Capital Management

Size comes from the stop, not from conviction

The fixed-risk rule tells you how much you can lose if the stop is hit, not how much to buy. With a 10,000 account and 1% risk, each trade can cost at most 100. If the stop sits 2 away from the entry price, the size is 100 / 2 = 50 shares; if it sits 5 away, it is 20. A wider stop gives a smaller position, so every trade risks the same amount.

Why drawdown weighs more than it looks

Losses and gains are not symmetric. After losing 10% you need an 11.1% gain to get back to where you started; after 25%, 33.3%; after 50%, 100%. The deeper the hole, the harder it is to climb out, which is why the first job of money management is to keep the hole shallow.

Losing streaks are normal: size for them

With a 50% win rate, the probability of at least one run of five straight losses somewhere in 100 trades is 81%. That is not bad luck, it is what to expect. At 1% risk per trade those five losses leave the account 4.9% down; at 5%, 22.6% down. The system can be exactly the same; what decides whether you survive the streak is position size.

What you will learn

1-2% Rule Per Trade

Never risk more than 1-2% of your capital on a single trade. With this, you can withstand 50 consecutive losses before losing everything. Example: $10,000 account → Maximum risk $100-200 per trade.

Kelly Criterion (Optimal Size)

Optimal position-fraction formula: Kelly% = WinRate − (1−WinRate)/PayoffRatio. It maximises long-term growth, but full Kelly is extremely volatile (50%+ drawdowns) — use ¼ or ½ Kelly. Try your own numbers in the interactive Kelly calculator above.

Position Pyramid (Scaling)

Add to winning positions (not losing ones). Scaling In: Enter with 50%, add 25% on confirmation, final 25%. Scaling Out: Take 50% at TP1, let 50% run with stop at breakeven.

Diversification

Don't have all your trades in the same asset or sector. Correlation: If BTC falls, many altcoins fall too (high correlation). Diversify among assets with low correlation.

Correlation: real diversification

Holding five positions is not diversifying if all five rise and fall together. Buying Apple, Microsoft, Nvidia, a Nasdaq ETF and bitcoin at once is, in practice, one leveraged bet on tech 'risk-on': when one falls, they all fall. Real diversification requires lowly or negatively correlated assets (stocks + bonds + gold, or sectors that don't move in unison). Before opening another position, ask: 'is this a new bet, or more of the same?'.

Sequence-of-returns risk

The ORDER in which your gains and losses arrive matters as much as their average, especially if you add or withdraw money. Two accounts with the same average return can end up very differently: the one that suffers a big losing streak early (when capital is already high, or just as you start withdrawing) may never recover, because a 50% drop needs a +100% to get back. That is why limiting drawdown early and not over-exposing at the start is not excessive caution: it is mathematical survival.

Risk of ruin

The probability that a losing streak alone leaves the account with no capital left to trade. Win rate does not decide it on its own: it comes from how much you risk per trade, your reward-to-risk ratio and how many trades you take. Two systems with the same 55% win rate have very different risks of ruin if one risks 1% and the other 10%. The consequence is uncomfortable: a system that is profitable on average can ruin you before its edge ever shows up, which is why position size is settled before the entry. The Monte Carlo simulator estimates it on your own numbers.

Total open risk, not just per trade

The 1-2% rule applies to each entry, but what takes an account down is the sum of everything open at once. Five positions at 2% is 10% of capital at risk simultaneously, and if they are correlated they behave like a single 10% bet. That is why you need a second ceiling — a maximum aggregate risk at any moment — respected even when every entry, taken separately, obeys the rule. It is the limit that is missing when someone follows the 1% rule to the letter and still takes a 15% drawdown in one session.

Minimum Recommended Ratios

There is no universal "minimum" R:R — it depends on your win rate. Break-even R = (1−WinRate)/WinRate: at 50% you need >1:1, at 40% you need >1:1.5, but at 65% accuracy you can be profitable below 1:1. Match the target R:R to your system's real hit rate, not to a fixed rule.

Calculation Before Entry

Measure every trade in R (1R = your risk from entry to stop). Place the stop where the idea is invalidated (structure or ATR), size the position so that distance equals a fixed % of capital, then read the reward as a multiple of R. Whether a given R:R is "enough" is decided by your win rate and expectancy — not by a 1:2 rule.