Risk Management and Trading Psychology

Risk Management

Q: How much should I risk per trade?

Between 1% and 3% of capital for most profiles, and toward the lower end the smaller the account or the less proven the strategy. The specific figure should come from a calculation, not an instinct: look at the structure’s maximum loss, divide the capital you accept losing by that figure, and round down. If the result is zero contracts, the trade is too large for your account — and that is a valid answer.

Q: How do I control correlation risk?

With two complementary tools. First, beta-weighted delta: translate the whole book into index deltas and get your real net directional bet as a single number. Second, explicit sector limits: no theme — technology, energy, banks — above a defined share of total risk. And remember that in crises correlations tend toward 1, so the scenario you must be able to withstand is everything falling together.

ES: Gestión del Riesgo PT: Gestão de Risco

The discipline that decides whether a statistical edge ever materialises: how much to risk, how to spread it, and what to do when something goes wrong.

Why it is the variable that matters most

Two traders with the same strategy and the same signals can finish the year with opposite results, and the difference is almost never in the analysis. It is in how much they risked per trade, in how that risk was spread, and in what they did when a position went against them. The mathematical reason is the asymmetry of losses: a 50% drawdown requires a 100% gain to recover, and a 70% drawdown requires 233%. Because recovery is disproportionately harder than decline, avoiding large losses is worth more than capturing large gains. That asymmetry — not a preference for caution — is the entire foundation of the discipline.

Risk per trade

The first decision is how much capital a single trade may lose. Tested references place that limit between 1% and 3% of capital, with 5% as an aggressive ceiling. The reasoning is survival: at 2% per trade, a run of ten consecutive losses — entirely possible even with a good strategy — costs around 18% of the account, painful but recoverable. At 10% per trade, that same run leaves the account at 35% of its value and requires nearly tripling to get back to par. The practical conversion is direct: divide the capital you are willing to lose by the structure’s maximum loss, and that quotient is your contract count, always rounding down.

Portfolio risk and correlation

Controlling each trade is not enough if they all bet on the same thing. Twenty positions at 2% are not a 2% risk: if correlated, they are a 40% position in disguise. Three controls complement the per-trade limit. Beta-weighted delta, which translates the whole book into index deltas and reveals the real net directional bet. Sector and underlying limits, so no single theme concentrates more than a defined percentage. And total deployed risk — what you would lose if every position hit its maximum simultaneously; a common limit places that figure between 15% and 25% of capital.

Exit rules, defined before entry

Management that works is management decided before emotional pressure exists. Four elements are worth fixing in writing at entry. The profit target: in credit structures, closing at 50% of the credit is the most tested guideline. The loss limit: closing when the loss reaches one or two times the credit collected, or a defined percentage of the debit paid. The time limit: closing or rolling at 21 days to expiration, where gamma begins making the position ungovernable. And the invalidation criterion: what specific fact would demonstrate the original thesis is broken, regardless of price.

The part that is not arithmetic

No rule helps if it is not followed, and non-compliance has recognisable patterns. Revenge trading after a loss, doubling size to win it back in one go. Moving the stop because "it will surely bounce". Averaging down into a position whose thesis is already invalidated. Increasing size after a winning streak, precisely when confidence is highest and judgement poorest. The three antidotes that work best are structural rather than willpower-based: writing the plan before entry, using automated orders that execute the exit without an intervening decision, and keeping a log of every trade with its thesis and outcome, because what is not measured is remembered selectively and favourably.

Frequently Asked Questions

How much should I risk per trade?

Between 1% and 3% of capital for most profiles, and toward the lower end the smaller the account or the less proven the strategy. The specific figure should come from a calculation, not an instinct: look at the structure’s maximum loss, divide the capital you accept losing by that figure, and round down. If the result is zero contracts, the trade is too large for your account — and that is a valid answer.

Should I use stop-loss orders on options?

With care. On illiquid options, a market stop can fill on the wrong side of a very wide bid-ask spread and turn a moderate loss into a large one, especially in stress when spreads blow out. The more common alternatives are price alerts that prompt you to close manually with a limit order, and conditional orders on the underlying rather than on the option, which tend to fill better.

How do I control correlation risk?

With two complementary tools. First, beta-weighted delta: translate the whole book into index deltas and get your real net directional bet as a single number. Second, explicit sector limits: no theme — technology, energy, banks — above a defined share of total risk. And remember that in crises correlations tend toward 1, so the scenario you must be able to withstand is everything falling together.

What do I do after a losing streak?

Cut size, not increase it. The instinctive reaction — doubling to win it back at once — is what turns a normal streak into structural damage. A simple and effective protocol: after three consecutive losses, halve position size until you string together two winners. And review the log to distinguish statistical bad luck, which is normal and expected, from something that changed in the market or your execution, which does require correction.

Does diversification reduce risk on its own?

Only if the assets are genuinely uncorrelated, and in equities that is rarer than it looks. Fifteen different technology stocks diversify almost nothing against a sector selloff. Diversification that works crosses asset classes — equities, bonds, commodities — structures with different Greek profiles, and staggered entry times, not simply the number of open tickets.