Position sizing with Kelly criterion and bankroll management
Overall
score
61%
Does it follow best practices?
If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:
npx tessl skill review --optimize ./path/to/skillValidation for skill structure
Calculate optimal position sizes using Kelly criterion, fractional Kelly, and portfolio-level allocation.
/kelly 0.45 0.55 10000 Market price, your prob, bankroll
/kelly "Trump 2028" 0.55 --bank 10k Calculate for specific market
/kelly --half 0.45 0.55 10000 Half Kelly (safer)
/kelly --quarter 0.45 0.55 10000 Quarter Kelly (conservative)/size 10000 --risk 2% Size for 2% risk per trade
/size 10000 --max-position 25% Max 25% in single position
/size portfolio --rebalance Rebalance to target weights/edge 0.45 0.55 Calculate edge (prob - price)
/edge "Trump 2028" --estimate 0.55 Edge vs market priceimport { createSizingCalculator } from 'clodds/sizing';
const sizing = createSizingCalculator({
// Bankroll
bankroll: 10000,
// Kelly fraction (1 = full, 0.5 = half)
kellyFraction: 0.5,
// Limits
maxPositionPercent: 25,
maxTotalExposure: 80,
});// Binary outcome (YES/NO market)
const size = sizing.kelly({
marketPrice: 0.45, // Current price
estimatedProb: 0.55, // Your probability estimate
bankroll: 10000,
});
console.log(`Optimal bet: $${size.optimalSize}`);
console.log(`Edge: ${size.edge}%`);
console.log(`Kelly %: ${size.kellyPercent}%`);
console.log(`Expected value: $${size.expectedValue}`);// Half Kelly (recommended for most traders)
const halfKelly = sizing.kelly({
marketPrice: 0.45,
estimatedProb: 0.55,
bankroll: 10000,
fraction: 0.5, // Half Kelly
});
// Quarter Kelly (very conservative)
const quarterKelly = sizing.kelly({
marketPrice: 0.45,
estimatedProb: 0.55,
bankroll: 10000,
fraction: 0.25,
});
console.log(`Full Kelly: $${sizing.kelly({...}).optimalSize}`);
console.log(`Half Kelly: $${halfKelly.optimalSize}`);
console.log(`Quarter Kelly: $${quarterKelly.optimalSize}`);// For markets with 3+ outcomes
const multiKelly = sizing.kellyMultiOutcome({
outcomes: [
{ name: 'Trump', price: 0.35, estimatedProb: 0.40 },
{ name: 'DeSantis', price: 0.25, estimatedProb: 0.20 },
{ name: 'Haley', price: 0.15, estimatedProb: 0.15 },
{ name: 'Other', price: 0.25, estimatedProb: 0.25 },
],
bankroll: 10000,
fraction: 0.5,
});
for (const alloc of multiKelly.allocations) {
console.log(`${alloc.name}: $${alloc.size} (${alloc.percent}%)`);
}// Optimal allocation across multiple markets
const portfolio = sizing.kellyPortfolio({
positions: [
{ market: 'Trump 2028', price: 0.45, prob: 0.55 },
{ market: 'Fed Rate Cut', price: 0.60, prob: 0.70 },
{ market: 'BTC > 100k', price: 0.30, prob: 0.40 },
],
bankroll: 10000,
correlations: correlationMatrix, // Optional
fraction: 0.5,
});
console.log('Optimal Portfolio:');
for (const pos of portfolio.positions) {
console.log(` ${pos.market}: $${pos.size}`);
}
console.log(`Total exposure: ${portfolio.totalExposure}%`);// Reduce size when less confident
const size = sizing.kellyWithConfidence({
marketPrice: 0.45,
estimatedProb: 0.55,
confidence: 0.7, // 70% confident in estimate
bankroll: 10000,
});
// Size is reduced proportionally to confidence
console.log(`Confidence-adjusted size: $${size.optimalSize}`);// Calculate edge
const edge = sizing.calculateEdge({
marketPrice: 0.45,
estimatedProb: 0.55,
});
console.log(`Edge: ${edge.edgePercent}%`);
console.log(`EV per dollar: $${edge.evPerDollar}`);
console.log(`Implied odds: ${edge.impliedOdds}`);
console.log(`True odds: ${edge.trueOdds}`);// Size based on risk per trade
const size = sizing.riskBased({
bankroll: 10000,
riskPercent: 2, // Risk 2% per trade
stopLossPercent: 10, // 10% stop loss
});
console.log(`Position size: $${size.positionSize}`);
console.log(`Max loss: $${size.maxLoss}`);| Fraction | Risk Level | Use Case |
|---|---|---|
| Full (1.0) | Aggressive | Mathematical optimum, high variance |
| Half (0.5) | Moderate | Most traders, good balance |
| Quarter (0.25) | Conservative | New traders, uncertain edges |
| Tenth (0.1) | Very Safe | Learning, small edges |
| Edge | Recommendation |
|---|---|
| < 2% | Don't trade |
| 2-5% | Small size (quarter Kelly) |
| 5-10% | Normal size (half Kelly) |
| 10%+ | Larger size, verify edge |
f* = (p * b - q) / b
Where:
f* = fraction of bankroll to bet
p = probability of winning
q = probability of losing (1 - p)
b = odds received (1/price - 1)Edge = Estimated Prob - Market Price
EV = Edge * Bet SizeIf you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.