Paper trading means practising trades with pretend money: you record what you would have bought or sold, and at what price, without placing a real order. It is one of the friendliest ways to learn, because mistakes cost nothing. But a simulation is a simplified copy of a market, and the parts it simplifies are exactly the parts that tend to surprise people later.
This article sorts paper trading into two honest piles: what it teaches well, and what it quietly leaves out.
What paper trading is, and is not
In paper trading you follow a set of rules, log each decision, and track an imaginary balance. Some people use a spreadsheet, some use a notebook, and many tools offer a simulated account.
What it is not is a preview of real results. U.S. commodity regulations require a specific warning whenever simulated or hypothetical trading results are advertised. Part of that required text says the trades "have not actually been executed," so the results may have under- or over-compensated for market factors "such as lack of liquidity," and that simulated programs are "designed with the benefit of hindsight." That is a useful sentence to keep in mind for any paper result, including your own.
The two columns at a glance
| What paper trading teaches well | What paper trading leaves out |
|---|---|
| How a rule behaves across many situations | How it feels to lose real money |
| The habit of writing down a plan before acting | Whether your order would really have been filled |
| Position sizing: how much to put in one idea | The exact price you would have received |
| How often a rule trades, and why that matters | Costs that are easy to forget: spreads, fees, slippage |
| Keeping a record you can review honestly | Liquidity: whether anyone was there to trade with you |
| Patience: waiting for the conditions you defined | The pressure to change the rules mid-way |
The rest of this guide takes each column in turn.
What it teaches well
Rules, not hunches
The biggest gift of paper trading is that it forces you to write rules down. "Buy when the price is above its 20-day average" is testable. "Buy when it feels strong" is not. Once rules are written, you can see how often they fire and whether they behave the way you expected.
Sizing and record keeping
Paper trading is a safe place to practise position sizing, the decision of how much to put into one trade. You can see how a rule that risks 2% per trade feels very different from one that risks 20%, even when the signals are identical. Keeping a simple log (date, rule, entry, exit, reason) also builds a habit that is valuable whatever you do next.
How often a rule trades
Many beginners are surprised by how often a simple rule switches on and off. Paper trading makes that visible, and it leads naturally to the most important missing piece: costs.
What it leaves out
Emotions
Clicking "sell" on an imaginary loss is easy. Doing the same with real savings is not. People who follow rules perfectly on paper often break them under real pressure, holding losers too long or exiting winners too early. A simulation cannot recreate that feeling, so it cannot tell you how you will behave.
Fills: whether the order really happens
A fill is the actual completion of an order. In a simulation, every order fills at the price you chose. In real markets, that is not certain. The SEC's investor education site explains that a market order is meant to be executed immediately but does not guarantee the price, and that the last traded price is not necessarily the price you will get. A limit order guarantees a price or better, but it may never execute at all. Paper trading usually assumes the best of both worlds.
Costs: the spread, fees and slippage
Every real trade has costs that paper trades tend to skip:
- The spread: the gap between the bid (the highest price a buyer will pay) and the ask (the lowest price a seller will accept), as Investor.gov defines them. Buying at the ask and selling at the bid means you start each trade slightly behind.
- Fees: commissions, platform charges or subscriptions.
- Slippage: the difference between the price you expected and the price you actually got.
Fees compound over time. An SEC investor bulletin shows that a hypothetical 100,000 investment growing 4% a year for 20 years ends at about 208,000 with a 0.25% annual fee, but about 179,000 with a 1.00% fee. For an active strategy, per-trade costs work the same way, only faster.
Liquidity: is anyone on the other side?
Liquidity describes how easily something can be bought or sold. Investor.gov describes liquid investments as ones that can be sold readily without paying a hefty fee. A paper account can "buy" any amount instantly. A real market might not have enough buyers or sellers at your price, especially for small or thinly traded assets, which can move the price against you while you trade.
A worked example: when costs eat a small edge
All numbers in this example are illustrative, not market data.
Suppose a paper log shows 20 trades, each using 1,000, with an average gain of 0.30% per trade before costs. On paper:
- Gross result: 20 × 1,000 × 0.30% = 60
Now add plausible round-trip costs (buying and selling once each):
| Cost item | Per round trip | Over 20 trades |
|---|---|---|
| Spread | 0.10% | 20 |
| Slippage | 0.08% | 16 |
| Fees | 0.05% | 10 |
| Total | 0.23% | 46 |
The net result is 60 − 46 = 14, less than a quarter of what the paper log suggested. If the rule traded twice as often for the same total gross gain, costs alone could erase the result entirely. If the true costs were a little higher than estimated, the same "winning" log becomes a losing one.
This is exactly what the "Small edges and costs" lesson in Aerolith's interactive lab is built to show: a cost slider on synthetic data, and the moment where a small positive edge turns negative.
How to make paper practice more honest
You cannot add real emotions to a simulation, but you can close some of the other gaps:
- Always subtract costs. Add a fixed cost per trade to every entry in your log, and err on the high side.
- Assume a slightly worse price. Record buys a little above and sells a little below the price you saw.
- Write the rules before you start, and do not change them mid-test. If you must, start a new log.
- Count the trades. A result based on 8 trades says much less than one based on 80.
- Note your reactions. Even on paper, jot down when you felt the urge to break a rule. It is a small preview of real pressure.
Where paper trading fits in a learning path
Paper trading is best seen as a classroom, not a rehearsal. It is excellent for understanding how rules, sizing and costs interact, and weak at predicting real outcomes. If you are still getting comfortable with the parts of a trading system, start with our guide to how an AI trading bot works, then come back to paper practice with a clearer idea of what each rule is meant to do.