$npx -y skills add SKE-Labs/agent-trading-skills --skill drawdown-managementManage account drawdowns with limits and recovery protocols. Use when protecting capital during losing streaks, implementing loss limits, or developing recovery plans.
| 1 | # Drawdown Management |
| 2 | |
| 3 | Drawdown management protects capital and ensures trading longevity through loss limits and recovery protocols. |
| 4 | |
| 5 | ## Drawdown Types |
| 6 | |
| 7 | | Type | Definition | |
| 8 | | ---------------- | ------------------------------ | |
| 9 | | Max Drawdown | Largest peak-to-trough decline | |
| 10 | | Daily Drawdown | Loss in single trading day | |
| 11 | | Weekly Drawdown | Loss in single week | |
| 12 | | Open Drawdown | Unrealized loss on open trades | |
| 13 | |
| 14 | ## Drawdown Policy |
| 15 | |
| 16 | Obtain daily, weekly, strategy, and maximum-drawdown limits from the portfolio mandate. Define the equity series (realized plus marked-to-market P&L, fees, funding, deposits/withdrawals), peak reset rule, timezone, and whether open-position stress is added. If no limits exist, propose examples for approval rather than enforcing invented percentages. |
| 17 | |
| 18 | ## Loss Response Protocol |
| 19 | |
| 20 | At a limit, block new risk as specified by the mandate and alert the owner. Do not automatically liquidate positions: compare orderly reduction with gap, spread, tax, and hedge consequences. Resume only after the documented cause, data/execution health, model validity, and revised risk budget pass an authorized review—not after an arbitrary pause or number of winning days. |
| 21 | |
| 22 | ## Recovery Math |
| 23 | |
| 24 | Larger drawdowns require exponentially larger gains to recover: |
| 25 | |
| 26 | | Drawdown | Return Needed to Recover | |
| 27 | | -------- | ------------------------ | |
| 28 | | 10% | 11% | |
| 29 | | 20% | 25% | |
| 30 | | 30% | 43% | |
| 31 | | 50% | 100% | |
| 32 | |
| 33 | Formula: `Recovery % = Drawdown / (1 - Drawdown) x 100` |
| 34 | |
| 35 | This is why preventing large drawdowns matters more than maximizing gains. |
| 36 | |
| 37 | ## Prevention Strategies |
| 38 | |
| 39 | 1. **Size to mandate** -- include gap and correlated exposure |
| 40 | 2. **Enforce approved limits** -- block new risk and escalate exactly as documented |
| 41 | 3. **Manage correlations** -- avoid stacking similar positions (see correlation-risk) |
| 42 | 4. **Reduce in losing streaks** -- cut size early, don't wait for the limit |
| 43 | 5. **Diversify strategies** -- single-strategy risk is concentrated risk |
| 44 | |
| 45 | ## Monitoring |
| 46 | |
| 47 | Track these metrics continuously: |
| 48 | |
| 49 | - Current drawdown from equity peak |
| 50 | - Number of consecutive losing days |
| 51 | - Deviation from the strategy's confidence interval for return and loss distribution |
| 52 | - Equity curve slope (flattening = warning) |
| 53 | |
| 54 | ## Evidence and Validation |
| 55 | |
| 56 | - Treat the setup as a testable hypothesis, not a prediction. Define thresholds, entry, invalidation, and exit before evaluating outcomes. |
| 57 | - Calibrate on the same instrument, venue, session, and timeframe. Use closed candles and a held-out or walk-forward sample; record every variant tried. |
| 58 | - Include spread, fees, slippage, borrow or funding, partial fills, and latency. Reject the setup when net expectancy is not positive or depends on one narrow parameter. |
| 59 | - Return observed inputs, missing data, cost assumptions, entry, invalidation, exit, and a valid, watch, or no-trade status. |
| 60 | - Research basis: [Risk-constrained Kelly research](https://web.stanford.edu/~boyd/papers/kelly.html) explicitly trades expected growth against drawdown probability; loss limits should be chosen from mandate and tested distributions, not copied as universal percentages. |
| 61 | |
| 62 | ## Key Rules |
| 63 | |
| 64 | - Never add risk to recover a loss unless a preapproved sizing policy independently calls for it. |
| 65 | - Preserve realized/unrealized P&L, cash flows, peak, limit, and response in an immutable audit record. |
| 66 | - Distinguish strategy deterioration, expected variance, concentration, data faults, and execution faults. |
| 67 | - Do not label a drawdown percentage normal without strategy-specific simulations and mandate approval. |
| 68 | - Use forward-looking stress and remaining risk budget, not winning/losing streak folklore, to set size. |
| 69 | |
| 70 | ## Related Skills |
| 71 | |
| 72 | - **position-sizing** -- proper sizing is the primary drawdown prevention tool |
| 73 | - **correlation-risk** -- correlated positions amplify drawdowns |