Trading bot portfolio management (running multiple strategies well)

One bot on one market is a single bet with a single failure mode — when its regime disappears, your whole account suffers. The professional move is a portfolio: several bots running different strategies across different markets, so that when one is bleeding, another is earning. Done right, diversification is the closest thing to a free lunch in trading — it can smooth the equity curve without sacrificing return. This guide covers correlation, capital allocation, rebalancing, and the code to run a fleet.

On this page
  1. Why a portfolio
  2. Correlation is the key
  3. Allocating capital
  4. Rebalancing
  5. The code
  6. Portfolio-level risk
  7. FAQ

Why run a portfolio of bots

Every strategy has a regime where it wins and a regime where it bleeds — trend following loves trends and hates chop; mean reversion is the opposite. Run them together and the chop that hurts your trend bot is exactly when your reversion bot earns. A diversified set of bots that don't fail at the same time produces a smoother equity curve than any single one — and a smoother curve means smaller drawdowns and the ability to size up safely.

Correlation is the whole game

portfolio (smoother) strategy Astrategy B
Two choppy, uncorrelated curves combine into a steadier portfolio — the diversification benefit.

Diversification only helps if the bots are genuinely uncorrelated — if their returns don't move together. Five bots that are all long-only crypto momentum are one bet wearing five costumes; they'll all crash together. Real diversification mixes strategy types (trend, reversion, breakout), markets (crypto, equities, FX), and timeframes. Measure pairwise correlation and treat anything above ~0.7 as effectively the same bet.

Allocating capital

Rebalancing

Winners grow and losers shrink, so allocations drift. Periodic rebalancing — trimming the bot that's grown oversized and topping up the laggard — keeps the portfolio at its intended risk and quietly enforces “sell high, buy low” across your strategies. Rebalance on a schedule (monthly/quarterly) or when an allocation drifts past a band, not constantly (fees add up).

The code

python · allocate.py# risk-parity weights: inverse of each bot's volatility
def risk_parity(vols):
    inv = [1/v for v in vols]
    total = sum(inv)
    return [w/total for w in inv]   # weights sum to 1

weights = risk_parity([0.30, 0.12, 0.20])   # vol of each bot
capital = [w*account for w in weights]    # allocate per bot

Portfolio-level risk

Watch the total, not just each bot

Individually safe bots can be collectively dangerous if they all pile into the same trade — five “different” bots all long Bitcoin is one giant Bitcoin bet. Enforce a portfolio-wide max exposure and a portfolio-wide drawdown kill-switch, on top of each bot's own limits. Size every bot's positions with the position calculator and govern the whole fleet with the discipline in risk management.

Before committing capital, backtest each strategy in the backtester and study their correlation — a portfolio of overfit bots is still overfit. Stress the combined result with a Monte Carlo simulation.

Not financial advice. This content is educational. Automated and algorithmic trading carries a real risk of financial loss. Never trade money you cannot afford to lose. Review the SEC investor.gov and CFTC resources before trading.

Frequently asked questions

What is trading bot portfolio management?

It is running multiple bots across different strategies and markets as a single portfolio, so that when one strategy's regime hurts it, another earns. Managing correlation, capital allocation and rebalancing across the fleet smooths the combined equity curve and reduces drawdowns.

Why run multiple trading bots instead of one?

One bot on one market is a single bet with a single failure mode — when its regime disappears, the whole account suffers. A diversified set of uncorrelated bots that don't all fail at once produces a smoother equity curve, smaller drawdowns and safer position sizing.

How important is correlation between bots?

It is the whole point. Diversification only helps if the bots are genuinely uncorrelated; several bots all doing long-only crypto momentum will crash together. Mix strategy types, markets and timeframes, and treat any pair with correlation above about 0.7 as effectively the same bet.

How do you allocate capital across trading bots?

Common methods are equal weight (split evenly, hard to beat), risk parity (each bot contributes equal risk, so volatile bots get less capital), and cautious performance-weighting. Whatever the method, rebalance periodically and enforce a portfolio-wide exposure and drawdown limit.

MB

Mustafa Bilgic

Algorithmic trading practitioner · Founder, AITradingBot.us

Mustafa builds and backtests automated trading systems and writes about them without the hype. Every tool on this site is free and runs entirely in your browser.