Portfolio thinking: several strategies, several accounts
- Why combining uncorrelated strategies smooths results
- How to allocate risk across strategies
- When adding a strategy makes things worse
One strategy has bad seasons
A trend strategy suffers in ranges; a range strategy suffers in trends. Run both and, when one is in drawdown, the other may be earning. That's the core of portfolio thinking: combine edges that don't fail at the same time.
What to combine
- Different styles: trend and mean reversion.
- Different markets: FX, gold, indices.
- Different timeframes: swing and intraday.
Check correlation of the equity curves, not just of the markets. Two strategies trading different pairs can still lose together if both are bets on the dollar.
Allocating risk
Split your total risk budget. Example: a total of 2% daily risk, split 1% to strategy A and 1% to strategy B. Strategies with deeper historical drawdowns get a smaller share.
When adding makes it worse
- The new strategy is unproven: it adds noise, not edge.
- It's highly correlated with what you already run: it adds risk, not diversification.
- You can't monitor it properly.
Several accounts
Many traders separate strategies into different accounts (or prop accounts) so results are clean and one strategy's drawdown can't trigger another's limits. A copier can then route each master's trades to the right accounts.
- Adding strategies that all fail in the same conditions.
- Giving every strategy the same risk regardless of its drawdown.
- Running more strategies than you can monitor.
How this works when you copy trades
Check yourself
- Combine strategies whose bad periods don't overlap.
- Allocate risk by drawdown, not equally by default.
- Unproven or correlated additions make a portfolio worse.