Manual vs. automated backtesting isn't a question with one right answer. They're two tools that solve different problems. Here's how they really differ, when each one makes sense, and why most new traders should start with manual backtesting.
What manual backtesting is
You replay price history one candle at a time and decide for yourself when to enter, where to place your stop and when to exit, exactly as you would on a live account, but without risking any money. You don't need to write any code. It's the approach we use in our step-by-step guide.
What automated backtesting is
You code your strategy's entry and exit rules (for example, in Pine Script, TradingView's scripting language, or MQL, the one used by MetaTrader). The software then runs them over thousands of historical trades in seconds and gives you summary statistics at the end. It's fast and objective, but you need to know how to code, or pay someone to do it for you.
If your strategy is discretionary, manual backtesting trains exactly what you'll need when you trade live.
See plansSide-by-side comparison
| Criteria | Manual | Automated |
|---|---|---|
| Requires coding | No | Yes |
| Speed to test an idea | Slow (candle by candle) | Very fast (thousands of trades in seconds) |
| Trains your execution and discipline | Yes | No |
| Main risk | Hindsight bias | Overfitting ("curve fitting") |
| Best for | Discretionary strategies | Systematic or algorithmic strategies |
A quick note on terms: a discretionary strategy relies on your judgment to decide each trade, while a systematic one follows fixed rules a computer could apply on its own. Hindsight bias means letting what you already know about the future chart influence your decisions.
Overfitting, explained with an example
Say you code a strategy and the automated backtest gives you mediocre results. You tweak one setting, like the trend filter, and things improve a little. You tweak another, and they improve more. You keep going until you find the combination of settings that gives the best possible result on that specific stretch of history.
Here's the problem: that combination probably doesn't capture a real market edge. It has been tuned to the random noise of that particular period. That's overfitting, also called curve fitting. When you test it on a different stretch of history, or trade it live, the results usually collapse. It's the biggest risk of automated backtesting, and a strong reason not to blindly trust a "perfect" result.
When to use each one
If your strategy depends on reading context (market structure, confluence, your own judgment), manual backtesting is the only kind that truly tests it, because it also tests your ability to carry it out. If your strategy is a set of 100% objective rules that can be coded, automated backtesting gives you a statistical answer much faster.
A common middle ground: use automated backtesting as a quick filter to throw out ideas with no statistical basis. Then save manual backtesting for the few that pass and that you plan to trade live, using your own judgment.
Why most new traders should start with manual backtesting
New traders rarely fail for lack of a statistically sound strategy. They fail because they can't follow it with discipline when price moves against them. Automated backtesting doesn't address that problem at all. Manual backtesting trains exactly that, session after session. Once you've logged enough trades, the next step is learning how to read your backtesting statistics.
Train the part automated backtesting can't touch: your execution.
See plansWhat you need to get started with each approach
| You need | Manual | Automated |
|---|---|---|
| Coding skills | No | Yes (Pine Script, MQL or another language) |
| 100% objective rules that can be coded | No, your own judgment is allowed | Yes, essential |
| Time | Short, repeated sessions | Initial setup, then it's fast |
If your strategy still includes some judgment calls, like "wait for confirmation" or "assess the context," you can't code it without simplifying it. Until you make it 100% objective, manual backtesting is your only real option.
Frequently asked questions
- Is automated backtesting more reliable than manual backtesting?
- Not necessarily more reliable, just faster at testing the pure logic of a system. Manual backtesting is still the only way to train your execution and discipline, which automated backtesting doesn't touch at all.
- Can I combine manual and automated backtesting?
- Yes, and many serious traders do. They use automated backtesting to quickly rule out ideas with no statistical basis, and manual backtesting for the discretionary strategies they actually plan to trade, where human execution matters.
- Is manual backtesting more prone to bias?
- Yes. Hindsight bias (accidentally peeking at what happens next) is the main risk of manual backtesting. A tool that replays price history one candle at a time, without letting you skip ahead, greatly reduces that risk compared with reviewing a full chart by hand.
- Which is better for a new trader?
- Manual, almost always. It trains analysis and execution at the same time, which is what new traders lack most. Automated backtesting gives you fast answers, but it doesn't build the discipline you need to trade live.
- What exactly is overfitting (curve fitting)?
- It's tuning a strategy's settings so closely to one specific stretch of price history that it stops capturing a real market edge and only captures the noise in that data. The result looks excellent in the backtest and falls apart in live trading.
- Does Trade & Repeat offer automated backtesting?
- Not for now. We focus on manual backtesting, because it's the kind that trains your execution and discipline, not just the logic of a strategy.
Keep reading: How to backtest a forex strategy step by step · How to read your backtesting statistics · The best free backtesting platform: an honest comparison · Confirmation bias in backtesting: why you only review your winning trades · From backtesting to a funded account: what really changes when you go live · Dynamic vs. fixed position sizing: how to backtest risk that changes with your streak
Educational content only; it is not financial advice. Trading financial markets involves the risk of losing capital.