Home › Learn › Guides

How to Forward-Test a Strategy Before Risking Real Money

Updated 28 July 2026 · 7 min read · PipTax education

If you want to forward-test a trading strategy properly, you need a process that runs in real time, on a real (or demo) account, with no peeking at future price data. Forward-testing is the bridge between a strategy that looks good on a spreadsheet and one you can actually trust with your own money — and skipping it is one of the most common reasons new traders blow up accounts that "should have worked."

What Forward-Testing Actually Is

Forward-testing means executing your strategy's rules going forward from today, trade by trade, without knowing the outcome in advance. It's different from backtesting, which runs a strategy against historical data all at once.

Backtests are useful for a first-pass filter, but they're vulnerable to:

• Curve-fitting — tweaking rules until they match past data perfectly
• Hindsight bias — subconsciously choosing entries/exits because you know what happened next
• Missing execution reality — no slippage, no requotes, no emotional hesitation

Forward-testing fixes this by forcing you to make decisions with the same uncertainty you'll face live. You place the trade, you don't know the outcome, and the market decides. This is the only way to see how a strategy behaves under genuine unknowns, including your own discipline under pressure.

Think of it as the strategy's probation period — a fixed window where it has to prove itself before you trust it with meaningful capital.

Setting Up a Proper Forward Test

Before you place a single trade, define the test in writing so you're not tempted to move the goalposts halfway through:

• Fixed rule set — entries, exits, stop-loss, take-profit, position sizing, all written down
• Fixed sample size — decide upfront: 30-50 trades, or 4-8 weeks, whichever comes first
• Fixed account type — demo first, then minimum live size, not skipping straight to full capital
• Fixed instrument and timeframe — don't switch pairs or charts mid-test because results look weak
• Fixed review point — a specific date or trade count when you'll sit down and assess results

Write this on one page. If a rule isn't on the page, it's not part of the test — and if you find yourself breaking your own rules mid-test, that's data too. It tells you the strategy may be sound but you're not yet disciplined enough to run it consistently.

Demo First, Then Small Live Size

Demo trading is the cheapest place to catch obvious flaws — a strategy that loses money constantly on demo has no business going live. But demo has real limits: there's no genuine fear of loss, so your execution discipline on demo often doesn't transfer to live trading.

A sensible progression:

1. Demo phase — confirm the rules are mechanically sound and you can execute them without confusion 2. Minimum live size — smallest position size your broker allows, real money, real emotions 3. Scaled live size — only after the minimum-size phase holds up, increase gradually

The minimum-size live phase is where most "great on paper" strategies reveal their real weakness: hesitating on entries, moving stops out of fear, or closing winners too early. These are behavioural problems no backtest or demo run will ever show you.

Recording What Actually Happens

A forward test without a journal is just trading with extra steps — you need a record to review objectively rather than relying on memory, which tends to remember wins and forget losses.

For every trade, log:

| Field | Why it matters | |---|---| | Entry price & time | Confirms you followed the rule, not a gut feeling | | Exit price & time | Same, for exits | | Position size | Checks sizing discipline stayed consistent | | Reason for entry | Flags rule creep over time | | Spread/commission paid | Real cost, not assumed cost | | Emotional note | Did you hesitate, override, or panic-close? |

That spread/commission column matters more than most journals give it credit for. A strategy that nets +2% over 40 trades on paper can turn flat or negative once real costs are subtracted — which is exactly why comparing actual costs paid against what the /audit.html tool models is worth doing partway through your test, not just at the end.

Measuring Results Honestly

Once you hit your fixed sample size, review it without moving the goalposts. Look at:

• Win rate — but don't judge a strategy on this alone
• Average win vs average loss — the real driver of expectancy
• Maximum drawdown — the worst stretch, not just the average
• Consistency — did results cluster in one lucky week, or spread evenly?
• Cost drag — how much of the gross return spreads and commissions actually consumed

A strategy that's marginally profitable before costs but negative after them isn't a strategy — it's a way to fund a broker. This is precisely where a genuine cost comparison across venues matters; two brokers can produce very different net results from the same gross trades, which is the whole reason /brokers/index.html and /audit.html exist as a first stop before choosing where to run the live phase.

Common Mistakes That Invalidate the Test

Watch for these — they quietly ruin an otherwise good process:

• Stopping early because of a bad run — a real drawdown is data, not a reason to quit
• Stopping early because of a good run — three great trades isn't a strategy, it's variance
• Changing rules mid-test — restart the count if you do this, don't just carry on
• Ignoring real costs — testing on a demo with unrealistic spreads gives false confidence
• Skipping the journal — memory is not a substitute for a written record

If you're unsure how a rule change should be tested cleanly, PipTax's /methodology.html walks through how we structure comparisons — the same discipline applies to your own strategy testing.

Final Checks Before Going Fully Live

Once your forward test is complete and the numbers hold up:

• Confirm the broker you tested on matches the one you'll trade live with, or re-check costs via /audit.html if not
• Re-read your journal for any rule you quietly bent and fix it before scaling size
• Increase position size gradually, not all at once
• Keep the journal running even after "graduating" — live trading is really just forward-testing that never stops

Learning to forward-test a trading strategy properly is less about finding a shortcut to profits and more about building a repeatable, honest process — one that tells you the truth about a strategy before real money is on the line. For more structured lessons on building and testing strategies, PipTax's /school/index.html is a good next stop, and remember that trading always carries risk of loss regardless of how well a strategy has tested.

Keep going: Audit Methodology Index Index