Skip to main content
Common Risk-Management Mistakes

Why Abandoning a Sound System During Drawdown Guarantees Losses

Pomegra Learn

Why Do Most Traders Abandon Profitable Systems at the Worst Possible Time?

Your system has a 58% win rate and a 1.3:1 risk-reward ratio. Backtests show a historical maximum drawdown of 18%. You begin trading it in January. By February, you're up 4%. By March, a regime shift hits. VIX spikes. Your mean-reversion system, designed for calm markets, breaks. By late March, you're down 16%. By April, you're down 18%—right at your historical maximum. The drawdown feels permanent, despite knowing mathematically that maximum drawdown is temporary. You make a decision: this system has failed. You stop trading it. You switch to a different approach. Then, over the next three weeks, markets normalize. Your old system would have recovered to +12% by May. But you exited at -18%, locking in losses forever. This is abandoning a sound system during drawdown, and it transforms an inevitable and temporary loss into a permanent one.

Quick definition: Abandoning a sound system during drawdown means stopping trading a strategy that has a documented edge (verified on backtests and out-of-sample data) because of a temporary loss, causing the trader to exit at the worst possible time and miss the recovery that statistics predicted.

Key takeaways​

  • Drawdown is the price of an edge, not the failure of an edge; every profitable strategy has maximum drawdowns
  • Traders typically abandon systems 60-80% of the way through the drawdown, missing the subsequent recovery that lasts 1-3x the drawdown duration
  • A system with 15% max historical drawdown will hit that drawdown roughly once every 3-5 years; exiting every time guarantees you'll miss 60%+ of annual returns
  • The psychological trigger for abandonment (pain of watching your account decline) happens exactly when the system is about to recover (the trough is where reversals begin)
  • Documenting your drawdown tolerance in writing before drawdowns occur is the only mechanism that prevents emotional abandonment

The Inevitability of Drawdown: It's Not Failure, It's Statistics​

Every profitable strategy with a measurable edge experiences drawdowns. This is not bad luck. It's not a sign the system is "broken." It's probability distribution. Here's why:

A system with a 58% win rate will produce clusters of losses. Not random loss distribution, but clusters. This is a proven phenomenon in statistics called "runs analysis."

A 58% win rate system might produce:

  • 5 consecutive wins, then 2 consecutive losses, then 6 wins, then 3 losses, etc.

The clustering is expected, not abnormal. A sequence of 3-5 losses in a row on a 58%-win-rate system happens 2-3 times per 100 trades. This is not broken. This is math.

Here's the drawdown arithmetic:

Assume your system risks $1,000 per trade (loss = $1,000 on a bad trade). Your average win is $1,300.

  • In normal sequences: +$1,300, -$1,000, +$1,300, -$1,000 (net: +$600 every 2 trades)
  • In loss clusters: -$1,000, -$1,000, -$1,000, -$1,000, +$1,300, +$1,300, +$1,300 (net: -$700 first 4 trades, then +$2,900 next 3 trades)

The loss cluster produces a deeper temporary drawdown (-$700) than the normal sequence, even though the expected value is identical. This is just how variance works.

If you trade 100 times per year, you'll experience 2-3 major drawdown clusters (3-5 losses in a row). Each will produce a -$3,000 to -$5,000 trough. This is the price of the edge. It's unavoidable. The choice is: do you stay through the trough and recover, or do you exit at the trough and convert it to permanent loss?

Drawdown vs. Return Cycles: Why Recovery Takes Longer Than People Expect​

Traders often underestimate how long recovery takes. They think: "If I lose 18%, I need 22% gain to recover" (because a 22% gain on $82 brings you back to $100). But they don't understand compound recovery timing.

If a system makes $1,500/month in normal times, but entered a drawdown phase (making -$1,500/month), here's the recovery math:

Month 1: Start at +$12,000 (annual running profit). Drawdown begins. End: +$10,500.
Month 2: Drawdown continues. End: +$9,000.
Month 3: Drawdown peak. End: +$7,500 (18% down from $12,000 peak).
Month 4: System recovers. Profit resumes. End: +$9,000 (+$1,500).
Month 5: End: +$10,500 (+$1,500).
Month 6: End: +$12,000 (+$1,500).
Month 7: End: +$13,500 (+$1,500).
Month 8: End: +$15,000 (+$1,500).

Recovery took 5 months (months 4-8) to get back to previous high. Drawdown took 3 months to develop. Recovery took 1.7x longer than drawdown. This is typical. A study of CTAs found average recovery time is 1.5-2.0x the drawdown duration.

Traders exiting at month 3 (peak pain, the trough) miss the recovery and lock in losses. By month 4, if they'd stayed, they'd be back to +$9,000. By month 8, they'd be at +$15,000 (new high). Instead, they're at -$12,000 (permanent loss) because they exited.

Historical Examples: Systems Abandoned at the Worst Time​

Example 1: Trend-Following CTAs in 2015

Trend-following CTAs (systematic traders who ride momentum in commodities, currencies, and equity indices) experienced a brutal June-August 2015 drawdown. The reason: oil collapsed, China devalued, U.S. rates stayed low. Trend-following systems, which make money during directional moves, were whipsawed by sudden reversals and low liquidity.

Many CTAs experienced -8% to -15% drawdowns. Their track records over the prior 2-3 years had shown max drawdowns of 12-14%, so -15% was slightly worse than expected but not unprecedented.

What happened next:

  • Many CTA clients panicked. "This system is broken. Commodity collapse isn't coming back. This is a new normal."
  • Clients redeemed (withdrew money) from CTAs right at the drawdown nadir.
  • The CTAs liquidated positions, locking in losses.
  • By October 2015, volatility normalized. Trend-following systems resumed normal operations and profits.
  • From October 2015 to September 2016, trend-following CTAs returned +8% to +12% as commodity recovery trends emerged.

The clients who abandoned at August 2015 trough locked in -12% to -15% losses. The clients who stayed earned +8% to +12% recovery gains. The difference: 20-27% gap—purely from timing the exit wrong.

Example 2: The Options Seller's Abandonment (February 2018)

A retail trader ran a weekly SPY short-put-spread system. The system had crushed it for 18 months: 89 consecutive winning months. Win rate: 92%. Average monthly profit: 0.8% of account. Max drawdown (backtested): 6%.

Then February 2018 "Volmageddon" hit. In a single day (February 5), VIX spiked from 16 to 40+. The trader's short puts, positioned for calm vol, got hammered. Instead of small losses (he'd been accepting 0.2% monthly losses on the 8% of losing months), his account dropped 4.2% in one day.

He panicked. "This system doesn't work anymore. Vol changed." He closed all positions, locking in the loss. What happened next:

  • February 6-8: Vol crashed back down to 22-25.
  • February 12+: Vol normalized to 18-20 (back to pre-spike levels).
  • By February 28: His positions, if left open, would have recovered 75% of the loss.
  • By March 31: Full recovery + new profit.

Instead, he'd exited at the worst moment. His 18-month winning streak became an 18-month track record with a nasty -4.2% crater right at the end.

Example 3: The Momentum Trader in 2022

A momentum-based long-only system (long the best-performing stocks, short the worst) made money in 2018-2021. Max drawdown: 12% (in March 2020, during COVID crash; recovered fully by June 2020).

2022 was brutal: Fed tightening, growth stocks sold off. The system's drawdown reached -18% by July 2022. The trader had backtested and knew max drawdown was 12%, so -18% was "broken." He abandoned the system in July, closing all positions.

What actually happened:

  • August 2022: Cooling inflation data emerged. Fed began pivoting.
  • September 2022: The bottom formed. The system would have started recovering.
  • October 2022 - December 2022: Massive rally in previously sold-off growth stocks. The system would have returned +22% in Q4.

The trader who abandoned in July missed a +22% recovery. His net 2022 result: -18% (locked in loss). The traders who stayed through: -18% (Jan-July) + 22% (Aug-Dec) = +3.2% for the year.

The Pain Peak Coincides With the Recovery Start​

Here's the cruel truth of drawdown psychology: the moment of peak pain (maximum drawdown trough) is statistically where the recovery begins. This is because:

  1. Mean reversion is real. Extremes tend to reverse. Maximum drawdown is an extreme. By definition, it's when the system is most over-extended in the "wrong direction."

  2. Loss clustering ends. The cluster of losses that created the drawdown is statistically unlikely to continue. The next cluster is likely to be wins, which produces recovery.

  3. Market regimes shift. A regime that breaks your system (e.g., low-vol regime breaks mean-reversion systems) will eventually shift. When it shifts, the system re-engages.

The trader's brain, however, experiences peak pain at the trough and concludes the system is broken. But the system is actually about to recover. The emotional response (abandon) is perfectly timed with statistical recovery (return). This is the cruelest misalignment in trading psychology.

The Pre-Drawdown Contract: Your Only Defense​

The only reliable defense against abandoning systems is writing down your drawdown tolerance before drawdowns occur. This is called a "trading plan" or "system rules." It should include:

System Name: [Name]

Expected Performance:
- Win rate: 58%
- Average win: $1,300
- Average loss: $1,000
- Expected annual return: 12%

Drawdown Parameters:
- Historical max drawdown: 18%
- Tolerance: I will NOT abandon the system until drawdown exceeds 22%
- Recovery expectation: Recovery typically takes 1.5-2x the drawdown duration
  (An 18% drawdown typically recovers within 3-4 months)

Abandonment Trigger:
- Only abandon if: (a) Backtested edge no longer exists (win rate drops to <50%), OR
                   (b) System parameters fundamentally change (e.g., new software breaks, settings corrupted)
- DO NOT abandon simply because drawdown is deep

Recovery Plan:
- If drawdown reaches 18%, I will email my accountability partner: [name/email]
- My partner will review my system rules and confirm I should continue
- I will NOT make any decisions during peak drawdown; decisions happen 2 weeks after trough

This contract, signed before drawdown, acts as a circuit-breaker on emotional decisions. When drawdown hits and your brain screams "exit," you instead re-read the contract, which says "recovery typically takes 3-4 months," and you white-knuckle through.

The traders who survive multiple market cycles are those who:

  1. Document system rules before deployment
  2. Have an external accountability partner review decisions during peak pain
  3. Separate "system broken" (mathematical failure) from "system in drawdown" (temporary variance)

Distinguishing System Failure From Normal Drawdown​

Not all drawdowns are temporary. Sometimes a system actually breaks. How do you tell the difference?

Signs the System Is Working Normally (Temporary Drawdown):

  • Win rate is still at or near the expected level (58% ± 3-5%)
  • The reason for drawdown is market regime shift, not system malfunction (e.g., low-vol system gets hit by vol spike—expected)
  • Backtested max drawdown was similar to current drawdown
  • Profit factor (gross wins / gross losses) is still positive
  • The drawdown is within the tolerance range you documented pre-deployment

Signs the System Actually Broke (Not Just Drawdown):

  • Win rate has dropped significantly (58% → 42%, sustained over 50+ trades)
  • Losses are larger than expected (expected avg loss $1,000, now $3,000)
  • Back-to-back winners have turned into consistent losers despite setup looking identical
  • Slippage or fill quality has degraded (your execution environment changed)
  • Market structure shifted and your system can't adapt (e.g., liquidity dried up, new regulations, competitor strategies)

If you see signs of actual breakage, investigate. Possibly quit. But if you see signs of normal drawdown, stay the course. The math says recovery is coming.

Compound Effect: Abandoning Systems vs. Persisting​

Start with a $100,000 account. System returns 12% annually with 18% max drawdown.

Trader A: Stays through drawdowns

  • Year 1: Start $100K. Hit 18% drawdown (trough: $82K). Recover. End year: $112K.
  • Year 2: Start $112K. No major drawdown. End: $125,440.
  • Year 3: Start $125,440. Hit 15% drawdown (trough: $106,624). Recover. End: $140,493.
  • Year 4: Start $140,493. No major drawdown. End: $157,352.
  • Year 5: Start $157,352. Hit 18% drawdown. Recover. End: $176,235.
  • 5-year total: $176,235 (76% gain)

Trader B: Abandons at peak drawdown, loses year of returns

  • Year 1: Hit 18% drawdown. Panic. Abandon. Lock in -18%. End year: $82,000.
  • Year 2: Switch to "safer" strategy (returns 8%). End: $88,560.
  • Year 3: Switch back to original system. Hit 15% drawdown. Abandon. Lock in -15%. End: $75,276.
  • Year 4: Switch to safe strategy. Return 8%. End: $81,298.
  • Year 5: Back to original system. No drawdown. +12%. End: $91,054.
  • 5-year total: $91,054 (−9% loss)

The difference: $176,235 vs. $91,054 = $85,181 difference purely from not abandoning during drawdowns. That's 94% more wealth from persistence.

The irony: the trader with the less frequent drawdowns (Trader B) does worse because they abandon systems at inopportune times. The trader with the same drawdowns who persists (Trader A) compounds better.

FAQ​

How do I know if 18% drawdown is "normal" vs. "system is broken"?​

Check the backtests. If historical max drawdown was 18%, then current -18% is expected. It's not proof of breakage; it's confirmation that you're on a known statistical path. If current drawdown is larger than historical max (e.g., -24% when backtest said -18%), then investigate. Likely causes: (a) smaller sample size (you haven't traded long enough to see true max), (b) regime shift (market changed, system needs tuning), or (c) execution degradation.

Should I ever abandon a system during drawdown?​

Only if the system broke—meaning win rate collapsed (58% → 35%), losses exploded, or market structure shifted irreversibly. If the drawdown is within historical bounds, no. If drawdown is 1.2x larger than historical max, start investigating but don't abandon yet. Only abandon after confirming that the mechanism of profit (the edge) has truly failed, not just hit a statistical rough patch.

How deep can drawdowns get? Can they be deeper than the backtest suggested?​

Yes. Backtests often underestimate tail drawdowns because:

  • Backtests assume perfect fills; real trading has slippage
  • Backtests ignore gaps; real markets gap overnight
  • Backtests assume you can exit at stop price; liquidity sometimes prevents this
  • Backtests use historical volatility; new regime spikes can be worse

A system with -18% historical max drawdown might experience -22% to -25% in real trading. Plan for this. Use position sizing that keeps you sane even if realized drawdown is 1.3-1.5x worse than backtest suggested.

What if I hit my drawdown tolerance? Is it then safe to quit?​

Not necessarily. "Drawdown tolerance" (e.g., 22%) is not the same as "abandon decision threshold." It's the point where you:

  1. Halt new entries (stop adding to the position)
  2. Review the system (is it broken or just in drawdown?)
  3. Consult your trading partner
  4. Consider scaling back position size by 30-50%

But you don't automatically abandon. You slow down and investigate, then continue at reduced size if the mechanism is still working.

Can I ever be too early in the system's history to know if drawdown is normal?​

Yes. If you've only traded 20 times and hit a -15% drawdown, and the system was backtested on 500+ trades, you have limited statistical power. The -15% might be the true max or just a coincidence from a small sample. This is why new system traders should:

  • Start with minimal position size (1/4 of intended final size)
  • Trade for at least 100 real trades before committing full size
  • Allow for 1-2 years of live trading before drawing conclusions

Is there a mathematical formula for "optimal" drawdown tolerance?​

Not exactly, but the Calmar Ratio (annual return / max drawdown) guides it. A system with 12% return and 18% drawdown has Calmar = 0.67. This is acceptable. If Calmar falls below 0.4, the drawdown is too large relative to return—consider quitting or reducing size. If Calmar is 1.0+, the return-to-drawdown ratio is excellent, and you should stay through drawdowns.

How do I handle psychological pain during deep drawdown?​

  1. Review the pre-written contract you drafted before drawdown began
  2. Run a backtest on the current year to confirm win rate is still near expected
  3. Email your accountability partner and say "I'm in drawdown; partner, confirm I should stay"
  4. Reduce position size by 30-50% to lower the emotional pain (smaller daily swings hurt less)
  5. Take a 1-week break from the system (but don't abandon it) if you're completely emotionally depleted
  6. Track recovery timeline: Most recoveries take 1.5-2x the drawdown duration. Document that your recovery matches prediction.

Summary​

Abandoning a sound system during drawdown is the most expensive mistake a trader can make. Drawdown is not failure; it's the price of an edge. Every profitable system experiences drawdowns. A system with 58% win rate will experience 3-5 loss clusters per 100 trades, producing temporary declines of 10-20%. These are not signs the system is broken; they're signs the system is working in statistically expected ways.

The cruel irony of drawdown psychology is that peak pain (maximum drawdown trough) coincides with statistical recovery beginning. Traders who abandon at peak pain lock in losses and miss the recovery. Traders who persist through peak pain earn the recovery gains and compound wealth.

The only effective defense is a pre-written contract, signed before drawdowns occur, that documents:

  • Expected maximum drawdown (from backtests)
  • Drawdown tolerance (a threshold above max, where you'll investigate but not abandon)
  • Recovery timeline (typically 1.5-2x the drawdown duration)
  • The definition of "system broken" vs. "system in drawdown"

When drawdown hits (and it will), you'll be tempted to abandon. Your contract will remind you that you've already lived through this statistically and decided to persist. You'll re-read it, white-knuckle through the trough, and emerge richer on the other side.

The traders who build wealth over decades are not those with the smoothest equity curves. They're those with the discipline to endure drawdowns and the documented rules to prevent panic exits at the worst possible moments.

Next​

→ Not Knowing Your Max Drawdown Tolerance