Free prompt library

You are a careful quant developer. Build me a backtestable trading bot for SPY using a 20-day / 50-day simple moving-average crossover: go fully long when the 20-day SMA closes above the 50-day SMA, and sell to cash when it closes below. Backtest from 2019-01-02 to today. Constraints: Python 3.11, pandas, yfinance for free daily data. Long-only, no leverage. Compute signals on the daily close and fill at the NEXT day's open (no look-ahead). Charge 5 bps slippage per fill on ETFs and 10 bps on crypto. Start with $10,000. Output a trade log CSV (entry/exit date, price, qty, P&L, exit reason) and an equity curve, and compare against buy-and-hold of the same asset. Then add an optional Alpaca PAPER trading mode (endpoint https://paper-api.alpaca.markets only) that reads keys from environment variables. Never hard-code keys and never use a live trading endpoint.

You are a careful quant developer. Build a regime-switching trading bot for SPY. Features: daily log return and 20-day rolling volatility, standardized using ONLY the training window. Fit a 3-state Gaussian HMM (hmmlearn, diagonal covariance) on the previous 756 trading days and refit every 63 days (walk-forward, never train on future data). Label states by mean return: lowest = bear, middle = sideways, highest = bull. Rules: bull = hold SPY; bear = cash; sideways = RSI(2) mean reversion (buy when RSI(2) < 10, sell when RSI(2) > 70). Backtest from 2019-01-02. Constraints: Python 3.11, pandas, yfinance for free daily data. Long-only, no leverage. Compute signals on the daily close and fill at the NEXT day's open (no look-ahead). Charge 5 bps slippage per fill on ETFs and 10 bps on crypto. Start with $10,000. Output a trade log CSV (entry/exit date, price, qty, P&L, exit reason) and an equity curve, and compare against buy-and-hold of the same asset. Then add an optional Alpaca PAPER trading mode (endpoint https://paper-api.alpaca.markets only) that reads keys from environment variables. Never hard-code keys and never use a live trading endpoint.

You are a careful quant developer. Build a multi-asset daily-bar trading bot with four sleeves: (1) SPY and QQQ mean reversion: buy when the close is below the lower 20-day, 2-standard-deviation Bollinger band; exit when it closes at or above the 20-day mean. (2) BTC-USD momentum breakout: buy when the close exceeds the prior 20-day high; exit when it closes below the prior 10-day low. (3) GLD trend following: buy when close > 50-day SMA > 200-day SMA; exit when close < 50-day SMA. Position sizing: risk 1% of total equity per trade with a hard stop 1 ATR(14) below entry (cap stop at 20%), and no sleeve larger than 25% of equity. Check stops against each day's low. Run all sleeves on the US trading calendar. Backtest from 2019-01-02. Constraints: Python 3.11, pandas, yfinance for free daily data. Long-only, no leverage. Compute signals on the daily close and fill at the NEXT day's open (no look-ahead). Charge 5 bps slippage per fill on ETFs and 10 bps on crypto. Start with $10,000. Output a trade log CSV (entry/exit date, price, qty, P&L, exit reason) and an equity curve, and compare against buy-and-hold of the same asset. Then add an optional Alpaca PAPER trading mode (endpoint https://paper-api.alpaca.markets only) that reads keys from environment variables. Never hard-code keys and never use a live trading endpoint.

You are a careful quant developer. Build a crypto dip-buying ladder bot for BTC-USD and ETH-USD on daily bars. Split equity into 5 equal slots per coin (10 total). When a coin closes above its 100-day SMA AND is down 3% or more on the day, buy one empty slot for that coin at the next open. Each slot has a +5% profit-target sell (check against each day's high). No stop-loss. Report open (unrealized) slots separately from closed trades. Backtest from 2019-01-02. Constraints: Python 3.11, pandas, yfinance for free daily data. Long-only, no leverage. Compute signals on the daily close and fill at the NEXT day's open (no look-ahead). Charge 5 bps slippage per fill on ETFs and 10 bps on crypto. Start with $10,000. Output a trade log CSV (entry/exit date, price, qty, P&L, exit reason) and an equity curve, and compare against buy-and-hold of the same asset. Then add an optional Alpaca PAPER trading mode (endpoint https://paper-api.alpaca.markets only) that reads keys from environment variables. Never hard-code keys and never use a live trading endpoint.

You are a careful quant developer. Build a mean-reversion bot for BTC-USD on daily bars: buy with all equity when the close is 7% or more below its 20-day SMA; sell when the close gets back to or above the 20-day SMA. Backtest from 2019-01-02. Constraints: Python 3.11, pandas, yfinance for free daily data. Long-only, no leverage. Compute signals on the daily close and fill at the NEXT day's open (no look-ahead). Charge 5 bps slippage per fill on ETFs and 10 bps on crypto. Start with $10,000. Output a trade log CSV (entry/exit date, price, qty, P&L, exit reason) and an equity curve, and compare against buy-and-hold of the same asset. Then add an optional Alpaca PAPER trading mode (endpoint https://paper-api.alpaca.markets only) that reads keys from environment variables. Never hard-code keys and never use a live trading endpoint.