Qué cambia de verdad al pasar de la demo a la cuenta real

A demo account is a faithful copy of the platform and a partial copy of the market. The charts are real, the order tickets are real, and the prices track a real reference feed. What a demo cannot copy is what happens when your order meets other people's orders, or what happens to your own decision-making when the money at risk stops being imaginary. Both gaps can be mapped, and traders who cross from demo to live with a map tend to do it more carefully than traders who assume nothing changes.

What a demo fill actually is

Press buy on a demo and software books you a position at the displayed price. No counterparty took the other side. No liquidity was consumed. Nothing in the market moved because you acted. The fill is an accounting entry, not a transaction. That is exactly what a simulator should be, but it quietly idealises three things.

• Queue position. A real order competes with every other order arriving at the same moment. A demo order competes with nothing; it is always first in line.
• Liquidity consumption. In a live market, size interacts with the depth available at each price. On a demo, a position ten times larger fills exactly as cleanly as a small one, because no depth is being used up.
• Slippage. Demo environments fill you at the price you saw, or apply simulated slippage that is gentler and more symmetric than live conditions, especially around scheduled news, when real spreads widen and fills move away from you. Requotes are rarely modelled at all.

None of this is deception; it is the nature of simulation. But it flatters two habits in particular: trading through high-impact news releases, and using size the live market would punish. A demo track record built on either tells you less than it appears to.

What changes at live execution

A live order is a real instruction that meets real liquidity. From that moment, every cost behaves exactly as modelled in the full cost of a CFD trade: the spread is paid on entry, swap charges accrue on positions held overnight, and slippage becomes a genuine two-way variable: sometimes in your favour, and often against you at precisely the moments you most want a clean fill. Because a CFD is a leveraged instrument, all of these act on the full notional size of the position, not on the margin you posted. None of this is a surprise. All of it was visible on demo. The difference is that live, it compounds into your realised results instead of your simulated ones.

The part prospect theory predicted

The mechanical gap is the smaller one. In 1979, Daniel Kahneman and Amos Tversky published prospect theory in Econometrica, documenting how people actually decide under risk rather than how classical theory said they should. Its most durable finding is loss aversion: in their words, “losses loom larger than gains”. The displeasure of losing an amount outweighs the pleasure of winning the same amount, and the value function people act on is steeper for losses than for gains.¹

On a demo, a losing trade is information. On a live account, the same trade at the same size is a threat — and prospect theory predicts what people do about threats: cut winners early to bank the sure gain, hold losers past the plan rather than realise the sure loss, and increase size to win a loss back. Nothing about the strategy changed. The stakes changed, and decision quality moved with them.

A demo tests your understanding of the platform. A live account tests your understanding of yourself.

The population evidence is consistent with that. When ESMA agreed EU-wide restrictions on retail CFDs in March 2018, the national regulators' analyses behind the decision showed 74–89% of retail CFD accounts losing money.² The FCA's 2016 analysis of a representative sample of client accounts at CFD firms put the figure at 82%.³ Those analyses graded live outcomes, not demo records: majority losses are what live populations produced, whatever their rehearsal looked like. We will not tell you a demo record predicts your result in either direction. What a loss percentage does and does not tell you is its own subject.

Retail CFD accounts losing money, EU regulators' analyses (2018)

74–89%^†

Average losses per client across those analyses

€1,600–€29,000^†

CFD clients losing money in the FCA's 2016 sample

82%^†

Position size is the bridge variable

If loss aversion scales with the size of the potential loss, then size is the one variable that connects the two worlds. You cannot reason yourself out of the asymmetry Kahneman and Tversky documented. You can, however, size a position so that the loss it risks sits inside the range where you can still think, where a stopped-out trade is data rather than a threat demanding to be won back.

That is why the standard discipline for going live is to start smaller than your demo size (often much smaller) and let size grow with demonstrated process rather than with confidence. A stop-loss placed where the idea is invalidated, with the position sized backwards from the distance to that stop, keeps risk per trade constant while you find out how live conditions and your own reactions differ from the rehearsal. The arithmetic is set out in position sizing before prediction and why position sizing beats prediction; what a string of losses does to an account balance is covered in drawdown arithmetic.

What a demo is genuinely good for

None of this makes demo trading pointless. It makes it specific. A demo is the right tool for every part of trading that is mechanical:

• Platform fluency: where every order type lives, how to modify and close positions, and what the margin figures on the ticket mean, before any of it costs money.
• Order behaviour: the practical difference between market, limit and stop orders, and how each one triggers.
• Process rehearsal: running your full routine end to end. Pre-trade checklist, position sizing with the calculators, journalling entries and exits, weekly review.
• Strategy plumbing: whether an idea generates the signals you expected, at the frequency you expected, in the sessions you expected.

What a demo cannot certify is the one thing it structurally excludes: your behaviour when a loss is real. What it can do is make the mechanical parts automatic. That matters, because on a live account you will want your attention free for the parts that are not.

The honest way to frame the move is not as a graduation but as a change of experiment. On demo, you were testing a strategy against reference prices. Live, you are testing a smaller version of the same strategy against real liquidity and against yourself, with real costs, at a size deliberately chosen so that early mistakes are tuition rather than damage. The traders who make the transition well tend to change as little as possible about their process, and as much as necessary about their size.