Forward testing a strategy on demo or small size

What forward testing is

Forward testing means running a fixed set of trading rules on live market prices as they happen, and recording every trade. You can do it on a demo account, which uses virtual money, or on a live account with very small position size.

It differs from backtesting, which applies rules to past data. A backtest can be repeated and adjusted endlessly, so it is easy to fit the rules to history. A forward test cannot see the future, so it shows how the rules behave on data they have never met.

Neither method proves a strategy will make money. A forward test only produces evidence, and that evidence can still mislead.

Demo account testing vs small live size

Many traders begin with demo trading practice and then move to a small live account before increasing size. If you use a demo, set its balance close to what you would actually trade. A large virtual balance makes losses feel unimportant. See the Demo account page for how demo accounts work on MetaTrader.

Remember that leveraged forex and CFD trading can lose money fast, even at small size. A small live test is still real risk.

How to forward test a strategy

Write the rules before the test starts

A rule is testable only if two people reading it would take the same trade. Write down:

• Market and timeframe: which instruments and chart period you trade.
• Entry condition: the exact signal, for example a close above a moving average after a pullback.
• Stop-loss: where it goes and why, in price terms, not "when it feels wrong".
• Exit: take-profit level, trailing rule or time exit.
• Position size: a fixed percentage of the account at risk per trade. The Calculators page helps convert this into a lot size.
• Filters: sessions you avoid, news events you skip.
• Maximum exposure: how many open trades or how much total risk is allowed at once.

Put these rules in a trading plan and do not change them during the test.

Record every trade

Keep a journal with date, instrument, direction, entry, stop, exit, size, spread at entry, swap paid, result in account currency and result in "R". R is the amount you planned to risk on the trade. A short note on whether you followed the rules is also useful. Skipped signals count as data too, so record them.

Review only at the end of a block

Decide in advance when you will review, for example after a set number of trades. Reviewing after each loss tempts you to change rules, and then you are no longer testing a single strategy.

Sample size: how many trades are enough

A short run of trades says very little. A random series can produce a long winning or losing streak, and this tells you nothing about the rules. The fewer trades in the sample, the more likely the result is luck.

Practical guidelines:

• Count trades, not weeks. A slow swing system may need many months to build a useful sample.
• Cover different market conditions: trending, ranging, quiet and volatile periods.
• Look at the spread of outcomes, not only the total: average win, average loss, win rate, longest losing streak and largest drawdown.
• Compare the forward results with the backtest. A large gap suggests overfitting or unrealistic backtest costs.

If the sample is too small to judge, the honest conclusion is "not enough data", not "it works".

Hypothetical example

This is a hypothetical illustration, not a real result or a recommendation.

A trader writes a trend following rule for EURUSD on the hourly chart: enter in the direction of a longer moving average after price pulls back to a shorter one, with the stop beyond the recent swing, a target set as a multiple of the risk, and a fixed small risk per trade. They run it on a demo account with a balance close to their intended live balance, log every signal, and agree not to review until a set block of trades is complete. After the block they compare win rate, average R and drawdown with their backtest, and note how many trades they skipped or managed by hand.

How forward tests lose money or mislead

• Too few trades: a lucky streak looks like an edge.
• Changing rules mid-test: each change resets the test, and the trader often never notices.
• Demo fills: demo execution may not show live slippage, requotes or wider spreads around news.
• Ignored costs: spread, commission and swap fees can turn a small gross edge into a net loss.
• Size too large on live: one bad run hits the account hard. See risk management.
• Cherry-picked periods: starting the test in a calm trending market and stopping before conditions change.
• Discretion creeping in: skipping "ugly" signals that would have been losers or winners.
• Scaling up too fast: moving from demo to full size without a small live stage, where trading psychology changes behaviour.
• Risky money management: martingale or grid methods can look smooth on a short test and then fail all at once.

Legal context in Pakistan

Before moving any test to a live account, read the Law and regulators page. Per the SECP, offshore platforms have no Pakistani licence or oversight and offer no legal protection or recourse.

Frequently asked questions

Is forward testing the same as demo trading practice?

Not exactly. Practice can be loose. A forward test follows fixed written rules and records every trade so the results can be judged.

Can a good forward test guarantee live profits?

No. Live execution, costs, emotions and changing markets can all produce different results. Losses can be fast with leverage.

Should I change my rules if I hit a losing streak?

Not during the test. Finish the planned block, then review. Changing rules midway means you are testing something else.

Is a demo account enough on its own?

It is useful for checking rules and platform skills. A small live stage shows real fills, real costs and how you behave with money at risk.

Where do I learn the basics before testing?

Start with Trading basics and the Trading glossary, then read about Trading strategies.