/alpaca-js-backtesting

Backtest trading algorithms built for Alpaca with JavaScript and Node.js

Primary LanguageJavaScriptMIT LicenseMIT

@kendelchopp/alpaca-js-backtesting

Disclaimer

This package is in no way affiliated with Alpaca

Getting Started

Installing the package

$ npm install @kendelchopp/alpaca-js-backtesting

Writing the code

  1. Initialize your Alpaca object with the keys
const Alpaca = require('@alpacahq/alpaca-trade-api');

const alpaca = new Alpaca({
  keyId: process.env.API_KEY,
  secretKey: process.env.SECRET_API_KEY,
  paper: true,
  usePolygon: false
});
  1. Initialize your Backtest object

Note: When passing in your alpaca object, you are giving this package the ability to use your keys to make API requests. This package will make API requests for the market data so that it can run the simulation. Be careful when sending out your alpaca object.

const Backtest = require('@kendelchopp/alpaca-js-backtesting');

const backtest = new Backtest({
  alpaca,
  startDate: new Date(2020, 0, 1),
  endDate: new Date(2020, 1, 1)
});
  1. Run your algorithm replacing Alpaca/websockets accordingly.

For example (commented out lines are the old lines to be replaced):

// const client = alpaca.data_ws;
const client = backtest.data_ws;

// alpaca.createOrder({...})
backtest.createOrder({...})
  1. Add a disconnect handler to see how your algorithm performed
client.onDisconnect(() => {
  console.log(backtest.getStats());
});

Alpaca Compatibility

This table lists the version of @kendelchopp/alpaca-js-backtesting with the version of alpaca-trade-api-js it was tested with. Using other versions may require some syntax changes.

@kendelchopp/alpaca-js-backtesting @alpacahq/alpaca-trade-api-js
1.0.0 1.4.1
1.1.0 1.4.1
1.1.1 1.4.1
1.1.2 1.4.1
1.1.3 1.4.1

Relevant Functions

Backtest#createOrder

Creates an order using the same parameters as Alpaca#createOrder

Example

backtest.createOrder({
  symbol: 'SPY',
  qty: 300,
  side: 'sell',
  type: 'market',
  time_in_force: 'day'
})

Backtest#getStats

Gets the stats for the portfolio including starting and ending value and ROI

Example

console.log(backtest.getStats());

Websocket#connect

Simulates connecting to the API, loads the data, and runs the simulation

Example

client.connect();

Websocket#onConnect

Runs a function when establishing the connection. When backtesting, you will usually just use this to subscribe to channels

Example

client.onConnect(() => {
  client.subscribe(['alpacadatav1/AM.SPY']);
});

Websocket#onDisconnect

Runs a function when the simulation is completed. You can print out the stats here.

Example

client.onDisconnect(() => {
  console.log(backtest.getStats());
});

Websocket#onStockAggMin

This runs your function when a simulated minute occurrs and returns the relevant data. This is where your trading will likely take place. Subject will look something like AM.SPY and data will look like a standard bar data.

Example

client.onStockAggMin((subject, data) => {
  if (data.closePrice < 500) {
    backtest.createOrder({
      symbol: 'SPY',
      qty: 300,
      side: 'buy',
      type: 'market',
      time_in_force: 'day'
    });

    console.log('Bought SPY');
  }
});

Websocket#subscribe

Subscribes to the channels listed. Currently, only aggregate minute channels are supported.

Example

client.onConnect(() => {
  client.subscribe(['alpacadatav1/AM.SPY', 'alpacadatav1/AM.AAPL']);
});