Paper trading means recording trades without putting real money behind them — logging the entry, the stop, the target and the outcome exactly as if the position were live. We use it to score our own signals in public, forward in time, where we cannot quietly revise the result afterwards.
A paper trade is a real decision with imaginary money. You write down what you would buy, at what price, where you would place the stop, where you would take profit — and then you leave it alone and let the market decide the outcome, recording it honestly whichever way it goes.
The word comes from the era when you would literally track it on paper. The essential feature has not changed: the decision must be committed BEFORE the outcome is known. A "paper trade" reconstructed after the fact is not a paper trade. It is a story.
A backtest is a claim about the past made by someone who already knows how the past turned out. However carefully it is run, the person running it could have tried a hundred variants and shown you the best — which is what overfitting looks like from the outside, and you generally cannot detect it from a backtest alone.
A forward paper-trading log fixes precisely that hole. Each signal is published with its entry, stop and targets before anyone knows the answer, and the result is recorded when it closes — winners and losers alike. We cannot retroactively add the trades that worked or quietly drop the ones that did not. It is the difference between a claim and a commitment.
That is why our track record shows losing trades, why it shows the equity curve with its drawdowns coloured red rather than smoothed away, and why it prints "the edge is statistically unproven" while the sample is still small. A track record that only showed the wins would be worth exactly nothing — it would just be the backtest problem in a new costume.
For a strategy: it is a clean, honest scoreboard. Did the signals, followed mechanically, produce a positive result per unit of risk?
For you personally: it is the cheapest possible way to find out whether you understand a process before it costs anything. Placing a stop, sizing a position, sitting through a drawdown without abandoning the plan, and reading your own results in R-multiples instead of in feelings — all of that is learnable on paper, and learning it with real money is an expensive way to discover you had not.
A practical discipline that makes paper trading actually useful: write the invalidation level down before you enter. "I am wrong if it trades below X" committed in advance is a test. The same sentence written afterwards is a rationalisation, and you will not be able to tell the difference in your own memory.
It cannot tell you how you will behave. This is the honest, uncomfortable limitation and it is worth being blunt about. Watching an imaginary position go 15% against you is a mild irritation. Watching real savings do the same thing is a physical experience, and it makes people abandon perfectly good plans at exactly the wrong moment. Nobody ever panic-sold a paper position at the bottom.
It also flatters execution. Real fills are not always at the price you wanted; spreads and slippage take a bite, especially in fast markets and thinly-traded names, and a real order can miss an entry entirely while a paper one assumes it filled. Commissions, taxes and the cost of being unable to trade at all during a halt are likewise invisible on paper.
The consequence: read a paper record as evidence about the SIGNALS, and discount it as evidence about the LIVE RESULT. Both readings are honest. Only one of them is being claimed.
The commonest way to fool yourself is a small sample. A handful of winning paper trades feels like proof and is statistically nothing — which is exactly why we show a t-statistic and an expectancy figure alongside the log rather than a triumphant win count. Read those two before you read the wins.
The second is quiet rule-drift: nudging a stop "just this once" because the trade would have worked, or not logging the one you would obviously have skipped. Each individual adjustment is defensible. Together they convert an honest experiment into a flattering one, and you will not notice it happening. If a paper log is not mechanical, it is not evidence.
The third is treating paper success as a licence to size up dramatically when you go live. The strategy's expectancy does not change when real money is involved — but your ability to follow it does, and that is the variable you have not yet tested.
Recording trades — entry, stop, target, outcome — without committing real money, so a strategy can be tested honestly and forward in time.
Because a backtest is made by someone who already knows how the story ended, and can be tuned until it looks good. A forward paper-trading log commits each signal before the outcome is known and records losers as well as winners, so it cannot be quietly revised after the fact.
No, and the gap is mostly psychological. Paper trading removes the fear of real loss, which is the thing that actually causes people to abandon good plans. It also ignores slippage, spreads, missed fills and commissions. Treat it as strong evidence about the signals and weak evidence about how you will behave.
More than feels necessary. A dozen trades tells you almost nothing. Rather than counting wins, look at expectancy (average result per unit of risk) and the t-statistic, which is the check on whether the result is distinguishable from luck.
Browse all S&P 500 tickers to see this metric applied to individual companies.
Educational research only — not investment advice.