Silver · Yahoo Finance
Yahoo Finance - Silver Futures (SI=F) Historical Data
Datadory delivers yahoo finance silver futures si f historical data data covering the front-month COMEX silver contract as typed rows: session OHLCV bars from one-minute intervals out to the contract's multi-decade daily record, a live quote block with day range, last price and volume, settlement dating for the active month, and 52-week extremes spanning 38.51 to 121.30 US dollars per troy ounce. Delivered daily, weekly, or hourly.
API, files, or your warehouse. Daily, weekly, or hourly.
- Where it covers
- One instrument, globally priced: the COMEX (New York) front-month silver futures benchmark quoted in US dollars per troy ounce - no country dimension to manage
- How far back
- Two clocks: intraday bars at one-minute, five-minute, fifteen-minute and one-hour intervals reaching back roughly 60 days, and daily bars running across the contract's multi-decade history
- How fine
- Bar-level OHLCV at intraday-to-daily grain plus a point-in-time quote block - one row per interval, one quote snapshot per pull, nothing below bar level (no ticks)
What is the Yahoo Finance Silver Futures (SI=F) historical data dataset?
It is the default silver chart on the consumer web, rebuilt as analysis-ready rows. The quote page for Silver Sep 26 (SI=F) follows the front-month COMEX silver futures contract - the instrument hedgers transact and the number headlines reach for when white metal moves. Everything the page knows about it lives in two layers.
The quote block holds the current session: open at 68.29, a day's range running 67.96 to 70.08, last printed at 68.11, an offer standing at 69.49, volume displayed as 38.31k contracts - plus trailing performance ranges from one day out to the full series. The bar-history layer serves proper OHLCV: one-minute through one-hour intervals reaching back roughly sixty days, and daily bars stretching across the contract's multi-decade life.
What separates this record from an anonymous price feed is that the contract mechanics arrive as columns rather than captions. shortName names the active month, settlementDate (2026-09-28) fixes when it expires, chartPreviousClose anchors the window, and fiftyTwoWeekHigh against fiftyTwoWeekLow - 121.30 against 38.51 - frames the whole year's argument in two numbers. Datadory scores the record 6 out of 10: indicator breadth is narrow by design, and the value lies in turning the world's most-opened silver screen into a table. Get a sample of this dataset with your date windows named, and it comes back cut to them.
What do the sample rows look like?
Observed during the August 2026 research pass, exactly as the two surfaces report them:
# quote block - one pull, current session
symbol : SI=F
shortName : Silver Sep 26
settlementDate : 2026-09-28 # the active front month, dated
open : 68.29 # USD per troy ounce
dayLow : 67.96 dayHigh : 70.08
lastPrice : 68.11 ask : 69.49
volume : 38.31k # contracts, display-rounded
# bar history - one row per interval
timestamp : 1787284800 # unix epoch seconds, normalized to ISO on delivery
open=68.29 high=70.08 low=67.96 close=69.365 volume=38310
# shape of the series
fiftyTwoWeekHigh : 121.30 fiftyTwoWeekLow : 38.51
chartPreviousClose: 68.105 priceHint : 2 # decimals renderedThree things those lines settle. First, the range is the story: 67.96 to 70.08 spans 2.12 dollars, roughly 3.1 percent of the open - a session with genuine motion in it, not a placeholder print. Second, the surfaces disagree politely: last prints at 68.11 while the window's prior close anchors at 68.105 and the offer stands at 69.49, because quote and bar captures land minutes apart; anchor each calculation on one layer and treat cross-surface arithmetic as approximate. Third, volume arrives twice - display-rounded on the quote side, raw contract counts on the bar side - so dashboards keep their readable string while models keep their integer, and neither has to guess.
What fields does the dataset include?
Fourteen documented fields form the delivered core, split cleanly across the two surfaces. The bar half is textbook OHLCV: epoch timestamp, then open, high, low, close and volume, all denominated in US dollars per troy ounce except the contract count. The quote half carries identity and mechanics - symbol and shortName, settlementDate, ask, chartPreviousClose and the day's-range pair, with priceHint quietly declaring how many decimals everything renders with. fiftyTwoWeekHigh and fiftyTwoWeekLow stand slightly apart as the year's frame.
Because the upstream surface publishes no data dictionary, every definition above was written against observed payloads during the August 2026 pass and carries a worked example from the same capture. Anything adjacent folds under additional fields on request rather than being promised blind: the performance-range percentages, the completed two-sided quote block, series metadata such as first trade date, and derived builds - a roll-spliced continuous series, annualized volatility, gold-contract pairing for ratio work. Name the ones your analysis needs and they ship as extra columns in the same delivery.
Where does coverage run across geography, time and granularity?
Geography - one instrument, globally priced. SI=F is the COMEX (New York) front-month silver benchmark quoted in US dollars per troy ounce; there is no country dimension and no currency ladder. For local-currency framings or auction benchmarks, the neighbours below cover those jobs.
Temporal - two clocks on one ticker. The intraday intervals (one-, five-, fifteen- and sixty-minute bars) reach back roughly sixty days, which is precisely the horizon an event study wants. The daily bars run across the contract's multi-decade history - enough tape for cycle-length backtests without stitching vendors together. The quote block rides on top as the current session's state, so a single pull yields history and present tense together.
Granularity - one row per interval on the bar side, one snapshot per pull on the quote side. Nothing sits below bar level: there are no individual ticks and no order-book depth beyond the quoted sides. That is the honest boundary of the record - breadth of time at bar grain, not microstructure.
How is the data delivered?
API, files, or your warehouse. Daily, weekly, or hourly.
Name the intervals, the date windows and the quote fields you need, and the sample returns shaped to them - sixty days of one-minute bars for the event study, the full daily run for the backtest, or both stacked on one instrument key.
Every delivery ships flattened rows with ISO-normalized dates beside the epoch timestamps, so joins happen on plain calendar keys instead of scaling arithmetic. Cadence changes are a settings conversation, not a re-integration project, and the field dictionary above travels unchanged through every channel.
Who uses this data, and for what?
- Quant and volatility desks - sixty days of minute bars bracket any recent macro week, feeding range-based volatility estimators that need highs and lows, not just closes; part of the shelf serving quant backtesting.
- Systematic traders - the multi-decade daily record backs cycle-length strategy tests on one consistent dollar-per-ounce convention, with
settlementDateavailable to build honest roll rules instead of pretending the front month never turns over. - Hedgers and treasury teams - the quote block is the reference leg when pricing forward physical purchases against the exchange-traded benchmark.
- Cross-metal analysts - identically keyed bars exist for the neighbouring gold contract, turning the gold-silver ratio into a two-table join; see price monitoring workflows.
- Dashboard and product teams - current level, day's range and distance from the 52-week extreme arrive precomputed, which is why the record drops into boards without a transformation layer.
- Educators and journalists - a chart every audience already recognizes, now citable with exact dates, ranges and volumes attached.
Which personas get the most value?
Ranked by relevance in Datadory's persona tagging:
- Investors & Quant Researchers (relevance 3) - intraday resolution around events with decades of daily bars behind them, one instrument key throughout - see investors & quants x silver.
- Data Scientists & ML Engineers (relevance 2) - typed OHLCV with normalized timestamps is clean feature stock for volatility and trend models - data scientists x silver.
- Developers & Data-Product Builders (relevance 2) - a flat, stable schema keyed on one symbol that caches and diffs cleanly - developers & builders x silver.
- Market Researchers & Consultants (relevance 1) - quotable extremes and session color for metals commentary - market researchers x silver.
- Competitive Intelligence & Product Teams (relevance 1) - the metal's level and range as an external yardstick for silver-exposed product decisions.
- Journalists, Academics & Students (relevance 1) - the familiar public face of silver pricing, with numbers precise enough to cite - journalists & academics x silver.
How does it compare within silver data?
The silver slice splits along three lines: exchange-traded versus over-the-counter, bar grain versus snapshot, and contract mechanics versus market fundamentals. Stooq XAGUSD owns dollar-pair daily bars on the spot side. Trading Economics' silver series owns the long daily reference reaching back to 1975. LBMA precious metals prices own the institutional auction benchmark the bullion market anchors to. SLV holdings track ETF vault tonnage, and the Silver Institute's supply-and-demand statistics run the industry's own deficit ledger. USGS silver data owns the ground truth of who mined what.
This record occupies the quadrant none of them touch: exchange-traded silver at bar grain. It is the only silver dataset in the slice carrying intraday intervals, and the only one where the instrument itself - active month, settlement date, expiry-driven volume shifts - is part of the data rather than a footnote. Pair it with Stooq for basis work, or with Trading Economics when a 1975 starting line matters more than intraday detail.
What should I know before requesting a sample?
Four things, stated upfront.
First, this is a futures price, not spot. SI=F tracks the exchange-traded front month, and the basis against over-the-counter benchmarks widens and narrows with storage, rates and delivery dynamics. Pick one convention and hold it across your whole history - mixing the two manufactures phantom signals.
Second, the front month rolls. shortName and settlementDate change identity as expiries pass, and naive concatenation inherits a jump at every hand-off. Ask for the roll-spliced build under additional fields and the seam is handled for you.
Third, volume arrives in two costumes - display-rounded strings on the quote surface, raw contract counts on the bar surface. Deliveries keep both labeled, so no dashboard ever charts "38.31" as thirty-eight contracts.
Fourth, intraday depth is bounded at roughly sixty days. Deep intraday questions older than that need scoping at sampling rather than discovery later. Start at Get a sample of this dataset with your windows and intervals named; the cadence decision comes after the sample validates.
Field dictionary
Every field below is documented against real records. The full dictionary ships with the sample.
| field | type | definition | example |
|---|---|---|---|
symbol | string | Instrument key for the front-month COMEX silver contract - the stable handle every delivered row answers to. | SI=F |
shortName | string | Display name of the currently active contract month - the fastest visible signal that the front month has rolled. | Silver Sep 26 |
settlementDate | date | Expiry and settlement date of the quoted contract - the calendar anchor behind every roll-aware pipeline. | 2026-09-28 |
open | number | Session opening price of the contract in US dollars per troy ounce. | 68.29 |
high | number | Session high price in US dollars per troy ounce. | 70.08 |
low | number | Session low price in US dollars per troy ounce. | 67.96 |
close | number | Session close, or latest traded price, in US dollars per troy ounce. | 69.365 |
volume | integer | Contracts traded in the session - raw counts on the bar surface, display-rounded strings on the quote surface. | 38310 |
timestamp | integer | Unix epoch seconds marking each bar's interval start; deliveries carry ISO-normalized dates alongside. | 1787284800 |
ask | number | Current ask quote in the quote block - the offered side of the live market. | 69.49 |
chartPreviousClose | number | Prior close anchoring the requested window - the baseline behind change arithmetic on the chart layer. | 68.105 |
fiftyTwoWeekHigh | number | Rolling 52-week high of the continuous front-month series in US dollars per troy ounce. | 121.30 |
fiftyTwoWeekLow | number | Rolling 52-week low of the same continuous series in US dollars per troy ounce. | 38.51 |
priceHint | integer | Decimal-precision hint governing how many places prices render with downstream. | 2 |
regularMarketDayHigh / regularMarketDayLow | number | Day's-range endpoints as embedded in the quote payload - the same extremes the bar surface reports, restated for the session. | 70.08 |
Additional fields on request | - | Performance ranges across the 1D through All horizons, the bid leg and prior-settlement pair, series metadata (currency, exchange, first trade date), a roll-spliced continuous series, and derived volatility or gold-ratio columns computed on delivery. | on request |
What teams do with it
- Event-window volatility work Roughly sixty days of one-minute bars put any recent announcement week under a microscope without a tick-data engagement - minute resolution is usually enough to see whether silver led or followed.
- Multi-decade backtesting Daily OHLCV across the contract's full history gives strategy research one consistent dollar-per-ounce convention, with highs and lows intact for range-based estimators that closes-only feeds cannot support.
- Roll and liquidity monitoring `settlementDate` crossed against volume shows the front month thinning ahead of expiry - the practical early warning behind any roll policy or contract-switch rule.
- Gold-silver ratio construction Same-shaped bars exist for the gold contract on the same publisher's rails, so the ratio computes as a join between two identically keyed tables rather than a spreadsheet archaeology project.
- Dashboards and alerting The quote block plus the 52-week band drop straight into executive boards: current level, day's range, distance from the yearly extreme - three numbers every metals stakeholder actually asks about.
- Teaching and demonstration The most recognized silver screen on the web doubles as classroom material - students meet a familiar chart and discover the same rows carry a machine-readable spine underneath.
Questions buyers ask
What does yahoo finance silver futures si f historical data data include?
Two layers on one instrument. The bar history carries session OHLCV - open, high, low, close and volume in contracts - from one-minute through one-hour intervals back roughly sixty days, plus daily bars across the contract's multi-decade record. The quote block carries the current session state for the active front month, including day's range, last price and volume, alongside identity fields such as the settlement date and the 52-week band.
How far back does the history go?
Daily bars run across the contract's multi-decade history, deep enough for cycle-length backtests on one consistent dollar-per-ounce convention. Intraday intervals are shallower by design, reaching back roughly sixty days - the window an event study actually interrogates. State the oldest date your analysis needs and the earliest available bar comes back confirmed with the sample rather than assumed.
Is this spot silver or a futures price?
A futures price. SI=F tracks the front-month COMEX silver contract quoted in US dollars per troy ounce, so it carries storage, rates and delivery dynamics that over-the-counter spot benchmarks do not. Expect small persistent divergences from spot series - that basis is information rather than noise, provided you hold one convention across your history.
What does the settlement date field tell me?
When the active contract expires. The captured row names Silver Sep 26 settling on 2026-09-28 - the month label and the actual settlement date do not coincide, which is exactly why the field earns a column. Crossing it against volume shows the front month thinning ahead of expiry, the practical input to any roll rule or contract-switch policy.
How precise are the numbers, and what units are they in?
Prices are US dollars per troy ounce, with a priceHint field declaring the rendered decimal count - two places in the captured row. Volume is contracts traded, arriving as raw integers on the bar surface and display-rounded strings such as 38.31k on the quote surface; deliveries keep both conventions labeled so no pipeline mistakes a display string for a count.
Does the dataset contain intraday bars?
Yes - one-, five-, fifteen- and sixty-minute OHLCV bars reaching back roughly sixty days, which covers the recent-event window most volatility and announcement studies want. Deeper intraday history than that sits outside the record's design; daily bars carry the long arc instead.
Can a sample be scoped to specific dates, intervals or fields?
Yes. Name the date windows, the bar intervals and any folded fields - performance ranges, the roll-spliced continuous series, derived volatility columns - and the sample returns in exactly the fourteen-field shape documented above, with extras added as columns rather than a different schema.
Notes on this record
- Two surfaces, one key Quote block and bar history describe the same instrument at different grains. Anchor each calculation on one surface - they are captured minutes apart, so polite disagreement between last, close and ask is expected, not corruption.
- Rolls are data here `shortName` and `settlementDate` turn contract rotation from folklore into columns. Any multi-year backtest built on the front month needs a roll rule, and this record supplies the inputs to write one honestly.
- Volume speaks two dialects Display-rounded on the quote surface, raw contract counts on the bar surface. The delivered table labels both, so dashboards stay readable while models keep exact integers.
- The 52-week band does the framing At 68.11 against a 38.51-to-121.30 band, the contract sits roughly a third of the way up its yearly range - one arithmetic line that replaces a chapter of metals commentary.
Datasets that pair with this one
- Stooq - XAGUSD Silver Price Data Dollar-pair daily bars on the spot side - the counterpart for basis work against the exchange-traded benchmark.
- LBMA Precious Metals Prices The institutional auction benchmark the bullion market anchors to, silver at the noon fixing.
- iShares Silver Trust (SLV) Daily Holdings ETF vault tonnage day by day - the investor-demand bucket no exchange tape records.
- Best silver datasets The ranked shortlist across the vertical, this record included.
- Silver data hub Every cataloged silver dataset on one page, from price boards to federal fundamentals.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.