Multi-Strategy Manager

Start with 2-3 strategies. More isn't always better - you need genuinely different, uncorrelated strategies. 3-5 well-chosen strategies typically capture most diversification benefit. Too many creates complexity without additional benefit.

Choose strategies that: (1) Each have positive expectancy independently. (2) Work in different market conditions (e.g., trend following + mean reversion). (3) Have low correlation. (4) You can actually execute. Start with opposite-style strategies.

Equal allocation is a good starting point. It's simple, requires no estimation, and provides diversification. As you gain experience, consider risk parity (equal risk contribution) or performance-based weighting.

This is exactly what you want! It means strategies are uncorrelated. The winning strategies offset the loser. Don't abandon the losing strategy immediately - this diversification is the point of multi-strategy.

Daily: Quick P&L check. Weekly: Strategy metrics. Monthly: Full analysis, correlation check, consider rebalancing. Quarterly: Deep review, strategy evaluation, potential additions/removals.

Calculate each strategy's volatility (e.g., annualized standard deviation of returns). Weight inversely: Weight_i = (1/σ_i) / Sum(1/σ_j). Recalculate periodically as volatilities change. Higher vol strategies get less capital.

Correlation > 0.5 is concerning. Above 0.7 is high - strategies aren't providing much diversification. Ideally keep correlations < 0.3. If correlations are high, consider replacing one strategy or reducing combined exposure.

Estimate capacity by analyzing liquidity (trade < 1-5% of daily volume). Set maximum allocation based on capacity. Even if a strategy performs best, don't exceed capacity. For small accounts, capacity is rarely an issue.

Calendar method: Monthly or quarterly. Threshold method: When any strategy drifts >5% from target. Hybrid: Check monthly, only rebalance if threshold breached. Consider tax implications in non-registered accounts.

Contribution = Strategy Return × Strategy Weight. Sum contributions = Portfolio Return. Track monthly. Identify consistent contributors vs detractors. Use attribution to guide allocation decisions.

1) Define regimes (e.g., trend+low vol, trend+high vol, range+low vol, range+high vol). 2) Create regime indicators (ADX for trend, VIX for vol). 3) Define target weights per regime. 4) Detect current regime and apply weights. 5) Smooth transitions.

MVO is highly sensitive to input estimates. Small changes in expected returns dramatically change optimal weights. Estimation errors make 'optimal' weights suboptimal in practice. Use robust methods: shrinkage, resampling, constraints, Black-Litterman.

Monitor rolling correlations. Alert when correlations exceed thresholds. During spikes: reduce overall exposure, increase hedges, or activate defensive strategies. Stress test portfolio under high-correlation scenarios. Plan before it happens.

Retire when: sustained negative performance (12+ months underperformance vs expectations), correlation with other strategies increased materially, or clear evidence the edge is gone (market structure changed, crowded trade). Document lessons.

Define max portfolio drawdown (e.g., 20%), max strategy drawdown (e.g., 15%), max gross exposure (e.g., 200%), max net exposure (e.g., 100%), daily VaR limit. Implement automated monitoring. Kill switch when breached. Review limits periodically.