Demo vs live divergence: what to log in your first 30 days

Your backtest might look great, but the gap between paper trading and a live account is often where strategies go to die. The main culprits are usually execution-related. In a demo environment, you get filled at the mid-price or best available quote, which ignores the reality of order book depth. Live, you deal with slippage, partial fills, and latency that eats your edge alive.

During your first 30 days of live trading, don't just log the PnL. You need to capture the difference between your expected execution price and the actual fill. If you are using platforms like QuantConnect or custom Python wrappers, ensure your logging captures these specific data points:

• Expected entry price vs. actual fill price
• Latency measured in milliseconds from signal trigger to receipt of fill
• Bid-ask spread at the exact moment of execution
• Frequency of partial fills or rejected orders

Most traders find that their slippage is non-linear, meaning it gets worse during high volatility or low liquidity. If your slippage is consistently higher than your projected edge, you either need to tighten your entry criteria or move to a more liquid instrument. You can use a simple tracking sheet to compare your theoretical slippage against reality. If the difference is consistently wider than 1-2 ticks, your model is likely over-fitting to clean data.

Are you tracking the difference between your signal price and your fill price, or are you just looking at the account balance at the end of the day?