Home / Libraries / ML4T Live / Docs
ML4T Live
ML4T Live Documentation
Production trading with broker integrations
Source on GitHub
Skip to content

Quickstart

The recommended first run is shadow mode: your strategy executes through the live engine and risk checks, but no real orders are sent to the broker.

First Strategy

import asyncio

from ml4t.backtest import Strategy
from ml4t.backtest.types import OrderSide
from ml4t.live import AlpacaBroker, AlpacaDataFeed, LiveEngine, LiveRiskConfig, SafeBroker


class BuyOnceStrategy(Strategy):
    def on_data(self, timestamp, data, context, broker):
        bar = data.get("SPY")
        if bar is None:
            return

        if broker.get_position("SPY") is None:
            broker.submit_order("SPY", 10, side=OrderSide.BUY)


async def main():
    broker = AlpacaBroker(api_key="...", secret_key="...", paper=True)
    feed = AlpacaDataFeed(
        api_key="...",
        secret_key="...",
        symbols=["SPY"],
        data_type="bars",
    )
    safe_broker = SafeBroker(
        broker,
        LiveRiskConfig(
            shadow_mode=True,
            max_position_value=25_000,
            max_order_value=5_000,
        ),
    )

    engine = LiveEngine(BuyOnceStrategy(), safe_broker, feed)
    await engine.connect()

    try:
        await engine.run()
    finally:
        await engine.stop()


asyncio.run(main())

Why This Works

  • Your strategy stays synchronous, just like in ml4t-backtest
  • LiveEngine runs broker/feed I/O asynchronously
  • ThreadSafeBrokerWrapper lets strategy code call broker.submit_order(...) safely
  • SafeBroker enforces limits before any live order can be placed

First-Run Checklist

Before you move past this example, confirm that:

  1. the strategy receives bars from the feed you expect
  2. orders appear in shadow mode instead of hitting the broker
  3. positions and cash update through the virtual portfolio
  4. you can stop and restart the engine cleanly

Deployment Progression

  1. Shadow mode: shadow_mode=True
  2. Paper trading: shadow_mode=False with paper broker credentials
  3. Small live size: conservative limits and low notional exposure
  4. Gradual scale-up only after observing stable behavior

Common Variations

Interactive Brokers Feed

IBDataFeed needs a connected IB session object:

broker = IBBroker(port=7497)
await broker.connect()

feed = IBDataFeed(broker.ib, symbols=["SPY", "QQQ"])

Aggregate Ticks Into Bars

raw_feed = IBDataFeed(broker.ib, symbols=["SPY"])
feed = BarAggregator(raw_feed, bar_size_minutes=1, flush_timeout_seconds=2.0)

Next Steps