MAE and MFE Analysis for Crypto Bot Optimization

Maximum Adverse Excursion (MAE) and Maximum Favorable Excursion (MFE) are trade-level diagnostic metrics introduced by John Sweeney in his 1997 book "Campaign Trading." MAE measures the maximum loss a trade experienced at any point during its life before eventually exiting (whether at a profit or a loss). MFE measures the maximum profit a trade reached at any point during its life before eventually exiting. These two data points, collected across a large sample of trades, allow you to make evidence-based decisions about stop-loss placement and take-profit targeting — decisions that most traders make based on intuition rather than their own trade data. For crypto bot traders with hundreds of trades in their history, MAE/MFE analysis can materially improve strategy performance without changing any entry signals.

Related guides: stop-loss strategies, profit factor, trade expectancy, parameter optimization, trading journal.

Maximum Adverse Excursion (MAE)

MAE is the deepest drawdown a trade experienced between entry and exit, regardless of the final outcome. For example:

Trade: BTC Long, Entry $68,000
  During trade: Price dropped to $67,000 (-1.47% from entry)
  Final exit: Take-profit at $69,500 (+2.21%)
  MAE = -1.47% (the worst point reached before the trade recovered and hit take-profit)

Collecting MAE across 100+ trades reveals: how deep do your winning trades typically dip before recovering? If 90% of your winning trades never go below -1.5% MAE before recovering, your current stop-loss at -2.0% may be already well-placed. But if most winning trades dip to -0.8% maximum, a stop-loss at -2.0% is giving the trade twice as much room as it actually needs — and a tighter stop at -1.2% would have caught the same winners while stopping out the losers much earlier.

Maximum Favorable Excursion (MFE)

MFE is the maximum profit a trade reached at any point before it eventually exited. For example:

Trade: BTC Long, Entry $68,000
  During trade: Price reached $70,200 (+3.24% from entry)
  Final exit: Stop-loss hit at $67,500 (-0.74%)
  MFE = +3.24% (the best point reached before the trade reversed and stopped out)

This particular trade was a loser that had +3.24% profit on the table at its peak. Collecting MFE across losing trades reveals: did you have substantial profit available that a trailing stop would have captured? If many losing trades show MFE above +2.0% before reversing, a trailing stop at +1.5% profit-lock would have converted many of them from losses to small wins. See our Parabolic SAR guide for SAR as a trailing stop mechanism.

MAE Scatter Plot Analysis

The classic MAE/MFE visualization: plot all trades with MAE on the x-axis and final trade outcome (win/loss amount) on the y-axis. Two clear patterns should emerge in a good strategy:

• Winning trades cluster near MAE=0: Most winners never went significantly negative — they moved in the target direction almost from entry
• Losing trades cluster at large MAE values: Losing trades consistently experienced large adverse excursions before stopping out — at or near the stop-loss level

If you see winning trades that experienced significant MAE (deep dips before recovering to profit), your stop-loss may be too tight — you are stopping out some potential winners that would have recovered. If losing trades cluster at small MAE values close to your stop, your stop is correctly calibrated. If there is significant overlap between winning and losing trade MAE values, the strategy's edge may be weak — the trades cannot be reliably differentiated by their early behavior.

Using MAE/MFE to Optimize Stop-Loss Placement

The evidence-based stop-loss optimization process using MAE:

1. Collect MAE data for 50+ completed trades (use DennTech trade history export)
2. Separate winning and losing trades
3. Find the MAE percentile for winning trades: what is the maximum MAE experienced by 95% of your winners?
4. Set stop-loss at that level plus a small buffer (e.g., 95th percentile MAE of winners = -1.3%, set stop at -1.6%)
5. This stop stops out almost no winners while catching losers at or near their typical reversal level

Compare this against the ATR-based stop methodology — MAE analysis provides empirical confirmation that the ATR multiple you chose is correctly calibrated for your specific strategy and pair combination.

Frequently Asked Questions

How many trades do I need for MAE/MFE analysis to be meaningful?

A minimum of 50 trades is needed to identify basic patterns; 100+ trades provides more statistically reliable conclusions. With fewer trades, individual outliers can distort the percentile calculations significantly. Use DennTech's trade history export to gather the full trade dataset, then add MAE and MFE columns to your trading journal spreadsheet. Calculate MAE by finding the minimum price relative to entry during each trade, and MFE by finding the maximum price relative to entry during each trade, using the candle high/low data from the exchange history. See the DennTech docs for trade history export format.

Can DennTech automatically apply MAE-optimized stop-losses?

DennTech's ATR-based stop-loss functionality provides a dynamic, volatility-calibrated stop that serves as the primary stop mechanism. After performing MAE analysis on your strategy's trade history, you can use the MAE results to calibrate the ATR multiplier — if your MAE analysis shows winning trades rarely exceed -1.5% MAE and ATR(14) is 1.2%, your ATR multiplier should be approximately 1.25× (1.25 × 1.2% = 1.5%). This bridges the gap between ATR methodology and empirical MAE calibration. See our ATR guide and optimization guide. Compare editions at the pricing page.

Should I use MAE analysis to change my take-profit targets as well as stop-losses?

Yes — MFE analysis directly informs take-profit calibration. If MFE data shows that your winning trades consistently reach +3.5% before pulling back, but your current take-profit is at +2.0%, you are leaving 1.5% of average winner profit on the table. Extend the take-profit to +3.0% (leaving a small buffer below the median MFE) to capture more of each winner without overshooting into reversals. Alternatively, implement a trailing stop that locks in profit at +2.0% and trails up to capture the remainder. The combination of MAE-calibrated stops and MFE-calibrated take-profits is the most evidence-based approach to stop and target optimization available. Get started at the pricing page and explore performance metrics at the live demo.

Optimization stack: backtesting, parameter tuning, MAE/MFE (this guide), trading journal. All strategies at the strategies page.

For a deeper dive into MFE specifically, read the Maximum Favorable Excursion deep-dive guide.