Risk Management in Trading: Position Size, Stops and Drawdown

Trading performance is path-dependent. Two strategies can produce the same final return but expose the account to very different drawdowns along the way. If losses become too large to tolerate financially or psychologically, the trader may stop before the strategy has a chance to recover.

Risk management aims to keep individual mistakes, unusual market events, and ordinary losing streaks within boundaries the account can survive. The objective is controlled participation, not the elimination of losses.

A stop should represent the level or condition where the original trade thesis no longer holds. For a breakout, that may be a return inside the prior range; for a trend pullback, it may be a break of structure; for a volatility system, it may be an ATR-based distance.

Placing a stop solely to achieve a preferred dollar loss can distort the setup. A more coherent process defines the market-based invalidation first, then adjusts position size so the monetary risk fits the account rule.

A simple framework is to choose a maximum amount the account is allowed to lose if the stop is reached and divide that amount by the risk per unit. Futures, options, leveraged products, and contracts with multipliers require the correct point or tick value in the calculation.

The result should also respect minimum lot sizes, broker margin requirements, liquidity, and any maximum contract or share cap. If the correct risk-based size is below the minimum tradable unit, the trade may simply be too large for the account at that stop distance.

There is no universal percentage that is appropriate for every trader, product, or strategy. The suitable amount depends on drawdown tolerance, strategy variance, leverage, account size, trade frequency, and how many correlated positions can be open at once.

Historical testing can estimate losing streaks and drawdowns, but future outcomes can be worse. Conservative sizing leaves room for model error, slippage, gaps, and a sequence of losses that exceeds the backtest.

Several individually small positions can create a large combined risk. Account-level controls can cap the number of open trades, total notional exposure, total stop risk, or contracts per connection.

Correlation matters as well. Long positions in several closely related equity indices or metals may behave like one concentrated trade when markets move sharply. Portfolio risk should therefore look beyond symbol count.

Broker margin tells you how much capital is required to hold a leveraged position; it does not tell you how much the position can lose. A highly leveraged position can move far beyond the margin amount if price gaps or volatility expands.

Risk should be measured from price movement, contract value, size, and the planned exit. Margin is an operational constraint that must be respected, but it should not be used as a substitute for a loss calculation.

A strategy with a healthy long-run expectancy can still produce consecutive losses. The higher the variance and the lower the win rate, the more important it is to model sequences rather than focusing only on the average trade.

Reducing risk during a significant drawdown can slow the rate of further losses, although it also slows recovery if the strategy rebounds. Any drawdown-based sizing rule should be defined and tested in advance rather than invented during stress.

A maximum daily loss can stop new entries after the account has reached a pre-defined boundary. This can be especially useful for high-frequency or intraday approaches where many trades can occur in a short period.

The limit should reflect the normal variance of the strategy. If it is too tight, ordinary losing sequences may repeatedly stop the system before its edge can play out; if it is too loose, it may fail to meaningfully contain abnormal behaviour.

Test what happens when slippage is worse, spreads widen, stops fill beyond the requested level, several correlated trades lose together, or the strategy experiences a longer losing streak than seen historically. Stress testing asks whether the account remains operable when assumptions deteriorate.

Monte Carlo resampling and parameter sensitivity can also help estimate a range of possible equity paths. These tools do not predict the future, but they can expose whether a risk plan depends on an unusually smooth historical sequence.

A good review should ask both whether the strategy signal was valid and whether the amount of risk was appropriate. A trade can be correctly identified but oversized, or poorly identified but small enough that the damage remains controlled.

Track actual risk, planned risk, slippage, maximum adverse excursion, drawdown, and breaches of risk rules. Over time, these records reveal whether losses come from the strategy itself or from inconsistent sizing and execution.