Where to backtest on longer history for free
- A broker's API usually serves months to a year of intraday history. That's enough to reject a bad idea, not enough to trust a good one.
- Free tools with longer history exist: thinkorswim for Schwab clients, TradingView (daily bars on the free plan), NinjaTrader's free platform, and QuantConnect's free cloud backtests with years of futures, stock and crypto data.
- Each has limits, and some need a paid plan for the deepest intraday tests. We say which below.
- More history only helps if you test honestly: keep data back, walk forward, get enough trades and charge real costs.
A strategy that looks great on six months of one market may only be great at those six months. Markets move through regimes (trending, choppy, calm, panicky) and a short history can sit entirely inside one of them. Before you trade a strategy, or pay for a prop firm evaluation to trade it, it's worth checking it on years of data, not months. You don't need to pay for that. Here are the free options we'd point a friend to, with what each one actually gives you as of November 2026. None of the links below earn us anything.
The free options at a glance
- thinkorswim: free for Schwab clients (no minimum deposit), or a 30-day Guest Pass. Intraday charts up to 360 days; daily charts back as far as the data goes. Best for chart strategies in thinkScript.
- TradingView: the free Basic plan; deeper tests on paid plans. 5,000 bars on intraday charts; all available data on daily and higher. Best for daily and swing strategies in Pine Script.
- NinjaTrader: a free platform for backtesting and simulation, no deposit. NinjaTrader's own historical data (depth not published). Best for futures strategies in C# (NinjaScript).
- QuantConnect: a free tier with unlimited cloud backtests. Minute and daily data: US stocks from 1998, US futures from 2009, crypto from 2015. Best for coded strategies in Python or C# on many years.
thinkorswim (if you're a Schwab client)
Schwab gives its clients the thinkorswim platform at no charge, and a Schwab account has no minimum deposit to open. If you're not a client, a 30-day Guest Pass lets you try it without opening an account (it can't be extended, and your setup is lost when it ends).
- Backtesting: write the strategy in thinkScript, add it to a chart, then right-click a signal arrow and choose Show Report. The Strategy Report estimates the profit and loss of every buy and sell signal over the chart's loaded time frame.
- How far back: intraday charts load from the current day up to 360 days; daily charts can load all available history. So it's a year of intraday bars, like most brokers, but many years on daily bars.
- OnDemand replays the market from a trading day you pick, so you can practise trading a past session by hand in a simulated account. paperMoney gives Schwab clients $100,000 of virtual buying power for live practice.
- Futures: trading futures at Schwab needs a separate application and approval.
TradingView
TradingView's free Basic plan includes the Strategy Tester with its basic report, for strategies written in Pine Script.
- The catch is bar count. A chart loads a fixed number of historical bars per plan: 5,000 on Basic, 10,000 on Essential and Plus, 20,000 on Premium. On 5-minute futures bars, 5,000 bars is only a few weeks. On daily and higher timeframes, every plan shows all available data, so the free plan is genuinely useful for daily and swing strategies.
- Deep Backtesting tests on all available history (up to two million bars), but needs the Premium plan or higher.
- Bar Replay on daily and higher bars is on every plan; replaying minute bars needs Essential or higher.
- CME futures data is free with a 10-minute delay. That doesn't matter for backtesting, which is all past data.
NinjaTrader
NinjaTrader's desktop platform is free to use for charting, strategy backtesting and simulated trading, with no time limit and no deposit; you register for an account to download it. Strategies are written in NinjaScript (C#) and backtested in the Strategy Analyzer on NinjaTrader's own historical futures data.
- NinjaTrader doesn't publish how far back its free historical data goes, and says it can change, so check the range you get for your market before relying on it.
- Live data is a 14-day trial; after that, simulation runs on delayed data unless you subscribe or fund an account. Some extras, like Market Replay, may need a funded account (NinjaTrader's own pages differ on this).
QuantConnect
If you're comfortable coding in Python or C#, QuantConnect's free tier is the deepest free option: unlimited cloud backtests (one at a time, up to 200 a day, with a short wait before each starts) on its own historical data.
- Free data in the cloud, from minute to daily resolution: US stocks from 1998 (including companies that later delisted, so results aren't flattered by survivors), the most liquid US futures from 2009, and Coinbase crypto from 2015.
- Paid for: second and tick data, live and paper trading, the local command-line tools, and downloading the data to your own computer.
- It's a coding platform, not a chart, so the learning curve is steeper. The payoff is 15 years or more of minute bars for a futures strategy.
Other routes
- MetaTrader 5's Strategy Tester is free in the terminal and downloads history from your broker's server, so how far back it goes depends on the broker.
- Your own data and a free library. Python libraries such as backtesting.py run on any price history you can get, which is the hard part: free sources tend to be daily bars only.
Testing on more history, without fooling yourself
More history only helps if you use it honestly. Four habits matter more than which tool you pick:
- Keep data back. Settle the rules and settings on an earlier stretch, then run them once on a later stretch you haven't looked at (out of sample). If the later stretch falls apart, the first result was fitted.
- Walk it forward. Repeat that split through time: tune on one window, test on the next, move both forward, and judge the strategy only on the joined-up test windows. A strategy that needs new settings every window has no stable edge.
- Get enough trades, in different markets. Dozens at the least, ideally hundreds, spread over trending, choppy, quiet and fast periods. A long history with a handful of trades proves little.
- Charge real costs. Commissions on every round trip and a tick or two of slippage on market and stop orders. Many intraday edges disappear once both are in.
When the longer test agrees, forward test: run the unchanged strategy on new sessions as they happen. That's the one test no amount of tuning can fit. Our guide to backtesting without fooling yourself goes through these in more detail.
What TradeHarbor's backtester does well, and where it stops
We'd rather you knew both. What it does well:
- Your broker's own prices. It runs on the history your broker serves through your own API connection, the same feed your automations trade on, so the data you test on is the data you'll trade on.
- The rules you'll trade under. The Prop check replays a run under a firm's plan, with its drawdown, daily loss limit and profit target, and says when it would have failed or passed. A strategy can be profitable and still bust an evaluation.
- One engine from test to live. Your own strategies run on the same engine in the backtest and in automations, and a saved backtest can be forward tested on every new session.
- Honest checks built in. An out-of-sample split, a stress test, a slippage curve, and costs on every trade.
Where it stops: it only has the history your broker serves through its API, and we don't keep a store of long history ourselves. For futures through ProjectX that's a year at most: 60 days of 1-minute bars, 180 days of 5-minute bars and a year of 15-minute bars. Alpaca serves coin pairs and US stocks further back (up to 3 years of 15-minute coin bars and 5 years of 15-minute stock bars). A year of one market can all be one regime, so treat a good result in the Backtester as a first check, and confirm it on a longer history with one of the tools above before you trade it.
Next steps
Once a strategy holds up on a long history and in forward testing, how trading APIs work explains how a program talks to your broker, and how automated trading works follows one trade from signal to fill. Past results, on any amount of history, don't guarantee future ones.