Bar Replay vs Tick Data Backtesting: What Resolution Changes

One-minute bars are enough to replay a chart and judge most entries. Ticks matter when a single bar reaches both your stop and your target and you need to know which came first. A bar keeps four prices: open, high, low and close. Everything that happened between them is gone, so the tool, or you, has to fill the gap.

What a bar cannot tell you

Take a made-up long trade at 1.1000 with a stop at 1.0990 and a target at 1.1020. That is 10 pips of risk for 20 pips of reward, so the trade wins 2.0R or loses 1R. The next one-minute bar prints a low of 1.0988 and a high of 1.1022.

The bar's range covers the stop and the target. The bar says both were touched; it does not say in which order. If the stop came first the trade is −1.00R; if the target came first it is +2.00R. The two readings differ by 3.0R on that one trade. A tester working from bars has to choose, and a hand backtest has to choose too.

Suppose, with made-up numbers, that 10% of the trades in a test land in a bar like this. Choosing the most favourable reading every time instead of the least favourable moves the average result by 10% × 3.0R = 0.30R per trade, which can be the difference between a rule that looks worth trading and one that does not. How often it happens in a real test depends on how wide your stops and targets are compared with the bars: wide ones rarely sit inside one bar, tight ones often do.

• If you are using bars, pick one rule for every ambiguous bar, apply it to all of them, and say which in your log. One choice is to count the stop first, which errs against the result.
• If you have ticks, the order is in the data, so the question disappears. What you pay for it is size and speed: tick data is larger than the same period in bars and slower to run.

How five tools describe their own data

What each tool states about its data. This is a description, not a ranking.

Two things follow from the MetaTrader 5 and cTrader rows. First, a tick-level test is not one thing: MetaTrader 5 separates ticks it simulates from ticks the broker recorded, and treats them as different modes. Second, the same strategy can give different results in different modes, which is why the platform documentation itself suggests checking a quick mode against a more detailed one.

When bars are enough, and when to want ticks

• Bars are usually enough when you enter on a bar's close, hold for many bars, and your stops and targets are wide compared with one bar's range.
• Ticks matter when stops or targets are tight against the bar size, when you trade fast news moves, or when the order of events inside a bar is the whole question.
• Spread matters either way. A bar carries one price per field, and a tester has to take the spread from somewhere. MetaTrader 5's documentation says the spread “is not modeled but is taken from historical data”. See backtesting mistakes for what ignoring it does.

Where free tick and one-minute files come from, and what each source says about using them, is in free forex historical data and what the terms say. The by-hand method this page assumes is in how to backtest a trading strategy.

Trepeat today

Trepeat does not sell a replay tool or a backtester.

We're working on a free journal and a backtester, with no date yet. Today Trepeat is a trade copier for MT4, MT5 and cTrader, currently on demo accounts, and you can add notes, a grade and screenshots to any copied trade once it closes, with a calendar of your results.

Copies can fill at different prices from the original for reasons of their own, covered in why follower fills differ. How the journal works.

Written by the Trepeat team.