Managing a Drawdown
Concept
A drawdown is the peak-to-trough decline in your account equity curve: how far you have fallen from a prior high water mark. Maximum drawdown is the largest such drop over a period you care about—often measured from all-time peak for a strategy audit, or from year start for personal planning. Drawdowns are normal in any strategy with positive expectancy but nonzero loss probability; what separates professionals is how deep drawdowns get and how long recovery takes when rules are followed versus when behavior deteriorates.
Recovery math is non-linear and often underestimated. A 10% loss needs an 11.1% gain on remaining capital to recover. A 20% loss needs 25%. A 50% loss needs 100%—you must double what is left. A 75% loss requires a 300% gain on the remainder. The asymmetry means avoiding large drawdowns is far more efficient than clawing back from them. Small, repeated process losses hurt; single catastrophic losses can end participation entirely.
When to pause is a risk-management decision, not a sign of failure. Common pause triggers include: drawdown exceeding a predefined monthly or quarterly cap; three consecutive process violations (not merely losses); personal stressors that impair judgment; or market regime shifts that invalidate your edge (e.g., your mean-reversion system in a one-direction liquidity event). Pausing means no new risk, not necessarily closing the app forever—you replan, resize, or paper-trade until adherence returns.
Calculating drawdown from XT P&L requires a clear equity series. Start with daily or trade-by-trade equity: beginning balance plus cumulative realized PnL (and unrealized if you mark open positions—be consistent). Track the running maximum of equity; drawdown at any point is (peak − current) / peak, often expressed as a percentage. Deposits and withdrawals distort simple PnL if ignored—either remove those cash flows from the series for “trading-only” performance or adjust the equity path when capital is added or removed so you do not mistake a deposit for skill.
Risk of ruin rises as drawdown deepens if you increase size to recover faster—martingale behavior. The professional response to drawdown is often decrease size (risk half your normal per-trade risk until a new high water mark) or freeze until review. Leverage cuts both ways on the way down; reducing leverage lengthens recovery in % terms but prevents terminal outcomes.
Finally, time in drawdown matters psychologically. A 10% drawdown that lasts two days feels different from 10% over six months of chop. Both count in the stats. Your journal and XT history help you see whether drawdowns come from clustered strategy losses or one-off fat tails—each suggests different mitigations (correlation vs. tail hedging vs. simply smaller size).
Underwater recovery paths differ by style: mean-reversion books may snap back when volatility mean-reverts; trend books may need a new leg to print. If your drawdown profile shows long flat periods after loss, ask whether you are trading during unfavorable regimes out of boredom. Pausing is active risk control, not passivity.
Benchmark honestly: beating BTC in a bear phase is a different claim than absolute dollars up. Your XT P&L sheet should separate alpha attempts from beta bags if you run both—otherwise you confuse a market drawdown with a process drawdown, or vice versa.
Leverage and margin add non-linear drawdown risk: a 20% spot move against you with 5× leverage can approach account trauma even when the underlying move looks ordinary on a chart. If drawdowns accelerate near liquidation, the fix is often lower leverage and wider equity buffers, not cleverer entries alone.
Documentation of drawdown episodes—what you traded, news context, whether rules were followed—creates institutional memory for a one-person desk. Next time equity peaks, you will have a playbook for the next valley instead of rediscovering panic responses from scratch.
Observe on XT
Find PnL or profit reports for futures or spot trading (if separated). Note whether the UI shows daily, period, or per-trade realized figures.
Locate your account equity history if available (some platforms chart wallet value over time). If not available, plan to build equity from exports as in the journal lesson.
Identify deposits and withdrawals in transaction history so you can exclude them when computing trading drawdown versus net worth drawdown.
Practice
- Pick a starting date (e.g., first of last month) and record starting equity in your primary XT trading wallet(s).
- Export or sum realized PnL from XT for each day or each closed trade from that date to today.
- Build a simple equity curve: start + cumulative PnL each day (ignore deposits/withdrawals for this exercise, or note them if any occurred).
- Compute running peak and current drawdown % from peak: (peak − now) / peak × 100%.
- If current drawdown is D%, compute required gain on remaining capital to recover: 1 / (1 − D/100) − 1, expressed as a percentage (e.g., D=50% → need 100% on remainder).
- Write a pause rule: e.g., “If drawdown from month peak exceeds X%, halve position size and no new strategies for one week.”
Checkpoint
Q1: If your account falls 50% from its peak, what return on the remaining balance is required to return to the peak?
- A) 50%
- B) 75%
- C) 100%
- D) 25%
Correct: C. You must double the remaining capital (100% gain) to recover from a 50% drawdown.
Q2: Maximum drawdown measures:
- A) Only your largest winning trade
- B) The largest peak-to-trough decline in equity over the measured period
- C) Only fees paid in a month
- D) Volatility of a single candle
Correct: B. Max DD captures the worst equity path drop in the sample.
Q3: Why is increasing position size during a drawdown especially dangerous?
- A) It always guarantees recovery in one trade
- B) It raises risk of ruin and often reflects emotional “make it back” behavior rather than edge
- C) Exchanges forbid any trading after losses
- D) Drawdowns never happen to disciplined traders
Correct: B. Martingale-style escalation compounds tail risk; recovery should prioritize survival and process.