Max Drawdown Controller

20% is a common professional standard - it's recoverable (needs a 25% gain), psychologically manageable, and matches moderate market corrections. However, your limit should match YOUR situation: if you'd panic at 15%, set it at 12%; if you can tolerate 30% with a proven recovery strategy, that may be appropriate. Ask yourself: at what loss would I make irrational decisions? Set your limit below that point.

Not necessarily all at once. A graduated response is usually better - reducing exposure progressively as drawdown deepens (e.g., at 10%, 15%, 18%) rather than one big action at 20%. If you do hit your absolute limit, closing all positions and taking a cooling-off period to review is prudent. A Singapore note: because there is no capital gains tax, these protective sales carry no tax cost. Avoid the temptation to immediately jump back in trying to recover.

A minimum 5-10 day cooling period is recommended. This prevents emotional revenge trading. During this time: review what went wrong, check if market conditions have changed, ensure your strategy still makes sense. After the cooling period, resume gradually - 25% position sizing for 2 weeks, then 50%, then 75%, then full. Don't rush back.

No, it doesn't guarantee. Gaps (overnight, or during a security's circuit-breaker cooling-off) can cause losses beyond your limit before you can act. Highly correlated positions can all drop simultaneously. Illiquid positions may be hard to exit. Drawdown control significantly reduces the probability and severity of deep losses, but it's not a guarantee. Build in buffers and have contingency plans.

Not necessarily. Different strategies have different risk profiles. A volatile momentum strategy might need a 15% limit while a conservative income strategy might have an 8% limit. However, you should also have an overall portfolio limit. Strategy limits are typically tighter than the portfolio limit - this way you catch problems at the strategy level before they become portfolio problems.

With leverage, drawdowns are amplified. A 2x leveraged position experiences 2x the drawdown. Set limits based on actual portfolio impact, not theoretical exposure. Example: with S$100,000 capital and S$200,000 exposure (2x leverage), a 10% market move is a 20% portfolio drawdown. Either use tighter DD limits or calculate DD based on leveraged exposure. Also remember margin calls may force liquidation before your chosen limit.

It depends on your trading style: intraday traders real-time or every few minutes; swing traders hourly during market hours, at minimum daily; long-term investors daily or weekly is sufficient. During high-volatility periods, increase frequency. Automated systems should check every price update. The key is catching threshold breaches before they get much worse.

Peak DD measures decline from the highest value achieved - this is the standard definition. Cost-basis DD measures decline from the original investment. Example: you invest S$100,000, it grows to S$150,000 (HWM), now at S$120,000. Peak DD (150-120)/150 = 20%. Cost DD (100-120)/100 = +20% (still up!). Peak DD is more useful for ongoing risk management; cost basis is useful for overall profit/loss assessment.

Calculate net portfolio value including both long and short P&L. A long/short portfolio may have natural hedging, so observe actual portfolio volatility, not gross exposure. Drawdown limits may be tighter (less DD expected due to hedging) or watch gross and net separately. Long-side DD and short-side DD can be tracked independently to identify which side is causing problems.

You can use a 'profit cushion' approach: if you're up significantly from cost basis, you can afford a larger drawdown from current levels without losing your original capital. Example: up 30% from cost, you could allow a 25% DD from current (still up 5% from cost). This is the 'house money' effect. However, don't let unrealized gains make you reckless - they're still real money.

Use Monte Carlo simulation to analyze the tradeoff: run simulations with various DD limits (10%, 15%, 20%, 25%, etc.). For each, measure final-wealth distribution, probability of limit breach, actual max DD, and risk-adjusted returns. Find the limit that maximizes a utility function (e.g., return - λ×variance of outcomes). Also consider your actual risk tolerance and whether you can psychologically handle the DD levels.

The multiplier choice depends on asset volatility and gap-risk tolerance. A higher multiplier = more upside participation but higher floor-breach risk. Common rule: Multiplier = 1 / Max Expected Gap. If the max expected overnight gap is 20%, use a multiplier of 5 max. Conservative 2-3; Moderate 4-5; Aggressive 6+. Backtest with historical data including crisis periods (2008, 2020) to validate.

Use market indicators to identify regimes: VIX levels (the CBOE VIX as a global proxy for low/medium/high volatility), trend indicators (bull/bear/sideways), breadth indicators. Map regimes to DD parameters: in a crisis regime use tighter limits and lower CPPI multipliers; in a calm bull regime you can use wider limits. The challenge is that real-time regime detection is imperfect - consider probability-weighted combinations of regime-specific parameters.

Use historical data including crisis periods. For each day: calculate DD, check thresholds, simulate response actions (position changes), apply transaction costs (in Singapore, brokerage + SGX fees + GST, with no capital gains tax on the de-risking sales). Track actual max DD with control vs without, final wealth, frequency of threshold hits, time spent at each level. Walk-forward test: optimize parameters on one period, test on the next. Be wary of overfitting - simpler systems are often more robust.

In crises, correlations spike toward 1 - diversification fails exactly when needed most (and in Singapore, banks and S-REITs already tend to move together on the rate cycle). Account for this by: (1) stress-testing with crisis correlations, not normal; (2) using conditional correlation in risk models; (3) considering tail-dependence (copula models); (4) having crisis-specific DD limits (tighter because diversification won't help); (5) ensuring hedges are negatively correlated in stress, not just in normal times.