How to Structure Your EA Demo Testing Period — What to Verify in 3 Months

It's completely understandable to want to run an EA on a live account as soon as the backtest results look good. But there is always a gap between backtesting and real market behavior. Running a demo account for three months is the single most important step you can take to safely measure that gap before risking real capital.

Why Demo Testing Matters

What Backtests Cannot Reproduce

These differences compound over time, which is why a gap between backtest results and live performance is inevitable. The demo testing period is your window to measure that gap quantitatively.

Recommended Demo Testing Duration

Why 3 months? Market character — trending, ranging, high-volatility, low-volatility — rotates on a monthly cycle. Three months gives you exposure to multiple market environments. With only one month of data, you cannot rule out the possibility that results were simply "lucky (or unlucky) timing."

6 Checkpoints During Demo Testing

Checkpoint 1: PF Deviation from Backtest

Deviation = Demo PF ÷ Backtest PF × 100%

Checkpoint 2: Do Real Spreads Match Your Backtest Settings?

Log the actual spread values you observe on the demo account.

How to check in MT5:

1. In the Market Watch window, monitor the Ask/Bid for your target symbol.
2. Also observe spread behavior during news releases, when spreads often spike dramatically.

If your backtest spread setting was lower than actual observed spreads, re-run the backtest using the real spread value you measured.

Checkpoint 3: Is the EA Entering at the Correct Times?

If your EA uses a TimeFilter, confirm it is only opening trades during the configured hours.

The Experts tab log records entry events along with ATR values, EMA values, and other indicator readings — check these to make sure they fall within expected ranges.

Checkpoint 4: How Do You React to Losing Streaks?

Even on a demo account, pay deliberate attention to how you feel when losses pile up:

• Do you want to change parameters after 5 consecutive losses?
• Do you want to shut it down when DD hits 10%?

Training yourself not to make emotional decisions on a live account is one of the most valuable aspects of demo testing. If you cannot hold firm on a demo, the same situations on a live account will almost certainly break your discipline.

Checkpoint 5: 24-Hour Stability on VPS

If you plan to run your EA on a VPS, use the demo period to verify the following:

• Can the VPS run continuously for a week or more without issues?
• Does MT5 start automatically after a reboot?
• Are there any overnight errors? (Check the Experts tab every morning)

Checkpoint 6: Monthly Performance Records

At the end of each month, save an HTML report from your account history.

Key metrics to record:

Date:           [Month, Year]
Trades:         [count]
Net P&L:        [$ amount] ([%])
Max DD:         [%]
Win Rate:       [%]
Profit Factor:  [value]
BT Deviation:   [%]

Once you have three months of data, compare it against your backtest and make an informed go/no-go decision for the live transition.

Criteria for Going Live

You can consider transitioning to a live account once all of the following conditions are met:

• Demo operation has run for at least 3 months (for H1 timeframe)
• At least 50 trades have been executed
• Demo PF is within 70–130% of the backtest PF
• Maximum DD is within 1.5× the backtest maximum DD
• No persistent errors in the MT5 Experts tab
• You were able to continue operating without changing parameters through 5–10 consecutive losses
• 24-hour VPS operation has been stable

Gradual Lot Sizing After Going Live

Immediately after transitioning to a live account, start with the minimum lot (0.01) or half your intended lot size.

Month 1 after live:   RiskPercent = 0.3–0.5%  (lower than demo)
Month 3:              RiskPercent = 0.5–0.7%  (if no issues)
Month 6:              RiskPercent = 0.7–1.0%  (matching demo level)

Even after going live, keep the demo account running in parallel. This lets you continuously monitor the gap between demo and live performance.

Common Failure Patterns

Failure 1: Deciding "It's fine" after just one week and going live

One week is not enough time for market conditions to change even once. Statistically, you have learned nothing meaningful in that window.

Failure 2: Changing parameters after a loss and restarting the demo

If you restart the demo every time you tweak parameters, you will never actually transition to live. If the DD is within the range seen in backtesting, the right move is to hold course without making changes.

Failure 3: Not matching demo and live settings exactly

If your Lot size, RiskPercent, or MagicNumber differs between demo and live accounts, you cannot make a valid comparison. Always double-check your settings at the moment of transition.

Summary

Demo testing is not wasted time — it is an investment that protects you from costly mistakes on a live account.

Three months of demo testing gives you:

• A deep understanding of how your EA actually behaves
• Practice at removing emotional decision-making from the equation
• A quantitative measurement of the gap between backtest and live performance

Rushing to go live and suffering a large early drawdown that wipes out a significant portion of your capital will cost you far more in the long run than the three months it takes to verify things are working as expected.

FAQ

Q: Should the demo account balance match my intended live balance?

Yes. Set the demo and live account balances to the same amount. Even with an identical RiskPercent, a different account balance will produce different lot sizes, making performance comparison meaningless.

Q: Can I still lose money on a live account even after 3 profitable months on demo?

Yes. If your demo period happened to coincide with a specific market environment — a strong trending phase, for example — you may see different results when the market shifts to a ranging environment. This is precisely why demo testing is a useful reference, not a guarantee.

Q: Will performance differ between an XMTrading demo and an Exness demo?

Yes. Spreads and swap rates vary by broker, so performance will differ. It is important to demo trade with the same broker you plan to use on your live account.

Related Pages

• When to Move from Demo to Live Trading
• How to Interpret Forward Test Results
• GOLD EMA ATR EA — Demo Compatible
• Daily MT5 EA Operations Checklist