Almost every new algo trader makes the same handful of mistakes. Learn them here instead of paying tuition for each.

1. Overfitting the backtest

Tuning a strategy until it looks perfect on history means you've fit the noise, not the signal — and it dies live. A too-good backtest is a warning, not a green light (overfitting, look-ahead bias).

2. Ignoring costs

Slippage, spreads, and fees turn many “profitable” strategies into losers once real fills are counted. Always test with realistic costs (slippage).

3. Oversizing

The fastest account-killer. A good edge sized too big is a blown account waiting for a normal losing streak (position sizing, risk of ruin).

Nobody blows up because their strategy was slightly wrong. They blow up because it was slightly wrong and sized far too big.

4-5. No fail-safes, and blind trust

No fail-safes: a bot with no kill switch will happily trade on a stale feed or a bug (kill switches). Blind trust: “set and forget” until a bad day — automation frees your attention, it doesn't eliminate the need for it. Start tiny, monitor, and scale only on earned trust (how to start).