risk management
In Systematic Trading, Robert Carver argues that risk management is more important than signal generation. This insight stems from the recognition that trading is an exercise in surviving uncertainty. Even excellent signals with high predictive power will fail to deliver returns—or worse, lead to ruin—if the trader does not control risk. Conversely, even mediocre signals can yield long-term profits if supported by sound risk controls.
The essence of risk management lies in ensuring you stay in the game long enough for your statistical edge to manifest. Without risk controls, drawdowns can rapidly deplete capital, triggering forced liquidation or psychological capitulation.
First-Principles Definition of Risk Management
Risk management is not just a set of rules. It can be broken down using first-principles thinking — deconstructing the idea to its most basic, irreducible components.
- Uncertainty: All financial markets are probabilistic, not deterministic. Outcomes cannot be predicted with certainty, and variance is inherent.
- Asymmetry: Losses are nonlinear in impact. A 50% portfolio loss requires a 100% gain to break even, making large drawdowns exponentially more damaging.
- Capital is Finite: Trading capital is a limited resource. Once depleted, the ability to recover is gone.
- Return Is Conditional: Any long-term return depends on survival. Even positive expected value strategies cannot yield results if the account is blown up.
- Variance Kills: Excessive volatility, even with a winning system, can cause drawdowns that make it impossible to continue trading.
From this we derive the core purpose of risk management: to control exposure to adverse outcomes so that your strategy can perform over time.
Core Components of Risk Management
Robert Carver identifies several essential elements that form the foundation of robust risk management. These are not optional features; they are the mechanisms that translate signal into sustainable execution.
1. Position Sizing
Position sizing determines how much capital is allocated to a single trade. This decision should not be based on intuition or conviction but on an objective assessment of risk. Carver advocates for techniques such as volatility scaling or risk-budgeting, where the size of a trade is proportional to the estimated volatility of the asset.
First-principle rationale: Without position control, even a single bad trade can destroy an account. Position sizing acts as a throttle that limits exposure per unit of risk.
2. Diversification
Diversification spreads risk across uncorrelated or weakly correlated instruments. Proper diversification considers not just asset class but also time horizon and strategy type. True diversification reduces the likelihood that multiple trades will fail simultaneously due to the same market factor.
First-principle rationale: If exposures are highly correlated, a systemic market shock can cause correlated losses. Diversification reduces portfolio-level variance and smooths returns.
3. Volatility Targeting
Volatility targeting adjusts trade size based on the current market volatility of the instrument. Higher volatility implies greater potential loss, so exposure is scaled down. Lower volatility permits increased exposure without increasing risk.
Vol targeting aims to maintain a constant level of risk across instruments and time periods. This avoids unintended consequences such as outsized positions during high-volatility regimes.
First-principle rationale: Market risk is dynamic. Static position sizes fail to account for regime shifts. Vol targeting ensures consistent risk per trade.
4. Drawdown Controls
Drawdown controls are mechanisms for reducing exposure or halting trading when performance deteriorates. These can be fixed thresholds (e.g. a 20% portfolio loss triggers pause) or adaptive systems that adjust exposure based on recent underperformance.
Drawdown controls protect both capital and trader psychology. Prolonged or deep drawdowns increase the likelihood of abandoning a strategy at the worst possible time.
First-principle rationale: Even robust systems can underperform. Drawdown limits prevent small setbacks from escalating into irrecoverable losses.