What is paper trading,
and what it cannot teach you

What paper trading is

Paper trading means placing orders in a market you are watching, with money that does not exist, and tracking what would have happened. The name comes from doing it literally on paper, decades before software.

The three forms it takes

• Broker demo. A live data feed with a fake balance, on the same platform you would use for real. Best for learning the software.
• Chart replay. Historical data played back so you can practise reading it. Best for recognising structure, useless for pressure.
• Simulated market. A generated tape that behaves like the market without being it. Best for stacking reps quickly, and the quality depends entirely on how honestly it is built.

What it is not

It is not backtesting. Backtesting applies a rule to historical data mechanically and reports statistics. Paper trading involves you deciding in the moment, which means it measures a different thing: your behaviour, not the rule.

That distinction matters, because most of what goes wrong live is behaviour rather than rules. Paper trading is the only free way to look at yours.

What it trains well

Used deliberately, it does four things no amount of reading can do, and each is worth weeks of live tuition money.

Platform mechanics until they are automatic

Placing a bracket order, moving a stop, flattening in a hurry, reading the depth. These should take no thought at all before real money is involved, because in the moment you will need the thinking for something else. A hundred repetitions in demo costs nothing and fixes this permanently.

Recognising your setup

The gap between reading a description of a setup and spotting it live, in fifteen seconds, with the chart still moving, is enormous. Closing it takes volume, and paper is where you get volume cheaply.

Counting the cost of being in the market

If your demo charges spread and commission, twenty round turns a day makes the cost visible in a way no article can. If it does not charge them, you are learning a style that loses money live, which is the single most common defect in free platforms.

Finding out you overtrade

Nearly everyone takes far more trades than their own rule allows, and almost nobody believes it until they see it in a log. This is measurable on paper for free, and it is worth more than any strategy you could learn in the same time.

The four ways it lies to you

These are not opinions about mindset. They are structural differences between a demo and a live account, and each one inflates your results in a way you can predict.

One: your fills are too good

Most demos fill you at the price on screen, instantly, in any size. Live, a market order at the New York open can slip a tick or several, and a stop is a market order once triggered, so it fills wherever there is a buyer. A strategy with a six point stop and a six point target can be profitable in demo and negative live purely from this.

Two: costs are often missing or wrong

Plenty of free platforms charge no commission and a fixed spread that never widens. Real spread widens exactly when you most want to trade, around news and at the open. A scalping style that pays one tick of spread twenty times a day is paying 10 dollars a day on one micro contract before it is right about anything.

Three: there is no consequence for holding

In demo, a position 40 points offside is a number. There is no margin call, no partner, no feeling. So you hold it, it comes back, and you learn that holding works. That lesson is the expensive one, because live it works four times and then ends the account.

Four: you can restart

Reset the balance, reload the replay, take the trade again. Every restart removes the one thing that makes a decision a decision, which is that you have to live with it. This is why formats with a cut do something demos cannot.

The part no simulator can teach

There is one variable that is simply absent from every form of practice, and it is worth being blunt about it rather than pretending the right platform solves it.

Loss aversion does not activate over fake money

The documented pattern is that losses feel roughly twice as bad as equivalent gains feel good. That asymmetry is what makes people close winners early and hold losers, and it needs the money to be yours. In demo you close a 40 dollar loser calmly. Live, the same 40 dollars argues with you.

The disposition effect shows up with real skin

Studies of retail brokerage accounts find traders systematically realise gains faster than losses, which is precisely backwards for intraday work. You will not observe this in yourself on paper, because the mechanism requires real consequence.

What this means practically

It means paper trading can prove you are not ready, but it can never prove you are. A bad demo is decisive evidence. A good demo is weak evidence. The psychology guide goes into what actually changes when money is real.

The practical conclusion is not to skip paper trading. It is to stop treating demo profit as the graduation criterion, and use behaviour instead.

How to set it up so it counts

Most demo accounts are configured in a way that guarantees the results will not transfer. Five changes fix the majority of it.

Set the balance to what you will actually trade

A 100,000 dollar demo balance when you will start with 2,000 makes every position size meaningless and every loss invisible. Set it to the real number, even though the platform offers more.

Turn costs on, and overstate them

If the platform lets you set commission, set it slightly higher than your broker charges. If it lets you set spread, assume it widens. Being pleasantly surprised live is a much better failure mode than the reverse.

Trade the hours you will actually trade

Paper trading at 2pm when you will live trade the 9:30 open is practising a different market. The first thirty minutes of New York behave nothing like late morning, and that is where most intraday damage happens.

One position size, for the whole sample

Changing size mid sample makes the results uninterpretable. Fix it, run the hundred trades, then evaluate.

Log everything, before the outcome

Write the reason for entry before you know how it ends. A log written after the fact records a story, not a decision, and you will unconsciously write the version that makes you look reasonable.

What to measure instead of profit

Demo profit is the least informative number available, because it is the one most inflated by the four problems above. These five transfer.

• Rule adherence rate. What share of trades followed the written setup. Below 80 percent, nothing else you measure means anything.
• Stops moved. The count should be zero. One is a pattern, because you did it once when it was free.
• Trades per session, and whether it rises when losing. A rising count after a loss is revenge trading with a nicer name.
• Sessions ended by your own limit. Not by the close, not by boredom. This one predicts survival better than any win rate.
• Average holding time on winners versus losers. If losers are held longer, the disposition effect already has you, and real money will amplify it.

All five are behavioural, all five are free to measure, and all five keep their meaning when the money becomes real. Win rate and demo profit do not.

How long to stay on paper

Think in trades and behaviour rather than calendar time, because time on its own does not produce learning.

The number

Around a hundred trades with one setup and one size, logged. That is enough for a win rate to mean something and, more importantly, enough for a behavioural pattern to be undeniable.

The condition that matters more

Within that sample, zero stops moved and zero size increases after a loss. If either appears, the count restarts. Not as punishment, but because the sample is measuring something you are about to make worse with real money.

The trap of staying too long

There is a failure mode in the other direction: six comfortable months in demo, because demo is pleasant and live is not. At some point paper stops teaching you anything and becomes avoidance. The signal is that your last fifty trades taught you nothing new about your own behaviour.

The realistic timeline

Two hours a week gets most people to a hundred logged trades in two to three months. The full timeline guide covers what happens after that, and why the honest answer is longer than most people want to hear.

The step after paper trading

There is a gap between a clean demo and a live account, and jumping it directly is where most of the damage happens. Two intermediate steps close it.

Add consequence before you add money

Something with a clock, a cut and no restart trains the decision making part that demo cannot, while still costing nothing. Four rounds where a max loss ends your run produce more useful information about you than four weeks of relaxed demo.

Then go live at a size that feels trivial

One micro contract. Not because it will make money, but because it activates the only variable missing from everything before it. If your behaviour is identical, scale slowly. If it is not, you found the thing that was going to cost you an account, for the price of a few dollars.

The honest summary

Paper trading is a filter, not a qualification. It reliably tells you when you are not ready. It never tells you that you are. Treat it that way and it becomes the most valuable free tool in trading. Treat demo profit as proof and it becomes an expensive kind of confidence.

Frequently asked questions