Backtest any market, in your browser.
Forex, indices, commodities, stocks and crypto — replayed bar by bar with the chart hidden ahead of you, each sized in its own contract terms, all writing to one journal.
Pick a market.
Every symbol runs the same replay engine and produces the same report. What changes is the contract model — index points, dollars per ounce, shares per lot or standard pips — so position sizing stays honest to the thing you are testing.
US100 (Nasdaq 100)INDEX
US500 (S&P 500)INDEX
US30 (Dow Jones)INDEX
DXY (Dollar Index)INDEX
GER40 (DAX)INDEX
UK100 (FTSE 100)INDEX
FRA40 (CAC 40)INDEX
EU50 (Euro Stoxx 50)INDEX
JP225 (Nikkei 225)INDEX
HK40 (Hang Seng)INDEX
US2000 (Russell 2000)INDEX
AUS200 (ASX 200)INDEX
SUI20 (SMI 20)INDEX
ESP35 (IBEX 35)INDEX
ITA40 (FTSE MIB)INDEX
XAU/USD (Gold)COMMODITY
XAU/EUR (Gold in euro)COMMODITY
XAG/USD (Silver)COMMODITY
XPT/USD (Platinum)COMMODITY
XPD/USD (Palladium)COMMODITY
COPPER (Copper)COMMODITY
WTI (Crude Oil)COMMODITY
BRENT (Brent Crude)COMMODITY
COCOA (Cocoa)COMMODITY
SUGAR (Sugar)COMMODITY
COFFEE (Coffee)COMMODITY
SOYBEAN (Soybeans)COMMODITY
BTC/USD (Bitcoin)CRYPTO
ETH/USD (Ethereum)CRYPTO
LTC/USD (Litecoin)CRYPTO
ADA/USD (Cardano)CRYPTO
BCH/USD (Bitcoin Cash)CRYPTO
AAPL (Apple)STOCK
MSFT (Microsoft)STOCK
NVDA (Nvidia)STOCK
TSLA (Tesla)STOCK
AMZN (Amazon)STOCK
GOOG (Google)STOCK
AMD (Adv. Micro Devices)STOCK
AVGO (Broadcom)STOCK
NFLX (Netflix)STOCK
JPM (JPMorgan)STOCK
PLTR (Palantir)STOCK
META (Meta Platforms)STOCK
LLY (Eli Lilly)STOCK
BRK.B (Berkshire Hathaway)STOCK
Instrument list as configured in the live simulator, September 2026. History depth varies by market and timeframe; each instrument's own page states its coverage.
Contract terms, per asset class.
The replay is identical everywhere; the sizing model is not. This is what a lot means in each class, and it is why a strategy's risk does not transfer unchanged when you move it.
| Asset class | One lot | Smallest increment | Session |
|---|---|---|---|
| Forex majors & crosses | 100,000 units | 0.0001 | 24 hours, five days |
| Yen pairs | Scaled to land near $8 per pip | 0.01 | 24 hours, five days |
| Stock indices | 10 per point | 1 point | Near-24h, short daily break |
| Gold | 100 troy ounces | $0.01 | 24 hours, five days |
| Silver | 5,000 troy ounces | $0.001 | 24 hours, five days |
| Energy | 1,000 barrels | $0.01 | 24 hours, five days |
| US stocks | 100 shares | $0.01 ($1 per lot) | US cash session, gaps overnight |
| Crypto | 1 unit | $1.00 | Continuous, 24/7 |
Contract model as configured per instrument in the live simulator. Yen-quoted pairs use a scaled contract size so the constant pip model lands near the real per-pip value rather than a literal 100,000 units.
Reading a sample that spans markets.
Testing several markets only pays off if the results are comparable afterwards. Three things make that possible.
Market ranking
Edge Lab ranks every market you have tested by expectancy, profit factor and drawdown side by side, so the question stops being "did this work" and becomes "where does this work best".
Releases on the chart
The economic calendar marks each release at the bar it landed on and filters by currency and impact — so a news-avoidance rule can be tested on the markets that release actually moves, not assumed across all of them.
Expectancy in R
Because risk is set as a percentage and lot size derived from the stop, results land in R rather than raw currency. A stock session and a forex session become directly comparable despite completely different contract terms.
Why one hub, not one market.
Most traders find their edge is narrower than they thought the first time they run the same rules somewhere else. A breakout set that prints on the Nasdaq open falls apart on EUR/USD, because one has a session open that concentrates volatility and the other does not.
A mean-reversion rule that works on a forex cross gets destroyed by a gap on Tesla. A gold scalping model collapses when it meets Bitcoin's weekend.
Comparable, not anecdotal.
Because every session writes to the same journal in the same units, those failures are measurable rather than remembered. Expectancy compares across a currency pair, an index and a stock even though their contract terms differ completely.
So you can rank markets by how well your own rules actually travel — and stop testing the ones that do not. That is the honest answer to "which market should I backtest?": the one you intend to trade, and at least one more.
Backtesting FAQ
Which markets can I backtest?
All 93 — 43 forex pairs, 15 stock indices, 14 commodities, 16 US stocks and 5 crypto pairs, each with every timeframe from M1 to D1. No market is reserved for a paid plan; paid tiers raise saved sessions, simultaneous indicators and analytics depth instead.
Can I backtest stocks and crypto, or only forex?
Both. Eleven US stocks including Apple, Nvidia, Tesla, Microsoft and Palantir are modelled at 100 shares per lot, and five crypto pairs run a genuine 24/7 session rather than the five-day forex week. Mixing asset classes in one journal is the point — the same rule set rarely survives the move from EUR/USD to Bitcoin, and the report shows where it broke.
How far back does the history go?
It varies by market and timeframe. Daily bars go back furthest — gold to 2009, the forex majors to 2010, the US indices to 2013 — while the lowest timeframes cover shorter windows because the tick archive behind them is far denser. Gold carries roughly 25 months of M1, most other symbols under six. Each instrument page states its own coverage.
Is this the same engine on every market?
Yes, with per-market contract terms. Replay, execution and the report are identical everywhere, but contract size, price precision and pip size are set per instrument — a Nasdaq trade is sized in index points, gold in dollars per ounce, a forex pair in standard pips. Position size comes from your stop distance and risk percentage in each market's own units.
Do I need MetaTrader or a broker account?
Neither. The simulator runs in the browser on Windows, macOS, Linux, iPad and phone. Nothing to install, no data files to source, no broker account to open — the history is already loaded and you trade it with the mouse.
Can I save a session and come back to it?
Yes. A session saves its full state — market, timeframe, start date, every trade taken and the report — so a long backtest does not have to be finished in one sitting. Saved sessions accumulate into the journal and the analytics, which is where a sample across several markets becomes readable.
How is this different from TradingView bar replay?
Bar replay moves the chart; it does not model a trade. Here, positions carry real contract terms per market, lot size is derived from your stop distance and risk percentage, spread and commission are charged against the result, and the session ends in a 28-metric report rather than a number you tally yourself. The TradingView comparison guide goes through the differences in detail.
Pick a market and start.
All 93 markets, every timeframe, nothing to install.