How to Transition from Demo to Live Forex Trading (with an EA)

تعليمات خطوة بخطوة

1. الخطوة 1: Why a demo phase matters

   Backtest validates the EA's logic against historical data. Demo trading validates the EA's behaviour against the live data feed, broker connection, and your VPS configuration. Three things demo catches that backtest doesn't:

   1. EA-environment interaction bugs. Some EAs work perfectly in Strategy Tester but crash in live because they depend on features that work differently in Strategy Tester (e.g. Trade History query semantics). Demo surfaces these.

   2. Broker-specific symbol naming, leverage, and stop-level differences. Backtest uses tick data; demo uses your actual broker connection with all its quirks.

   3. Your operational setup. VPS uptime, monitoring, EA-attach persistence across restarts. Demo for 30 days reveals every operational weakness.

   The 30-day demo is not optional for serious money. Skipping it has saved exactly zero people from blow-up; it has cost many people their first month of live losses learning operational lessons they should have learned on demo.

2. الخطوة 2: Set up the demo correctly

   Use the same broker you'll go live with. Different brokers have different data feeds, different symbol names, different leverage, different commission structures. A demo at IC Markets is not a substitute for a demo at Pepperstone.

   Match the account type. If you'll go live on a Razor ECN account with 1:30 leverage and $7 commission per lot, get a demo with those exact settings. Many brokers' default demo is a Standard account with no commission — useful for nothing.

   Configure the demo account size to match your planned live size. If you'll go live with $5000, set the demo balance to $5000. Don't run a $100k demo when planning a $5k live — the absolute P&L numbers and psychological reaction will be wildly different.

   Deploy the EA with the same .set, same Magic Numbers, same VPS, same monitoring. Treat the demo as production rehearsal — anything you'd do for live, do for demo.

3. الخطوة 3: Run for 30 days and capture all data

   30 days is the minimum credible demo period. Less than that, you might be testing a single regime (a bull trend, a news-light week) and learning nothing about the EA's robustness.

   During the demo: • Don't intervene. The whole point is to observe the EA's autonomous behaviour. Manual closes invalidate the test. • Capture daily account screenshots. Time-stamp them. These prove the demo ran for the claimed period. • Run the EA's standard reporting. Compare weekly P&L against backtest expectation. Anomalies are signal — investigate. • Survive at least one high-impact news event (NFP, FOMC). The EA's behaviour around news is the highest-risk failure mode; if you don't see a news event in the demo period, extend until you do.

   After 30 days, export the full trade history and compare to backtest: • Net Profit within ±30% of backtest's same-period equivalent? Acceptable. • Drawdown ≤ 120% of backtest's? Acceptable. • Trade count within ±20% of backtest's? Acceptable. • Profit Factor within ±0.3 of backtest? Acceptable.

   If all four pass, proceed to small live. If any fails, investigate before continuing — the EA may have a real issue, or backtest assumptions may not hold.

4. الخطوة 4: Graduate to a small live account

   Open a real-money account funded with 10-25% of your intended final capital. If your plan is $20k live, fund the graduation phase with $2-5k.

   Why small live and not skip straight to full size:

   Demo and live differ in subtle ways that only surface with real money: • Slippage on real liquidity. Demo uses the broker's synthetic fills; live uses actual liquidity provider fills which can slip 0.5-2 pips on volatile entries. • Spread widening during news. Demo widens spreads; live widens them more, often briefly hitting 5-10× the typical spread during the first 30 seconds of NFP. • Broker risk-management interventions. Live accounts may face leverage cuts during news (e.g. 1:30 → 1:5 for 15 minutes around FOMC). Demo never does this. • Psychological pressure. The cognitive difference between watching a $200 demo loss and a $200 live loss is bigger than experienced traders admit. Decisions you'd make calmly on demo, you may second-guess on live.

   Run small live for another 30 days. Compare to demo: • Live P&L within ±20% of demo same-period? Good. • Drawdown ≤ 130% of demo? Good. • Anomalous trades you can't reconcile? Investigate before scaling.

   If small live tracks demo, you're cleared to scale. If small live materially underperforms demo, the gap reveals the live-trading friction your strategy actually faces — incorporate this into your performance expectations before scaling.

5. الخطوة 5: Scale to full size in stages

   Don't go from 25% to 100% overnight. Scale in 3 stages: 25% → 50% → 100%, with at least 14 days between each step.

   Why gradual scaling: • Some EAs degrade with size — at small lot sizes, fills are clean; at large lot sizes, liquidity friction kicks in and slippage degrades execution. Catch this early by observing per-trade slippage at each size step. • Your psychological response scales nonlinearly with position size. A 5% drawdown at $5000 ($250) feels different from 5% at $20000 ($1000). Confirm you can hold each step before stepping up. • Margin requirements scale linearly but the margin headroom you have at 25% is comfortable; at 100% it's tighter and a sudden spread widening can margin-call you. Verify margin headroom at each step.

   After reaching full size, maintain a monthly review cadence: compare actual performance to backtest + demo + each prior month. Drift > 30% is a signal to re-investigate — either the EA's edge is degrading (regime change) or your live conditions have changed (broker spread structure, leverage policy).

6. الخطوة 6: Warning signs that block graduation

   Don't proceed to live (or scale up) if you observe any of:

   • Demo profit factor < 1.3 (close to breakeven; live will likely be unprofitable after execution friction). • Demo drawdown > 150% of backtest drawdown (the EA isn't behaving as backtested; investigate before risking capital). • Repeated 'Cannot open position' / 'Off quotes' errors in the Journal log (operational issue that needs fixing). • EA hung or stopped trading for > 4 hours during normal market hours (the operational setup isn't reliable enough for live). • Anomalous trades you can't explain (trades with stops far from the EA's documented logic). The EA might have a bug; fix before live. • Material spread widening blew out your stop more often than expected. Either tighten broker selection (move to ECN), or widen stops in the EA config.

   These aren't 'maybe' signals — they are blockers. Address them on demo, where they cost nothing. Then resume the demo period from day zero of the fix.

أخطاء شائعة يجب تجنبها

• ✗ Skipping demo entirely because 'I trust the backtest'إصلاح: Backtest is necessary but not sufficient. Demo catches operational issues backtest can't.
• ✗ Demo on a $100k account when planning live $5kإصلاح: Demo balance must match planned live balance. Absolute numbers matter for psychological calibration.
• ✗ Treating demo as a casual test rather than production rehearsalإصلاح: Same broker, same EA, same .set, same VPS, same Magic. Anything different invalidates the validation.
• ✗ Going from demo to 100% live size immediatelyإصلاح: Stage 25% → 50% → 100% with 14-day observation between. Catches size-sensitive issues early.
• ✗ 30-day demo with no news events in the periodإصلاح: Extend until at least one NFP and one FOMC. EAs that pass calm weeks but fail at news are common.
• ✗ Not capturing daily demo screenshotsإصلاح: Future-you, debugging a live issue, will need to compare against demo behaviour. Daily screenshots are the evidence.

الأسئلة الشائعة