prior result and time since the previous exit on every trade; review re-entry rate and weekly R against frequency.
These limits aim to prevent the extra trades taken outside validated standards after a loss, leaving only entries that satisfy the tested rule.
Two pitfalls
Increasing size to recover in one trade. This is revenge trading expressed as size instead of frequency. Larger entries during a losing run make each loss a greater share of the account and quickly raise risk of ruin. Trying to win it all back resembles Martingale, concentrating failure in one final streak.
Assuming more trades mean more data. A sample is meaningful only when trades follow the same rule. Trades added with looser criteria come from a different distribution. Mixing them obscures the original system's expectancy rather than enlarging its sample. To gather a bigger valid sample, collect more trades under the same criteria over time.
Trade frequency can be managed only when recorded
Revenge trades can look individually reasonable. A quick re-entry after a loss feels like a chance to recover, and any one trade may seem close to the rules. The problem becomes visible when several accumulate after losses. Only the full record reveals the pattern. Record the preceding outcome and re-entry interval for every trade; calculate post-loss re-entry rate, weekly R against trade count and round-trip cost in R. Then “I trade more after losses” becomes a manageable measure. Set a limit on post-loss trades and verify it in the journal to break the cycle from recovery pressure to overtrading.