Define risk before entry, not after the loss

Risk management asks: if the analysis is wrong, how much capital could be lost, and can the account continue? A method can work repeatedly while a few oversized positions, price gaps, or execution errors undermine its results. Risk is therefore more than “having a stop.” It combines position sizing, execution quality, correlated exposures, and rules for pausing activity.

All numbers in this article are hypothetical, and the percentages are not personal recommendations. A suitable budget depends on financial circumstances, tolerance for loss, instrument liquidity, and actual evidence about the method. Essential living funds or money needed soon should not become the basis of a high-risk experiment. No formula eliminates the possibility of loss.

Five numbers that must not be confused

Account equity measures the capital available to absorb outcomes. Position notional measures exposure to price changes. Margin is collateral required to maintain certain positions. Stop distance is the difference between entry and the planned exit level. Monetary risk estimates the loss if that exit executes, including relevant costs.

These are different quantities. Posting 100 in margin for a position with a notional value of 1,000 does not mean the maximum risk is 100. Losses and settlement rules depend on the instrument, contract, and provider. Some leveraged products can produce losses beyond the initial deposit. Understand contract size, order increments, tick value, profit-and-loss calculation, and margin rules before placing an order.

Analytical invalidation and an exit order are different

Invalidation is a condition under which the trade hypothesis no longer holds—for example, loss of a structural low on the decision timeframe. An exit order is the execution mechanism used to respond. Identify logical invalidation first, then adapt quantity to its distance. Moving the stop arbitrarily closer merely to obtain a larger position changes the original analysis.

If invalidation requires a candle close, an adverse move before that close can exceed the paper distance. If the exit is an automated price-triggered order, a temporary penetration may activate it. State this distinction explicitly. A backtest based on closing prices should not be evaluated as though live execution used identical intrabar triggers.

Investor.gov explains that a conventional stop order becomes a market order when triggered, and its execution price is not guaranteed. A stop-limit order adds a price constraint but may not execute at all. Exact mechanisms differ across venues and platforms. Check the order type, last-trade or reference-price trigger, and behavior during disruptions rather than assuming every “stop” works identically.

Position sizing: convert a loss budget into quantity

For a simple spot purchase, the risk budget is equity multiplied by the selected risk fraction. Estimated loss per unit is the distance from entry to planned exit plus estimated costs and slippage per unit. Divide the budget by that loss and round quantity down to an allowed increment.

Suppose hypothetical equity is 10,000, the illustrative budget is 0.5%, entry is 100, and the exit level is 95. The budget is 50 and the price distance is 5. Ignoring costs gives 10 units and a position value of 1,000. Including estimated round-trip costs and slippage of 0.25 per unit reduces quantity to 50 divided by 5.25, or approximately 9.52 units.

Hypothetical before-cost example: equity 10,000, budget 50, exit distance 5 and quantity 10; including costs reduces the permitted quantity.
Hypothetical before-cost example: equity 10,000, budget 50, exit distance 5 and quantity 10; including costs reduces the permitted quantity.

If orders require whole units, the quantity is 9, not 10. Estimated loss is 47.25, and entry notional is 900. An unleveraged spot purchase also faces a buying-power constraint: quantity cannot exceed available funds after entry costs. Some charges are fixed or nonlinear; replace the simple cost assumption with the actual schedule in that case.

Contracts and derivatives

For a contract with linear profit and loss, estimate risk per contract as the number of ticks to the exit multiplied by monetary tick value, plus contract costs. Divide the loss budget by this amount. If the budget supports only 0.4 contracts but the minimum trade is one contract, the conclusion is that the instrument does not fit this budget—not that risk should automatically be increased.

Inverse cryptocurrency contracts, options, and instruments with nonlinear payoffs require their own calculations. Option delta and implied volatility can change, and a displayed leverage number alone does not explain the outcome. Do not apply the spot formula indiscriminately to every product.

Leverage, liquidity, and execution risk

Leverage generally changes collateral requirements; it does not remove the risk of a given notional exposure. Holding the same number of units leaves the same underlying price exposure, while financing costs and liquidation rules may add further risks. A liquidation threshold is not a substitute for a planned exit.

In a shallow market, an order may consume multiple order-book levels. Wide spreads, slippage, gaps when trading resumes, and halts can take realized loss beyond the intended budget. Estimate slippage conservatively for the instrument, but recognize that an extreme event may fall outside the estimate. Planned risk is not a guaranteed maximum.

Prepare for operational failure too: connectivity loss, rejected orders, partial fills, and one leg of a multi-part trade closing without the others. Determine whether protective orders reside on a broker's server or depend on local software. An acknowledgement that an order was submitted is not evidence that it filled.

Portfolio risk: three positions can be one shared bet

Different symbols do not automatically create diversification. Simultaneous exposure to Bitcoin, an altcoin, and a cryptocurrency-sensitive equity may depend on a common market factor. Correlations are not constant and can change under stress. Do not infer independence from different asset names.

Maintain a record of each position's open risk, aggregate planned loss to exit levels, and common factor exposures. Three positions each carrying an illustrative 0.5% planned risk add up to 1.5%. That sum is neither a guaranteed worst-case loss nor necessarily inclusive of gaps, operational failure, or changing correlations. Examine a simultaneous adverse-move scenario separately.

Set limits on aggregate open risk, daily or weekly loss, and activity following an execution error before starting. Appropriate thresholds depend on the method and the account. When a limit is reached, the next trade should not exist merely to recover losses. First determine whether the issue concerns normal variation, a deteriorating method, or an operational problem.

Drawdown and the mathematics of recovery

A loss and its recovery are not symmetric. After a 10% drawdown, roughly 11.1% growth is required to return to the starting balance. A 20% loss requires 25%, a 30% loss about 42.9%, and a 50% loss 100%. Subsequent gains are calculated on a smaller capital base.

Arithmetic, not a forecast: drawdowns of 10%, 20%, 30% and 50% require approximately 11.1%, 25%, 42.9% and 100% gains to recover.
Arithmetic, not a forecast: drawdowns of 10%, 20%, 30% and 50% require approximately 11.1%, 25%, 42.9% and 100% gains to recover.

Required recovery is d divided by one minus d, where d is drawdown expressed as a decimal. For example, 0.2 divided by 0.8 equals 0.25. This arithmetic explains the importance of limiting drawdown; it does not guarantee recovery or establish how long recovery will take.

If every unsuccessful trade loses exactly 1% of current equity, ten consecutive losses leave approximately 90.44% of starting capital—a drawdown of about 9.56%. The example assumes no fees or slippage and does not predict the probability of that sequence. Under the same assumptions, risking 5% produces approximately 40.13% drawdown after ten losses. Larger risk can make the recovery path substantially harder.

Reward-to-risk alone is not enough

R denotes the initial planned risk unit. A result of positive 2R means the recorded profit is twice that basis; specify whether the journal uses net or gross profit and keep the definition consistent. A target at 2R is a plan, not an achieved average. Early exits, costs, and partial fills can reduce realized rewards.

Simple expectancy is win probability multiplied by average win, minus loss probability multiplied by average loss. In a hypothetical sample with a 40% win rate, average win of 2R, and average loss of 1R, gross expectancy is 0.2R. If average costs are 0.1R per trade and are not already included in the outcomes, net expectancy is 0.1R. This is not a guarantee of future profitability: a small sample, changing conditions, and inaccurate cost assumptions can alter the estimate.

Review win rate, average wins and losses, drawdown, losing sequences, and actual costs together. A few successful trades or an attractive target ratio are not a risk assessment. Exposure should be consistent with documented performance and an acceptable loss tolerance, not only the appeal of a setup.

An execution protocol and a practice exercise

Before entry, record the instrument and its specifications, equity basis, trigger, invalidation, exit-order type, loss budget, estimated costs, rounded quantity, and risk in other open positions. After execution, verify actual quantity, average fill price, and the presence of the intended protective order.

As an exercise, recalculate the entry-at-100 example with an exit at 90 instead of 95. With a budget of 50 and illustrative per-unit costs of 0.25, quantity is approximately 4.87. A wider distance implies a smaller position. Do not increase the loss budget merely to preserve the old number of units.

  • Is the budget calculated using current equity and a clearly stated fraction?
  • Has quantity been rounded down after accounting for costs?
  • Is the stop distinct from liquidation and from a guarantee of execution price?
  • Have aggregate exposure and a gap scenario been examined?
  • Is there a pause-and-review rule after a breach or operational failure?

The Nexito principle: risk management does not seek to eliminate every losing trade. It seeks to prevent a tolerable error from becoming damage that interrupts the entire process. Define an acceptable loss first, then align the instrument, exit distance, and position size with it.