Profitable on Demo but Not Live: The Step a Demo Account Never Tests

By Sofia Harchich | Trading Psychologist & Behavioural Finance Writer

Being profitable on demo but not live is usually read as proof of something unflattering: that the strategy only works when nothing is at stake, or that the trader is the weak point in an otherwise sound method. The first reading is occasionally true. The second is almost never an accurate description.

A demo account is a faithful test of the analysis, the entry criteria and the execution mechanics. What it cannot test is the state a decision is made from, because that state depends on the money being real. The gap between the two accounts is a measurement of that missing step, and like most gaps in a trading record, it can be measured.

What a demo account actually tests

A demo session exercises almost everything a strategy needs: reading structure, waiting for criteria, placing orders, sizing by rule, managing an open position. Traders who are profitable on demo but not live have usually proved that this part works. The equity curve on the demo is evidence about the method, and it is reasonable evidence.

What it does not exercise is the body. A loss on a simulated account is information. A loss on a live account is information plus a physiological event — a change in heart rate, attention and urgency that arrives before any thought about the trade has formed. The demo runs the decision without that event, which makes it a test of a slightly different decision.

The evidence that real money changes the decision

The cleanest research on this comes from experimental economics rather than from trading. In Holt and Laury’s study of risk aversion and incentive effects, participants chose between safer and riskier lotteries. When the payoffs were scaled up and paid in real money, choices became markedly more cautious. When the same scaled-up payoffs were only hypothetical, choices barely moved.

Three limits apply. These were laboratory lotteries, not markets, and the participants were mostly students choosing from a short menu of lotteries rather than traders deciding hundreds of times. What transfers is the narrower finding: people do not discover how they will decide with real stakes by deciding with imaginary ones. A demo record is a hypothetical-payoff record, however seriously it was kept.

Where being profitable on demo but not live shows up first

The gap is rarely spread evenly across a record. It tends to concentrate in a small number of decisions, and they are worth checking in a particular order.

Exits are often the first to change. On demo, the stop is left alone and the target is allowed to fill. Live, the winner is closed at the first pullback and the loser is given room it was never meant to have. Size is often next, usually downward on the best setups and upward after a loss. Entries tend to change last — late, or skipped entirely, on setups the demo record would have taken without a second look.

That order is useful in itself. A trader who is profitable on demo but not live and whose entries are identical in both accounts has a management problem rather than an analysis problem, and management is where state does most of its work.

The demo account runs the whole of the instability loop except its second step. The trigger still arrives and a behaviour still has to be chosen, but without real money the activation never fires — so the demo record shows how the plan performs with the loop switched off.

The four ways the live account breaks the plan

How the gap presents depends on which surface pattern runs under pressure. The structure underneath is the same; the visible failure differs.

In the reactive pattern, the live account adds speed: the re-entry arrives within minutes of a loss that would have been logged and left alone on demo. In the controlled pattern it adds checking — one more confirmation, one more timeframe — until the entry is late or gone. In the avoidant pattern the platform simply stops being opened on difficult days, something no demo record ever shows. In the impulsive pattern the size grows after a stretch of restraint, on a real setup, at a level the plan never allowed.

One of them is often recognisable at once, and that one is the best guide to where to look in the next twenty live sessions.

Why going back to demo does not close the gap

The usual advice is to return to demo until the results are consistent again. For a trader who is already profitable on demo but not live, this repeats the test that was passed and avoids the one that was failed. More demo sessions train the analysis further and leave the state exactly where it was.

Moving straight to full size has the opposite problem. It exposes the state to the strongest version of the trigger before there is any way of noticing it, which is how a promising method gets abandoned after one bad week.

What closes the gap is graded exposure: live money at a size small enough that the activation is present but survivable, held there long enough for the response to become familiar, then increased in steps.

Building the bridge from demo to live

Start live at the smallest size the platform allows, or at a size where a full stop-out is unremarkable. The purpose of this size is not profit. It is to make the second step of the decision real while keeping it small enough to watch.

Add one column the demo never needed: a single word for your state when the position was opened and another when it was closed. Settled, activated or flat is enough. The point is to give the live account a measurement the demo account could not produce.

Increase size only when two numbers hold for twenty live sessions — the proportion of trades that met every written criterion, and the proportion of exits taken at the planned level. When both sit within a few percentage points of your demo record, the state has stopped changing the decision at that size. Then the next step up begins, and the same twenty sessions run again.

The verdict a live account seems to deliver

There is an understandable reason so many traders are profitable on demo but not live. A trader who has worked on a method for months has more than money riding on its first live results; there is a verdict on the months themselves. Closing a winner early protects that verdict for one more day. Skipping the setup means the method cannot fail today.

The private conclusion that tends to form is that the problem is personal — that the method works and the person running it does not. The record usually says something more precise and far less damning: that one step of the decision has never been trained, because until the account was live there was nothing to train it on.

Twenty live sessions, one extra column

Before changing a strategy that is profitable on demo but not live, run it live at minimum size for twenty sessions with the state column added. Then compare three numbers against your demo record: criteria met, exits at plan, and size relative to plan.

If entries match and exits do not, the work is in managing a position under activation. If size moves after losses, the work is in the hour after a loss. If entries are skipped, the work starts before the session opens. Each of these has a different fix, and none of them is more time on demo.

For the exit side of this gap, read exiting trades too early; for the state that turns a decision before the click, ventral vagal trading; for the record the bridge depends on, trading journal psychology.

The pattern quiz identifies which behaviour is running your trading under pressure, and the free sample it sends maps that result onto the four surface patterns — reactive, controlled, avoidant and impulsive — the four ways a live account most often parts company with the demo.

---

About the Author

Sofia Harchich is a Trading Psychologist and Behavioural Finance Writer with a Master’s in Psychology. She works at the intersection of Jungian shadow work, neuroscience, and market behaviour — helping traders understand the psychology driving their decisions, not just the strategy.

Read more at thewealthmirror.com/about