Paper Trading & Forward Testing

A backtest interrogates history; forward testing runs your rules on the live market going ahead — the only data that's guaranteed lookahead-free, survivorship-free, and unmined, because it didn't exist when you wrote the rules. Its simplest form is paper trading: executing the strategy with simulated money, logging every signal and fill as if real. This is the honest dress rehearsal — and it has its own characteristic lies, pointing the opposite way from a backtest's.

What paper trading tests brilliantly: the rules' operability. A strategy that looked clean in a backtest turns out, live, to have ambiguous signals ("is this a close above, or a wick?"), impossible timing (signals firing while you're at work), instrument quirks (the option strike your rules want has no liquidity — Market Structure school's spread lesson, met in person), and workflow gaps (where does the alert fire? where does the order go? — Chapters 4–6). Two weeks of paper trading surfaces more operational truth than two months of backtesting, because the market keeps generating situations your historical test never imagined. This is the rehearsal's genuine gift: **it debugs the process before money is attached to the bugs.**

Where paper trading lies — two directions at once: (1) Perfect fills. Most simulators fill you at the touched price, full size, always — no spread crossed, no slippage, no partial fills, no thin-book impact (the entire toll system of your Market Structure school, waived). Intraday and options paper results are inflated by default; mentally charge your journal's measured slippage against every simulated fill. (2) The fearless brain. The deeper lie: paper trading tests your rules with a you that doesn't exist. No loss aversion (nothing to lose), no revenge urge, no hot-hand sizing, no 2:30 PM red-position paralysis — the entire Behavioural Finance school's cast, absent from the rehearsal, guaranteed present on opening night. This is why the classic trajectory — three flawless paper months, then live chaos — isn't a mystery: the strategy didn't change; the operator did. Paper trading validates the system; it cannot validate you inside the system.

Bridging the gap — the graduated go-live: the professional answer to the fearless-brain lie is dosage, not courage: (1) Paper until operationally clean — signals unambiguous, workflow smooth, a pre-decided minimum sample logged (e.g., 20–30 signals, not "a few good weeks" — small samples flatter, and you know from the Behavioural Finance school who does the flattering); (2) Go live at token size — the smallest real position your instruments allow. The purpose isn't P&L; it's introducing the emotions under anesthesia: real money's feelings arrive, but at a dose your discipline can practice against. Watch the journal's discipline scores, not the returns; (3) Scale on behavioral evidence, by rule — size steps up only after N trades at current size with plan-adherence above your threshold (a stranger-enforceable rule, naturally); any adherence collapse steps size back down. You're titrating exposure to your own psychology — the one variable no simulator ships with; (4) Keep a paper lane forever — new strategy variants earn their live slot through the same pipeline, always. The rehearsal stage never closes; it just stops being the only stage.

One reframe to keep the whole module honest: backtest, paper, and token-size live aren't competing methods — they're three filters in one pipeline, each catching what the previous one structurally cannot: the backtest catches statistical fantasy, paper catches operational fantasy, token-live catches psychological fantasy. A strategy deserves your full size only after passing all three — and most won't, which is the pipeline working, not failing.