MeridianAlgo.js

A professional-grade quantitative finance framework for JavaScript and TypeScript, designed specifically for quantitative developers, researchers, and algorithmic traders.

Overview

MeridianAlgo is a comprehensive, modular quantitative finance framework that provides institutional-quality tools for algorithmic trading, quantitative research, and financial analysis. Built with TypeScript, it offers a complete suite of packages covering data ingestion, technical analysis, backtesting, risk management, portfolio optimization, machine learning, and execution.

IMPORTANT: Please read the DISCLAIMER before using this software. This framework is provided for educational and research purposes only.

Quick Start
Installation
Key Features
Architecture
Usage Examples
Documentation
Development
Contributing
License

Quick Start

# Install dependencies
npm install -g pnpm
pnpm install

# Build all packages
pnpm build

# Run examples
pnpm example:basic      # Basic backtest
pnpm example:advanced   # Advanced features
pnpm example:utils      # Utilities demo

Installation

For End Users

Install individual packages as needed:

# Core packages
npm install @meridianalgo/core @meridianalgo/indicators @meridianalgo/data

# Strategy and backtesting
npm install @meridianalgo/strategies @meridianalgo/backtest

# Risk and portfolio management
npm install @meridianalgo/risk @meridianalgo/portfolio

# Machine learning
npm install @meridianalgo/models

# Live execution
npm install @meridianalgo/execution

# Utilities
npm install @meridianalgo/utils @meridianalgo/cli

For Contributors

# Clone the repository
git clone https://github.com/MeridianAlgo/Javascript-Packages.git
cd Javascript-Packages

# Install dependencies
pnpm install

# Build all packages
pnpm build

# Run tests
pnpm test

Key Features

Technical Indicators (100+)

Classic Indicators

Moving Averages: SMA, EMA, WMA, DEMA, TEMA, HMA
Momentum: RSI, MACD, Stochastic, Williams %R, CCI, ROC
Volatility: Bollinger Bands, ATR, Keltner Channels, Donchian Channels
Volume: OBV, MFI, VWAP, Volume Profile
Trend: ADX, Aroon, Parabolic SAR, Supertrend

Advanced Indicators

Volatility Models: GARCH, EWMA, Realized Volatility, Parkinson, Garman-Klass
Regime Detection: Hidden Markov Models (HMM), Change Point Detection
Market Microstructure: VPIN, Order Imbalance, Kyle's Lambda, Bid-Ask Spread Analysis
Feature Engineering: PCA, Lagged Features, Rolling Statistics, Cross-sectional Ranks
Seasonality: Day-of-week Effects, Month-end Effects, Holiday Analysis

Data Management

Supported Data Sources

Yahoo Finance (fully implemented with caching)
Polygon.io (with rate limiting and WebSocket support)
Binance (REST and WebSocket streaming)
Alpaca (paper and live trading integration)
Extensible adapter pattern for custom data sources

Data Quality

Automated quality validation and gap detection
Corporate action handling (splits, dividends)
Missing data interpolation and normalization
Multi-timeframe resampling
Efficient caching mechanisms

Trading Strategies

Pre-built Strategy Templates

Trend Following (moving average crossovers, breakouts)
Mean Reversion (Bollinger Band, RSI-based)
Pairs Trading (cointegration-based statistical arbitrage)
Momentum (relative strength, time-series momentum)

Strategy Composition

Blend multiple strategies with custom weights
Voting mechanisms for signal aggregation
Regime-based strategy gating
Dynamic position sizing (Kelly Criterion, Volatility Targeting, Drawdown-aware)

Backtesting Engine

Engine Types

Time-based backtesting for bar-by-bar simulation
Event-driven architecture for tick-level precision
Walk-forward analysis for out-of-sample validation
Multi-scenario testing for robustness checks

Realistic Simulation

Configurable commission models (fixed, percentage, tiered)
Slippage modeling (fixed, volume-based, volatility-based)
Borrow fees for short positions
Corporate action adjustments
Margin and leverage constraints

Risk Management

Risk Metrics

Value at Risk (VaR): Historical, Parametric, Monte Carlo
Conditional Value at Risk (CVaR/Expected Shortfall)
Maximum Drawdown and Drawdown Duration
Volatility (realized, implied, conditional)
Beta, Correlation, and Tracking Error

Performance Metrics

Sharpe Ratio, Sortino Ratio, Calmar Ratio
Information Ratio, Treynor Ratio
Win Rate, Profit Factor, Payoff Ratio
Risk-adjusted returns and attribution analysis

Risk Monitoring

Real-time position monitoring
Automated alert system
Stress testing (scenario-based, historical, Monte Carlo)
Performance attribution (Brinson model, factor exposure)
Comprehensive reporting (PDF/HTML tear sheets)

Portfolio Optimization

Optimization Methods

Mean-Variance Optimization (Markowitz)
Black-Litterman Model (incorporating views)
Risk Parity (equal risk contribution)
Hierarchical Risk Parity (HRP)
Minimum Variance, Maximum Sharpe, Maximum Diversification

Constraint Handling

Weight constraints (min/max per asset)
Long-only or long-short portfolios
Leverage limits
Sector and factor exposure limits
Turnover constraints

Robustness Techniques

Covariance shrinkage (Ledoit-Wolf)
Resampled efficient frontier
Robust optimization
Multi-objective optimization

Machine Learning

Supported Models

Tree-based: Random Forest, Gradient Boosting (XGBoost-style)
Neural Networks: Feedforward, LSTM, GRU
Time Series: ARIMA, VAR, State-Space Models, Kalman Filter
Ensemble methods and model stacking

ML Infrastructure

Hyperparameter tuning (Grid Search, Random Search, Bayesian Optimization)
Cross-validation with time-series splits
Feature selection and importance analysis
Online learning with concept drift detection
AutoML for automated model selection
Labeling methods (Triple-Barrier, Meta-labeling)

Execution

Trading Modes

Paper trading with realistic simulation
Live trading via Alpaca API
Backtesting mode for historical analysis

Order Management

Market, Limit, Stop, Stop-Limit orders
Bracket orders (with take-profit and stop-loss)
Time-in-force options (DAY, GTC, IOC, FOK)
Pre-trade risk checks
Smart order routing
Position and order tracking

Compliance

Comprehensive audit logging
Trade compliance checks
Risk limit enforcement
Regulatory reporting support

Developer Tools

Command-Line Interface

Project initialization and scaffolding
Automated backtesting workflows
Parameter optimization utilities
Live trading deployment
Report generation

Utilities

Mathematical functions (statistics, linear algebra)
Time-series utilities (resampling, alignment)
Market calendar (trading days, holidays)
Logging and monitoring infrastructure

Architecture

MeridianAlgo is built as a monorepo with independent, composable packages:

meridianalgo-js/
├── packages/
│   ├── core/           # Core types, interfaces, and plugin system
│   ├── indicators/     # 100+ technical indicators
│   ├── data/           # Data adapters and management
│   ├── strategies/     # Trading strategy templates
│   ├── backtest/       # Backtesting engines
│   ├── risk/           # Risk metrics and monitoring
│   ├── portfolio/      # Portfolio optimization
│   ├── models/         # Machine learning models
│   ├── execution/      # Order execution and management
│   ├── optimize/       # Parameter optimization
│   ├── visualize/      # Charting and visualization
│   ├── pipeline/       # Workflow orchestration
│   ├── compliance/     # Audit and compliance
│   ├── cli/            # Command-line tools
│   └── utils/          # Shared utilities
├── examples/           # Example scripts and tutorials
├── docs/               # Comprehensive documentation
└── tests/              # Integration tests

Design Principles

Modularity: Each package is independently usable
Type Safety: Full TypeScript support with strict typing
Extensibility: Plugin architecture for custom components
Performance: Optimized for large-scale data processing
Testing: Comprehensive test coverage
Documentation: Extensive API documentation and examples

Usage Examples

Basic Backtest

import { YahooAdapter } from '@meridianalgo/data';
import { trendFollowing } from '@meridianalgo/strategies';
import { TimeBasedEngine } from '@meridianalgo/backtest';

async function runBacktest() {
  // Fetch historical data
  const yahoo = new YahooAdapter();
  const bars = await yahoo.ohlcv('AAPL', {
    start: '2023-01-01',
    end: '2023-12-31',
    interval: '1d'
  });

  // Create a trend-following strategy
  const strategy = trendFollowing({
    fastPeriod: 10,
    slowPeriod: 30,
    maType: 'ema'
  });

  // Run backtest
  const engine = new Time...

Releases: MeridianAlgo/Javascript-Packages

v2.5.9

Uh oh!

v2.5.3

Uh oh!

v2.5.2

Uh oh!

v2.5.0 - Beta

Uh oh!

v2.0.0-Public

MeridianAlgo.js

Overview

Table of Contents

Quick Start

Installation

For End Users

For Contributors

Key Features

Technical Indicators (100+)

Data Management

Trading Strategies

Backtesting Engine

Risk Management

Portfolio Optimization

Machine Learning

Execution

Developer Tools

Architecture

Design Principles

Usage Examples

Basic Backtest

Uh oh!