Build an Alpaca paper trading bot in Python (alpaca-py, step by step)
Alpaca gives every account a free paper (simulated) trading environment with an API. That makes it a good place to learn bot-building without risking money. This guide gets a small bot running in Python.
1. Get paper API keys
Sign up at Alpaca, switch the dashboard to Paper, and generate an API key pair. Store the keys as environment variables, never in your code:
export APCA_API_KEY_ID="your-paper-key"
export APCA_API_SECRET_KEY="your-paper-secret"
2. Install the SDK
python -m venv .venv && source .venv/bin/activate
pip install alpaca-py pandas
3. Connect and check the account
import os
from alpaca.trading.client import TradingClient
trading = TradingClient(os.environ["APCA_API_KEY_ID"],
os.environ["APCA_API_SECRET_KEY"], paper=True)
acct = trading.get_account()
print(acct.status, acct.buying_power)
paper=True points at the paper endpoint. Keep it that way while you're learning.
4. Read daily bars
from datetime import datetime, timedelta, timezone
from alpaca.data.historical import StockHistoricalDataClient
from alpaca.data.requests import StockBarsRequest
from alpaca.data.timeframe import TimeFrame
from alpaca.data.enums import DataFeed
data = StockHistoricalDataClient(os.environ["APCA_API_KEY_ID"],
os.environ["APCA_API_SECRET_KEY"])
req = StockBarsRequest(symbol_or_symbols="SPY", timeframe=TimeFrame.Day,
start=datetime.now(timezone.utc) - timedelta(days=120),
feed=DataFeed.IEX) # free data feed
bars = data.get_stock_bars(req).df.loc["SPY"]
5. Compute a simple signal
fast = bars["close"].rolling(10).mean()
slow = bars["close"].rolling(30).mean()
want_long = fast.iloc[-1] > slow.iloc[-1]
This moving-average cross is a teaching example, not a recommendation.
6. Place a paper order, with a kill switch first
from pathlib import Path
from alpaca.trading.requests import MarketOrderRequest
from alpaca.trading.enums import OrderSide, TimeInForce
if Path("STOP").exists():
raise SystemExit("STOP file present: not trading")
positions = {p.symbol for p in trading.get_all_positions()}
if want_long and "SPY" not in positions:
trading.submit_order(MarketOrderRequest(symbol="SPY", notional=1000,
side=OrderSide.BUY, time_in_force=TimeInForce.DAY))
elif not want_long and "SPY" in positions:
trading.close_position("SPY")
7. Run it on a schedule
Run the script once a day after the open with cron (Linux/macOS) or Task Scheduler (Windows). Log every decision to a file so you can review it later.
What to add before you trust it
- A dry-run flag that prints orders instead of sending them
- A check that the market is open
- Optional caps: max position size, max open positions, max daily loss
- A backtest on historical bars, and tests with a mocked API
- A read-only "evidence" script that snapshots account, positions, orders and fills
Skip the boilerplate
The Alpaca Trading Bot Starter Kit ($24) has all of the above already wired: config from env vars, dry-run, a STOP-file kill switch, optional risk caps, a backtester and 29 tests against a mocked API. It runs on paper by default; live mode is opt-in and needs an explicit confirmation flag. The Trader Bundle ($39) adds the read-only Safety Kit and a Paper Trading Journal template.
Educational software, not financial advice. Trading involves risk of loss, and backtests or paper results don't predict live results. Not affiliated with Alpaca.
