How the compound interest calculator works

Compounding means each period's return is earned on the balance that already includes previous gains. The math is simple: final = initial × (1 + r)ⁿ, where r is the return per period and n is the number of periods. Its consequences are not intuitive. At 1% per period, 100 periods do not give you +100% — they give you about +170%, because gains stack on gains.

Worked example: $10,000 at 2% per month for 3 years (36 months) grows to about $20,400 — a double — even though 2% sounds small. Push the same numbers to 10% per month and the calculator will show a fantasy figure in the millions. That contrast is the point: the formula is honest, but the assumption of a constant positive return is not. Real trading returns arrive unevenly and include losing months, which is why the calculator's note warns you to treat it as an illustration of consistency, not a forecast.

• Use it to: understand why surviving and staying consistent beats chasing one big year.
• Do not use it to: project a backtest's best month forward for a decade.

How the Kelly criterion calculator works

The Kelly criterion answers one question: given your edge, what fraction of your bankroll maximizes long-term growth? The formula is f* = W − (1 − W) / R, where W is your win rate and R is your average win divided by average loss. Example: a 45% win rate with a 2.0 reward-to-risk ratio gives f* = 0.45 − 0.55 / 2 = 17.5% of bankroll per trade.

Full Kelly is mathematically optimal and emotionally unbearable — drawdowns above 50% are routine. That is why the calculator highlights half Kelly: roughly three-quarters of the growth at half the volatility. Many professionals size below even that. If the calculator reports no positive edge, believe it: no position size can turn a negative-expectancy strategy profitable.

Where the inputs should come from

Both calculators are only as honest as their inputs. Win rate and reward-to-risk estimated from ten trades are noise; estimated from several hundred backtested trades across different market conditions, they start to mean something. That is the intended workflow on this site: run a backtest over seven years of data, read the win rate and profit factor from a large sample, then bring those numbers here to think about sizing.