At the same capital base, a chosen 2% budget is twice a 1% budget before volume constraints. After successive losses the base also changes. Under the stated model, ten full losses reduce capital by about 9.56% at 1% and 18.29% at 2%. These fractions are examples, not safe settings assigned by a trader’s experience or account age.

Smaller percentage, not necessarily smaller amount

In a fictional account, balance is PLN 10,000 and equity PLN 20,000 owing to positive unrealized P/L. A uses 1% of equity; B uses 2% of balance. Both rules currently give PLN 200. Compare percentages directly only when their bases agree.

Scroll sideways to see the full diagram

Different percentage bases1% of PLN 20,000 equity gives PLN 200. 2% of PLN 10,000 balance also gives PLN 200. On the same PLN 10,000 base, 1% gives PLN 100 and 2% PLN 200.A: 1% of PLN 20,000 equity20,000 × 0.01 = 200 PLNB: 2% of PLN 10,000 balance10,000 × 0.02 = 200 PLNOn the same PLN 10,000 base:1% = PLN 100; 2% = PLN 200
Bar lengths show bases, not budgets. Bases differ twofold while budgets are equal.

Record the base, measurement time and amount beside the percentage. Do not read “1%” alone as less risky than any “2%”. Account for specifications, volume steps, costs and the whole account. Equity changes with unrealized P/L, so this example’s equality is not permanent.

Assumptions: Balance and equity are fictional. Unrealized P/L is not a guaranteed closing result. We compare budgets under two explicit rules at one instant, without recommending either percentage or base.

What exactly are you risking under an "X percent rule"?

A percentage rule specifies a chosen modeled loss budget, not the share of capital committed as notional or margin. Stop distance and instrument value estimate price loss at an assumed fill. Ordinary stops can fill differently, and costs, gaps and currency conversion can change actual loss, so the modeled budget is not a guaranteed maximum.

Take a 10,000 USD account and a stop loss 30 pips away on EUR/USD:

  • For a hypothetical USD 100 price-loss budget under the same assumptions, unrounded volume is 0.3333 lot. At a 0.01 step, 0.33 lot models USD 99 and EUR 33,000 notional before costs. This is not a guaranteed maximum loss.
  • For a hypothetical USD 200 price-loss budget, 30-pip stop and USD 10 per pip per standard lot, unrounded volume is 0.6667 lot. At a 0.01 step, 0.66 lot models USD 198 and EUR 66,000 notional; 0.67 models USD 201 and exceeds that price-loss budget before costs.
  • For a hypothetical USD 50 price-loss budget under the same assumptions, unrounded volume is 0.1667 lot. At a 0.01 step, 0.16 lot models USD 48 and EUR 16,000 notional; 0.17 models USD 51 and exceeds that price-loss budget before costs.

In the stated price-loss model, position size equals budget divided by the product of stop distance and pip value for the specified volume unit. The percentage is one input, not a universal rule. Volume steps, minimums, separate costs and actual fills require additional checks.

Why the difference between 1 and 2 percent is not linear

With a constant fractional modeled loss f, exact fills and no other changes, remaining capital after N losses is C0(1−f)N. Monetary drawdown is C0[1−(1−f)N], and percentage drawdown is [1−(1−f)N] × 100%. The remaining-capital expression is not itself drawdown.

Drawdown after ten consecutive losses, by risk per trade
0.5 percent per trade−4.9% of capital
1 percent per trade−9.6% of capital
2 percent per trade−18.3% of capital
3 percent per trade−26.3% of capital
5 percent per trade−40.1% of capital
10 percent per trade−65.1% of capital

The ten-loss sequence is a stress scenario, not a forecast of its probability. Twenty exact fractional losses leave 0.98²⁰ ≈ 66.7608% of starting capital at 2%, a 33.2392% decline; at 1% the decline is 18.2093%. A historical win rate alone does not establish the frequency or dependence of future losing runs.

Recovery uses the smaller remaining base. After the model’s ten 1% losses, recovery needs about 10.57%; after ten 2% losses, about 22.39%. An exact 50% decline needs 100% growth on the remainder; exact 70% and 90% declines need about 233.33% and 900%. These identities do not promise recovery.

When is 1 percent too little, and 2 percent already too much?

A chosen risk fraction determines the arithmetic of a specified loss sequence. It does not establish a typical strategy win rate or guarantee psychological tolerance, recovery or a profitable edge. Treat 1% as an example risk budget to evaluate against the full strategy and portfolio.

Before considering a change, review the following inputs; they are not permission to raise risk after a fixed number of profitable months.

  1. Complete observations, costs and execution assumptions from a stated window. Demo can explain mechanics, but does not demonstrate live fills; an account’s age does not prove a durable edge.
  2. Documented wins, losses and their sizes in the trading journal. Win rate must be assessed with payoffs and costs rather than a universal 50% threshold.
  3. Assess payoff distribution and costs. Positive expectancy depends jointly on win probability, payoff sizes and costs, not a universal 50% win-rate or 1.5R threshold. In a fixed-base linear payoff model, doubling exposure doubles expected monetary P/L and standard deviation, while variance quadruples. Trade-level compounding and constraints require separate analysis.
  4. The financial consequences of adverse sequences, combined positions and execution deviations. A chosen stress loss is not a universal measure of psychological tolerance.

These checks do not derive a suitable percentage from the drawdown formula. A lower fraction can still be excessive for an account; positive estimated expectancy does not make a larger fraction safe. Keep base, measurement time and portfolio exposure explicit.

The risk-scaling trap

Hypothetical risk increase: every loss is exactly 3% of current capital, with no costs or other outcomes. Five losses produce 1−0.97⁵ ≈ 14.13% drawdown; ten produce 1−0.97¹⁰ ≈ 26.26%. Returning after those ten losses needs about 35.61% growth on the remainder. No frequency of such runs or past quarterly performance is assumed.

In a fixed-base linear model, exposure scales monetary outcomes; actual compounding, execution and costs can change that relationship. Financial loss and its effect on decisions are separate questions. Record departures from the plan rather than assuming that a particular percentage causes the same behaviour in everyone.

Moving an ordinary stop to entry changes the intended exit level but does not guarantee a zero net result. Costs and execution away from the trigger can leave a loss. The article on break-even and moving the stop loss explains the mechanism; choose any change under a prewritten procedure rather than assuming it improves every strategy.

How to scale risk sensibly over time

There is no experience calendar that automatically justifies increasing from 0.5% to 2%. Review the following dimensions before changing a rule:

  1. State the capital base. Balance and equity can differ; record the chosen base, time and monetary budget.
  2. Inspect the data. Include costs, payoff distribution and dependence. A fixed number of months does not establish a reliable estimate.
  3. Check executable size. Round down to the specified volume step and respect the minimum. If the minimum exceeds the budget, increasing size is not a solution.
  4. Test combined exposure. Include open positions, independently executable pending orders and adverse execution scenarios.
  5. Write change and review rules. Define the action, evidence and loss capacity required. Neither 1% nor 2% guarantees that the account survives future outcomes.

What to do tomorrow

  1. Record the complete sequence. Use the journal to identify realised losses and departures from the plan. The longest observed run is one sample fact, not a future risk bound or a substitute for payoff and exposure analysis.
  2. Recalculate remaining equity and drawdown separately. For n equal fractional losses, remaining equity as a share of the start is (1-r)^n; drawdown as a share is 1-(1-r)^n. For r=0.02 and n=10 these are 0.817073 and 0.182927, or 18.29% drawdown.
  3. Document the assumptions behind a proposed change. Recheck the base, costs, actual size and consequences of adverse scenarios. Do not use account age or a universal win-rate threshold as permission to increase risk.
  4. Write the rule before using it. State the fraction, base, rounding, costs, combined exposure and review actions. A calendar date or short profitable run alone does not establish suitability.