Kronos

Fast and versatile backtest framework to test your quantitative strategies. Here is a list of some of the features:

• Implement your strategy with any language that compiles to WebAssembly (Rust, Python, Typescript, C/C++, ...)
• Data source agnostic, can be used with a variety of assets (stocks, crypto, ...) by providing OHLCV data how you wish
• Fast (less than 10 seconds to backtest 1 year of data with a tick of 1 second which is around 31 million data points)
• Easy to use, send a simple POST request to start your backtest
• Place market orders, limit orders and stop orders
• Able to simulate down to a precision of 1 nanosecond for HFT strategies
• Detailed metrics about your strategy performance with per trade analysis
  • Strategy performance: ROI, Sharpe ratio, max drawdown, net profit
  • Buy & hold benchmark: Compare your strategy against a simple buy-and-hold approach
  • Trade statistics: Win rate, profit factor, average win/loss, trade duration
  • Execution metrics: Number of orders placed/executed, total fees and slippage
• Configurable slippage for a more realistic result

Important

You should be careful about stock split in your data. If it isn't ajusted, it might falsify the result of the simulation.

Usage

Here is an example of an execution of a strategy with a simple POST request

curl -i http://localhost:3000/run -H "Content-Type: application/json" -d '{
    "parameters": {
      "start_date": "2024-02-17 00:00:00",
      "end_date": "2025-02-17 00:00:00",
      "tick": "1m"
    },
    "data": {
      "source": "..."
    },
    "broker": {
      "cash": 10000.0,
      "fees": { "Flat": 1.0 },
      "slippage": {
        "min": 0.01,
        "max": 0.05
      }
    },
    "strategy": {
      "wasm": "..."
    }
  }'

Ideas and TODO

• Visualize your strategy using a dedicated frontend
• Calculate taxes impact on your strategy performance
• Support for derivatives (Futures, Options, Warrants, ...)
• Support for Stop-Limit orders
• Support multiple instruments in one simulation for complex strategies
• Support L2 data for a better overview of the order book with spread visualization etc
• Support other data types like FIX
• Support for better splippage predictions using liquidity and volatility
• Support paper trading with live market data
• Support for stress testing during significant market events
• Replay system with per trade visual and total history
• Better time management with timezones etc