Demo to Live Transition: The Complete Checklist for Expert Advisors
The demo to live transition is the moment an expert advisor stops being a theory and starts being an expense. Most traders treat it as a single event — switch accounts, switch terminals, done — when it is really a process with four stages: understanding why demo results will differ, serving a minimum demo period, stepping the account size down and comparing the first 30 live days against a baseline. This post works through each stage as a checklist.
Why demo to live transition results differ: execution, spreads and slippage
A demo account is a simulation of a market, not the market. Every difference between the two curves starts there. The first divergence is execution: demo orders fill at the quoted price in an instant, while live orders travel through the broker’s pipeline and meet whatever liquidity is available at that millisecond. In fast markets you get requotes, partial fills or a worse price. The second is spreads: demo platforms often run fixed or artificially tight spreads, while live spreads are variable and widen sharply around news releases and at session opens. The third is slippage: demo fills rarely slip, and live fills slip by definition when the market moves between your order and its execution — the question is only how much.
Add commissions and swaps, which some demo configurations waive or miscalculate, and the arithmetic of a strategy quietly changes: a strategy that earned two points per trade on demo can lose two on live purely through costs. The forward testing checklist shows how to measure these costs before deployment.
Demo to live transition: the minimum demo period
How long is long enough? Eight to twelve weeks, or a hundred completed round-trip trades, whichever comes later — and only if the period actually contained the market conditions your strategy needs. A trend-following EA tested through a three-month range has served no useful demo period at all. The period must include trending and ranging phases, at least one major news event and, ideally, a gap or a thin-liquidity session, because those are the moments execution quality is revealed.
Two further rules keep the period meaningful. First, run the demo on the same terminal, the same settings and the same broker you intend to use live — a demo that differs from the live setup validates nothing. Second, log the demo period like a live one: equity, drawdown, slippage and every settings change — the start here guides cover the routine. A demo curve without a log is a screenshot, not evidence.
Demo to live transition: account size step-down and risk scaling
The most common transition error is moving at full size. The fix is a step-down: begin live trading at ten to twenty-five per cent of your intended account size and scale up only against defined triggers — 10 per cent to start, 25 per cent after thirty live trades with slippage inside budget, full size after ninety trades with drawdown inside the tested maximum. Scaling becomes mechanical and undebatable instead of emotional.
Step-down also matters because position sizing behaves differently at small balances: the same EA can refuse orders or hit minimum-lot limits, producing trade frequency that differs from the demo regardless of the market. Starting small surfaces those mechanical problems cheaply. The ESMA risk analysis and the FCA’s market guidance set out the regulatory context for the trading costs that make this sizing discipline necessary.
Demo to live transition: VPS readiness and broker differences
Live trading makes hosting a requirement, not an optimisation. A home computer that sleeps or loses connection overnight turns a tested strategy into an unmonitored one, so the VPS trading guide becomes part of the transition: the VPS must run the EA for at least two full weeks of demo trading before the live switch, proving the infrastructure before the money is committed.
Broker differences deserve the same suspicion: execution model (market maker versus straight-through processing), spread structure, stop-out levels, margin requirements and symbol naming all change results, and all differ between brokers — sometimes between accounts at the same broker. Trading demo at one broker and live at another invalidates the transition. Choose the live broker first, then serve the demo period there.
Demo to live transition: what to compare in the first 30 live days
The first 30 live days are the only period in which demo and live are genuinely comparable, because conditions and account both change afterwards. Compare five things, each against a number you decided in advance: trade frequency against the backtest, average slippage and spread paid against the forward-testing assumptions, the shape of the equity curve, maximum drawdown against the tested maximum, and the error log for requotes, disabled trades and failed orders.
Decide the tolerance before day one and act only when it is breached. Slippage averaging double the assumption is a real execution problem; an equity curve that merely looks different while every cost number is inside tolerance is noise — the worst error in the first 30 days is re-optimising a system that is behaving as designed. The CFTC’s investor education material covers the wider retail trading risks before you commit capital.
Frequently asked questions about the demo to live transition
How long should I test an expert advisor on a demo account before going live?
A sensible minimum is eight to twelve weeks or at least a hundred completed round-trip trades, whichever comes later, provided the period covers the market conditions the strategy needs — both trending and ranging phases, plus at least one major news event. A strategy that has not survived those conditions on demo has no business being judged on live money.
Why is my expert advisor performing differently on live than on demo?
The three usual reasons are execution, spreads and slippage. Demo accounts fill at the quoted price instantly, often at fixed or reduced spreads, and typically incur no slippage, while live accounts are subject to variable spreads, latency, requotes and slippage during fast markets. Costs that are invisible on demo become real on live, which is exactly why the transition checklist treats the live curve as a new baseline rather than a continuation.
Should I move to a live account at full size?
No. Step the account size down instead: start at ten to twenty-five per cent of your intended allocation, trade a defined number of live trades, then scale up only when realised slippage and drawdown are inside the budget you set in testing. Full size on day one converts a normal learning curve into a forced stop-out.
What should I compare between demo and live in the first 30 days?
Five numbers: trade frequency versus the backtest, average slippage and spread paid versus the forward-testing assumptions, the shape of the equity curve, maximum drawdown versus the tested maximum, and the error log. Each has a tolerance decided in advance, and you only act when a number breaches its tolerance.
Transition on evidence, not on nerves. Follow the automation playbook for the deployment sequence, VPS setup and monitoring routine that turn the checklist above into a repeatable process for every EA you run.
Risk disclosure: Trading foreign exchange, commodities, CFDs and cryptocurrencies carries a high level of risk and may not be suitable for all investors. Past performance is not indicative of future results. AlgoTM provides trading tools and technology only and does not provide investment advice, portfolio management or guaranteed returns.
