- Open the simulator and pick XAG/USD from the instrument list - or gold, platinum, palladium, copper and the rest of the 105.
- Choose a start date and timeframe. The chart hides everything after it, so you read the same bars a trader had at the time rather than a chart you already know the answer to.
- Replay bar by bar and place trades with the mouse. Drag the stop and target on the chart and the lot size follows from your risk percent.
- Read the report. Win rate, expectancy, profit factor, Sharpe and drawdown across the session, saved to your trading journal so the sample accumulates.
Silver · XAG/USD
Backtest silver free, in your browser.
Replay XAG/USD bar by bar from 2011 to this week, place trades with the mouse, and get win rate, expectancy, profit factor and drawdown at the end of the session. No download, no broker account, no card - silver is unlocked on the free plan along with all 105 instruments.
The data
Fifteen years of silver, through the 2026 spike.
Tick-derived OHLC bars from the same archive our published research is computed on. Lower timeframes cover shorter windows because the underlying data is far denser - the table is exactly what you can replay today.
| Timeframe | History starts | Roughly | Plan |
|---|---|---|---|
| D1 / W1 / MN | February 2011 | 16 years, 4,000+ bars | Free |
| H4 | February 2011 | 16 years, 24,000+ bars | Free |
| H1 | February 2011 | 16 years, 92,000+ bars | Free |
| M30 | February 2018 | 9 years, 100,000+ bars | Free |
| M15 | May 2022 | 4 years, 100,000+ bars | Free |
| M5 | March 2025 | 19 months, 100,000+ bars | Free |
| M1 | April 2026 | 5 months, 137,000+ bars | Free |
Coverage as of October 2026 and growing nightly. One lot is 5,000 troy ounces and prices are quoted to a tenth of a cent, so a one-cent move on one lot is worth $50.
Measured
Silver moves about twice as far as gold - and 2026 broke the scale.
Median XAG/USD daily range as a percentage of price, against gold over the same days, every trading day from 2014 to September 2026.
| Year | Silver: median daily range | Gold: median daily range | Silver ÷ gold |
|---|---|---|---|
| 2014 | 1.89% | 1.11% | 1.70× |
| 2015 | 2.13% | 1.24% | 1.72× |
| 2016 | 2.11% | 1.25% | 1.69× |
| 2017 | 1.55% | 0.94% | 1.65× |
| 2018 | 1.46% | 0.87% | 1.68× |
| 2019 | 1.51% | 0.94% | 1.61× |
| 2020 | 2.85% | 1.36% | 2.10× |
| 2021 | 2.24% | 1.16% | 1.93× |
| 2022 | 2.63% | 1.25% | 2.10× |
| 2023 | 2.23% | 1.07% | 2.08× |
| 2024 | 2.60% | 1.22% | 2.13× |
| 2025 | 2.39% | 1.49% | 1.60× |
| 2026 (to Sep) | 4.92% | 2.21% | 2.23× |
| All years | 2.26% | 1.23% | 1.84× |
4,049 XAG/USD daily bars, 2011-02-17 to 2026-09-25, FxBacktest archive, weekend stubs excluded; the all-years row covers the whole archive. Range is high minus low divided by the open.
In every year since 2014 the typical silver day has covered between 1.6 and 2.3 times as much ground as the typical gold day. Over the whole archive the median is 2.26% of price for silver against 1.23% for gold. A gold strategy moved to silver unchanged takes on roughly double the risk per trade at the same lot size - re-derive the stop from silver's own range instead of copying gold's.
2026 is a regime of its own. The median silver day has ranged 4.92% of price this year, against 1.46% to 2.85% in every year from 2017 to 2025, and the single widest day in fifteen years came on 30 January 2026, at 38.31% from high to low. A backtest that runs from 2018 to 2025 and stops there is describing a calmer market than the one you would trade today - and one that includes 2026 can be dominated by it. Report the years separately.
When Silver movesmedian hourly range by 3-hour block, UTC, 2021-2026
33,916 Silver hourly bars, 2021-01-03 to 2026-09-25, FxBacktest archive. Each block is the average of its three hourly medians. Values are in cents per ounce. The 12:00-15:00 block holds the US data releases and the COMEX open.
Before you test
Three things that break a gold strategy on silver.
Each links to a longer read.
Guide
Gold is not silver
Gold and silver share a chart shape and a driver, but not a volatility. Read how the metals behave against the currency pairs before you port a setup across.
Read the guide Original researchSize in percent, not cents
Silver more than quadrupled in price between 2019 and 2026, so a stop measured in cents means something different in every year. The range study compares markets as a share of price.
Read the study GuideWide days arrive in clusters
Three of silver's six widest days on record fell in the first week of February 2026. Check risk of ruin before win rate, and size so a cluster is survivable.
Read the guideHow it works
Four steps, about a minute.
More markets with their own page: Nasdaq · Dow 30 · EUR/USD · Bitcoin · GBP/USD · USD/JPY · S&P 500 · Ethereum · Crude oil · DAX 40 · GBP/JPY · AUD/USD. All 105 are on the backtest hub.
Questions
Silver backtesting FAQ
Is silver backtesting really free here?
Yes. XAG/USD and all 105 instruments are unlocked on the free plan, with every timeframe, full trade execution and the session report. Nothing to download, no card. Paid plans add more saved sessions, more simultaneous indicators and deeper analytics - not the silver chart or its history.
How much silver history can I replay?
Daily, 4-hour and hourly bars go back to February 2011. 30-minute data starts in February 2018, 15-minute in May 2022, 5-minute in March 2025 and 1-minute in April 2026.
Is silver more volatile than gold?
Yes, by about two times. Across the archive the median silver day ranges 2.26% of price against 1.23% for gold, and the ratio has stayed between 1.6 and 2.3 in every year since 2012.
Can I backtest other metals?
Yes. Gold (XAU/USD and XAU/EUR), platinum, palladium and copper are on the same free plan with the same replay engine and report, alongside forex, indices, energy and stocks - all listed on the backtest hub.
Risk disclaimerTrading foreign exchange, CFDs, cryptocurrencies and other leveraged products carries a high level of risk and is not suitable for every investor - losses can exceed your deposits. Everything on this page is educational content, not financial advice. Backtest and simulator results are hypothetical: they do not represent live trading and past performance does not guarantee future results.