Best backtesting software? Judge it by the job — and by whether it's honest
“Best” depends entirely on the job you're doing and on whether the tool is honest about what it can't do. There is no single best backtesting tool for everyone — a futures scalper, a long-horizon allocator and someone sanity-checking one rule on one famous stock all need different things. So this page doesn't rank tools. It does two more useful things.
First, the five honesty checks you can run on any backtesting tool — the part that decides whether you can trust what it shows you. Then a guide to which kind of tool fits which job, with us placed honestly as the right fit for exactly one of them, graded on the same five checks, limits and all.
This page is independent. No affiliate links, no sponsored placements, no pay-to-win list. A roundup that gets paid to place tools isn't a guide, it's an ad — so this one has no reason to point you anywhere, and the only tool it sells you on is the idea of checking any tool honestly.
The five honesty checks — apply these to any tool
Before any tool is named, here is the part that actually matters: five questions you can run on any backtester — including the screenshots in a roundup — to tell an honest result from a highlight reel. Each one catches a specific way a backtest fools people.
- 1.Costs on? Does it charge realistic commission AND slippage on every fill, or print a frictionless result?
Friction quietly sinks many a rule that looked like a winner on paper — a rule that only comes out ahead at zero cost doesn't really come out ahead. Catches: the demo that draws a beautiful curve because it trades for free.
- 2.Look-ahead prevented? Do the rules only ever act on bars that have already closed?
Look-ahead and repaint are the single most common way a backtest manufactures an edge that never existed — a signal using a price it couldn't have known in real time. Catches: the curve that was impossible to actually trade.
- 3.Losses and drawdown shown? Does it foreground every losing trade and the full peak-to-trough drawdown?
The pain you'd have had to sit through is the result, not a footnote — a strategy is the stretch that would have made you quit, too. Catches: the highlight reel that shows only the happy ending of the equity curve.
- 4.Measured against buy & hold? Is the headline the EXCESS over simply holding, or an absolute return?
A rule that made money only because the market rose didn't earn its complexity — any rising market would have handed you that. Catches: the strategy that “made 200%” while buy & hold made 240% over the same window.
- 5.Whole universe, or cherry-picked? Does it test a fixed, pre-committed set and disclose how many variants were tried?
Enough variants will always throw up one lucky winner, and tuning-multiplicity is invisible unless it's counted. Catches: the one chart kept because it looked good after the fact — survivorship and p-hacking wearing a track record.
The honest ceiling: even a tool that clears all five can only tell you a rule survived its own history — never that it will make money next. Nothing honest can promise that, and a tool that implies otherwise just failed a sixth check it didn't tell you about.
Which tool for which job
The query hides a mix of very different jobs. Match the job to the kind of tool built for it — this routes by what each is for, not by a feature count or a price.
Best for working on the chart itself: it pairs charting with its own scripting language (Pine) and an on-chart strategy tester. Reach for it when the chart, the scripting flexibility and a big script-sharing community are what you want.
Best for futures and systematic strategies written in code: a desktop platform whose analyzer runs strategies over bar data one bar at a time. Reach for it when you trade futures or want deep systematic tooling for strategies you program yourself.
Best for questions about a mix of holdings rather than one asset: both work at the allocation and rebalancing level, and one of them can also drive automated rebalancing — a thing we deliberately don't do. Reach for these when the job is a long-horizon portfolio, not a single-asset entry rule.
Best for assembling and testing a strategy visually without programming: these let a non-coder click a rule together. Reach for one of them when you want to build by dragging and connecting rather than typing — the real peer cohort to us on the no-code side.
Best for unlimited flexibility: write the rules yourself, bring your own data and assumptions, and nothing a vendor can gate or change stands between you and the test. Reach for it when you can program and want no limits — the honest answer for a coder.
Best for one job: a free, no-code check of whether a simple rule on a famous liquid asset actually beat simply holding it — costs on, look-ahead prevented, losses and drawdown shown, judged across a whole fixed universe rather than one lucky chart. Reach for it for exactly that, and read the limits below for where it isn't the one.
Features and prices change constantly, and a stale claim about a live financial tool can do real harm — so this names what each is BUILT for, not a spec sheet. Check each tool's own site for current details.
We graded ourselves on the same five checks
Turning the framework on ourselves, plainly — how this tool does on each of the five, then where it's simply the wrong tool. Stating the limits is the point: a tool that hides what it can't do is the same tool that hides a strategy's losses.
- Costs on — every fill is charged realistic commission and tiered slippage; there is no zero-cost mode.
- Look-ahead prevented — rules act only on closed bars, and fills land at the next bar's open, never the signal bar.
- Losses and drawdown shown — every losing trade and the full drawdown are in plain sight, not a footnote.
- Measured against buy & hold — the headline is always the excess over simply holding that same asset.
- Whole universe — a fixed, pre-committed set of famous assets, with the number of variants you've tried counted against you.
Where we're the wrong tool
Daily bars only — no intraday, tick or minute data. For day-trading or fine execution detail, a chart-first or futures platform fits the job far better.
A fixed, famous-asset universe — famous stocks, ETFs and crypto only, with no options, futures or forex, and no arbitrary ticker or your own CSV. If you need that, your own code is the honest answer.
Non-custodial and alerts-only — we never execute, automate or rebalance anything. If you want a backtest that can also place or automate trades, that's a different kind of tool, and we deliberately aren't one.
So this tool is the right fit for exactly one job — a free, no-code honest reality-check of a simple rule on a famous asset — and openly the wrong one for the other jobs above. That's the whole point of grading it on the same checklist we handed you.
The honest ceiling
Even a tool that clears all five checks can't promise you'll make money — the most it can tell you is that a rule survived its own past. And most don't: tested honestly, most famous strategies on famous assets lag simply holding the asset. The value isn't being told you'll win; it's finding the rare rule that genuinely beats holding, and being able to trust the result when you do. See what actually beats buy & hold — every winner and every loser →
The fastest way to learn the five checks is to watch a tool apply them. This link runs a real backtest — RSI mean-reversion on Apple over the last five years — live, with costs on and measured against holding. Change the asset or the rule once it opens.
Already tested one elsewhere? Bring your trade list and run the checks on it →
Stratsemble is not affiliated with any other tool named here. TradingView, NinjaTrader, Composer and Testfol.io are named only to describe, in nominative fair use, what kind of job each is built for — not to rank them. Other tools' features and prices change over time; check their own sites for current details.
Hypothetical / simulated results — Past performance is not a reliable indicator of future results. This is an educational, analytical tool — not investment advice, and not a recommendation to buy or sell anything. You make all decisions and execute on your own broker. Educational and analytical, not investment advice, and not a recommendation that any tool is suitable for you. It is non-custodial: you make all decisions and execute on your own broker.