Manual vs Automated Backtesting: Which Fits the Way You Trade
By hand tests your judgement slowly; a tester runs coded rules fast. A side-by-side of what each can and cannot see, from the testers' own documentation.
3 Nov 2026 · 5 min read
On this page
Backtesting by hand tests your judgement slowly; an automated test runs coded rules fast but only knows what the code and the data give it. Neither is the more serious method. They answer different questions, and the choice usually comes down to one thing: can your rules be written completely as code?
What each one can and cannot see
The right-hand column quotes the testers' own documentation.
| What you are testing | By hand | Automated, in a tester |
|---|---|---|
| Rules that need judgement | Can be tested, if you apply them the same way each time you meet them. | Only what can be written as code. A rule you cannot put into words is left out. |
| Speed | Slow: you step through the data yourself. | Fast. MetaTrader 5's tester trades speed for accuracy by mode: it calls "Every tick" the most accurate and slowest, and says of a rougher mode "You should only use it for quick and rough strategy estimation". |
| Price path inside a bar | Only what your chart or replay tool shows. A bar replay shows bars. | Set by the data mode. MetaTrader 5's "1 minute OHLC" emulates four prices per minute bar; its "Every tick based on real ticks" uses ticks recorded by the broker. cTrader offers tick data from the server, one-minute bars or one-hour bars. |
| Spread | You enter it, trade by trade. | MetaTrader 5's documentation: "During testing, the spread is not modeled but is taken from historical data." cTrader lets you set a fixed spread, match the symbol's live spread, or draw a random one between a minimum and a maximum. |
| Commission | You enter it, trade by trade. | cTrader: set as a rate per million units traded. MetaTrader 5's tester help has a section on it: per lot or per deal, in money, percentage or points, charged at once or at the end of a day or month, or "Use predefined commissions" to take the current trading account's settings. |
| Slippage | Your own assumption, written into the log. | cTrader's backtesting help does not mention it. MetaTrader 5 lists execution delay as a separate setting for trade actions in an automated program. |
| Hindsight | The main risk: a glance at the next bar changes what you take. | Removed for the entry itself, but replaced by two others: a look-ahead mistake in the code, and tuning settings on the same data you then report. |
| Repeatability | Depends on your discipline. Two runs by the same person can differ. | Same data, same settings, same result. |
When by hand fits
- Your approach depends on reading the chart. If you pick levels, judge structure or skip setups on context, a coded version is a different strategy, and its result is not yours.
- You want a first look at an idea before you spend time coding it. A short, honest log shows quickly whether the idea is worth more work. The steps are in how to backtest a trading strategy by hand.
- You want to see the trades. Walking through each one teaches things a summary table does not.
When automated fits
- The rules are mechanical. Entry, stop, exit and size can all be stated without “it looks like”.
- You need a lot of history or many markets, which no one steps through by hand.
- You want to rerun after a change and know the only difference is the change.
What changes with automation, and what does not
An automated test removes hand errors and adds machine ones. A tester that builds ticks from one-minute bars is making up the path inside every bar; its mode decides how much it makes up (see bar replay against tick data). A tester that lets you try ten thousand settings will find some that fit the past beautifully, which says little about the future (see backtesting mistakes). And neither method tells you how you will act when the rules are losing. That is what a forward test is for: backtest, forward test and demo.
Tools that sit between the two, such as a chart that reveals one bar at a time, are compared in backtesting software compared.
Trepeat today
Trepeat does not sell a backtester or a replay tool.
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.
Written by the Trepeat team.