@ebowwa/quant

Multi-market quantitative trading tools and formulas for AI trading systems.

Features

Multi-Market Support

• Prediction Markets - Binary outcomes, AMM math, Kelly criterion
• Traditional Markets - Stocks, crypto, forex technical analysis
• Derivatives - Options, futures, perpetuals risk management

Modules

prediction - Prediction Market Tools

import { kellyCriterion, ammBuyCost, detectArbitrage } from '@ebowwa/quant/prediction';

// Kelly sizing for binary bets
const kelly = kellyCriterion(0.65, 0.55, 1000); // Your prob, market price, bankroll
console.log(kelly.kellyFraction); // Optimal fraction to bet

// AMM cost calculation (constant product)
const cost = ammBuyCost({ poolYes: 1000, poolNo: 1000, k: 1000000 }, 'yes', 100);

// Detect arbitrage
const arb = detectArbitrage(0.48, 0.48); // YES + NO < 1 = arb opportunity

indicators - Technical Indicators

import { sma, ema, rsi, macd, bollingerBands, atr } from '@ebowwa/quant/indicators';

const prices = [100, 102, 101, 103, 105, 104, 106];

// Moving averages
const sma20 = sma(prices, 20);
const ema12 = ema(prices, 12);

// RSI
const rsi14 = rsi(prices, 14);

// MACD
const { macd: macdLine, signal, histogram } = macd(prices, 12, 26, 9);

// Bollinger Bands
const { upper, middle, lower } = bollingerBands(prices, 20, 2);

risk - Risk Management

import {
  sharpeRatio, sortinoRatio, calculateDrawdown,
  calculateVaR, fixedFractionalSize
} from '@ebowwa/quant/risk';

// Position sizing
const size = fixedFractionalSize(10000, 0.02, 100, 95); // 2% risk

// Risk metrics
const sharpe = sharpeRatio(returns, 0.04, 252);
const { maxDrawdown } = calculateDrawdown(equityCurve);
const { var95, expectedShortfall95 } = calculateVaR(returns);

stats - Statistical Functions

import {
  pearsonCorrelation, linearRegression,
  distributionStats, autocorrelation
} from '@ebowwa/quant/stats';

// Distribution analysis
const stats = distributionStats(returns);
console.log(stats.skewness, stats.kurtosis);

// Correlation
const corr = pearsonCorrelation(assetReturns, marketReturns);

// Linear regression
const { slope, intercept, r2 } = linearRegression(x, y);

// Time series
const acf = autocorrelation(prices, 20);

patterns - Pattern Detection

import {
  findSupportResistance, detectDoubleTop,
  detectHeadAndShoulders, detectTriangles
} from '@ebowwa/quant/patterns';

// Support/Resistance levels
const levels = findSupportResistance(high, low);

// Pattern detection
const doubleTops = detectDoubleTop(high, low);
const headShoulders = detectHeadAndShoulders(high, low);
const triangles = detectTriangles(high, low);

Installation

bun add @ebowwa/quant

API Reference

Prediction Markets

| Function | Description | |----------|-------------| | kellyCriterion(yourProb, marketPrice, bankroll) | Calculate optimal bet size | | ammBuyCost(state, outcome, shares) | Cost to buy shares from AMM | | ammSharesReceived(state, outcome, cost) | Shares received for cost | | detectArbitrage(yesPrice, noPrice) | Find arb opportunities | | brierScore(predictions) | Measure forecast accuracy | | calculateCalibration(predictions) | Calibration analysis |

Technical Indicators

| Function | Description | |----------|-------------| | sma(data, period) | Simple Moving Average | | ema(data, period) | Exponential Moving Average | | rsi(data, period) | Relative Strength Index | | macd(data, fast, slow, signal) | MACD indicator | | bollingerBands(data, period, stdDev) | Bollinger Bands | | atr(high, low, close, period) | Average True Range | | adx(high, low, close, period) | Average Directional Index | | ichimoku(high, low, close) | Ichimoku Cloud |

Risk Management

| Function | Description | |----------|-------------| | sharpeRatio(returns, rf, periods) | Sharpe ratio | | sortinoRatio(returns, rf, periods) | Sortino ratio | | calculateDrawdown(equityCurve) | Max drawdown analysis | | calculateVaR(returns) | Value at Risk | | portfolioVolatility(weights, covMatrix) | Portfolio volatility | | beta(assetReturns, marketReturns) | Beta coefficient |

Statistics

| Function | Description | |----------|-------------| | distributionStats(data) | Full distribution analysis | | pearsonCorrelation(x, y) | Pearson correlation | | linearRegression(x, y) | Simple linear regression | | autocorrelation(data, maxLag) | ACF | | rollingStats(data, period) | Rolling statistics |

Types

The package includes comprehensive TypeScript types for all market types:

• MarketType - 'prediction' | 'equity' | 'crypto' | 'forex' | 'derivative'
• OHLCV - Standard candlestick data
• PredictionQuote - Prediction market quote
• AMMState - AMM pool state
• KellyResult - Kelly criterion output
• TradingSignal - Trading signal with confidence
• PatternMatch - Detected pattern with confidence

Future: MCP Interface

This package is designed to be wrapped by an MCP server for AI trading systems:

// Future MCP tool example
{
  name: "calculate_kelly",
  description: "Calculate optimal position size using Kelly criterion",
  inputSchema: {
    type: "object",
    properties: {
      yourProbability: { type: "number" },
      marketPrice: { type: "number" },
      bankroll: { type: "number" }
    }
  }
}

License

MIT