The Gold/Silver Ratio: How Traders Use It, and Its Limits

Short answer: the gold/silver ratio is gold price ÷ silver price, the number of silver ounces that buy one ounce of gold. Traders use it three ways: to judge whether one metal is stretched against the other, to build ratio trades (long one metal, short the other), and as a confirmation tool for gold moves. Our weekly data since 2006 has a median of 72.32, a low of 32.00 (2011) and a high of 125.89 (2020). Extreme readings did tend to come back, but slowly, and the ratio’s “normal” level shifted by about 20 points between decades. It is a context tool and a relative-value idea, not a timing signal.

In plain words: imagine two runners tied together by a long elastic rope. They usually run the same way, but the faster one (silver) keeps pulling ahead or falling behind. The ratio measures how stretched the rope is. A very stretched rope tends to pull back, but nobody can tell you when, and sometimes the rope is longer than you thought.

How this lesson differs from our other ratio pages. The gold/silver ratio tool shows today’s reading and its history. Our blog guide, the gold/silver ratio explained, covers what the ratio is and why it moves. This lesson is the trading version: what extreme readings actually led to, how a ratio trade is built, sized and exited, a real one played out from our data, and where the idea fails.

What the ratio measures

The ratio compares two prices that are both in dollars, so the dollar cancels out. What is left is relative performance:

• Ratio rising: gold is doing better than silver. That can mean gold up and silver flat, both down with silver falling more, or both up with gold rising more.
• Ratio falling: silver is doing better than gold.

That is the most common beginner mistake: a falling ratio does not mean gold is falling, and a rising ratio does not mean gold is rising. On a day when both metals fall, the ratio can rise sharply.

Why the two metals drift apart

Gold and silver share a lot: both are precious metals priced in dollars, both react to the dollar and to real interest rates, and both are bought as a store of value. But silver has a second life as an industrial metal used in electronics, solar panels and many other products. And its market is much smaller, so the same flow of money moves it further. The result:

• In fear and recessions, gold usually holds up better (it is the deeper, more trusted haven) and the ratio rises. The 2008 crisis and the March 2020 crash both pushed it sharply higher; 2020 set the record in our data.
• In strong economic upswings and speculative rallies, silver usually outruns gold and the ratio falls, as in 2010-2011 and in late 2025 to January 2026 in our data.

Our correlation trading lesson covers silver as one of gold’s confirmation markets.

A long history, and what our data covers

For much of history the ratio was set by law. The US Coinage Act of 1792 declared fifteen pounds of pure silver equal in value to one pound of pure gold, a ratio of 15 to 1, and several European countries used similar fixed ratios in the 1800s. When governments stopped fixing it, the ratio was free to move, and in the last century it ranged far above those old legal levels.

What our files hold. The site computes the ratio from COMEX gold (GC=F) and silver (SI=F) futures daily closes via Yahoo Finance. Our files contain 1044 weekly readings from November 2006 to 4 November 2026, and daily readings with full silver prices only from 23 October 2025. We don’t have reliable data before 2006, so this lesson’s numbers cover the last 20 years only, and the daily silver tests cover about one year.

The normal level moved

A ratio trader’s first question is “what is normal?” Our data gives an uncomfortable answer.

The average weekly reading was 60.0 from 2006 to 2015 and 81.7 from 2016 to 2025. A trader in 2016 who treated 60 as normal and sold gold against silver at 75 would have waited years, through readings above 100, before being proven even partly right. The ratio does tend to come back from extremes, but the centre it comes back to can move. Any “normal” figure depends on the period you pick. That is why we show percentiles and real outcomes rather than one magic number.

What extreme readings led to

We took every run of weekly readings at or above 90 (joining runs less than nine weeks apart) and looked at the reading 52 weeks after it started.

• There were 7 such episodes, all of them since 2019.
• A year later the ratio was lower in 6 of 7. In 2019 it was higher (it went on to the 2020 record). Several of the “lower” cases were only a few points lower.
• Readings at or below 50 happened in 5 episodes (four between 2006 and 2012, and January 2026). In all 4 of the 4 with a full year of data after them, the ratio was higher a year later.

Two honest caveats. First, seven and four episodes are tiny samples, and episodes overlap with big events (the 2020 crash, the 2011 silver spike), so they are not independent tests. Second, the ratio can stay extreme for months: the 2019-2020 run above 85 lasted more than a year. Even when the direction was right, the timing was slow and the path was rough. Readings in the top 15% or so of history (above 85, 185 of 1044 weeks, 18%) are a reason to look, not a reason to trade.

How traders use the ratio

1. Relative-value (ratio) trades

The classic ratio trade has two legs:

• Expect the ratio to fall (silver to outperform): long silver, short gold.
• Expect the ratio to rise (gold to outperform): long gold, short silver.

The trade profits from the gap between the two metals’ moves, not from the direction of metals in general. If both rise but silver rises faster, the long-silver/short-gold trade makes money. That is attractive in theory because much of the shared risk (the dollar, real yields) cancels out. In practice both legs cost money to hold, and silver’s extra volatility means the trade is rarely as “hedged” as it looks.

2. Metal switching for long-term holders

Some long-term holders of physical metal swap silver for gold when the ratio is low and gold for silver when it is high, aiming to grow their total ounces. This is slow, low-frequency use of the same idea. It has real costs (dealer spreads on physical metal are wide) and it depends on the ratio returning to a centre that may have moved.

3. Confirmation for gold trades

Gold traders watch silver as a second opinion:

• A gold breakout with silver breaking out too, and the ratio stable or falling, has broad support.
• A gold high that silver refuses to confirm, with the ratio jumping, is a warning that the move may be narrow or driven by fear.

Our SMT divergence lesson turns that idea into a setup and tests it.

4. A risk-appetite gauge

Because silver is half industrial, a sharply rising ratio often goes with risk-off markets, and a falling ratio with risk-on. It is one input among many, alongside equities and the dollar.

A real played-out example: one year, two extremes

Our daily silver data starts on 23 October 2025, which happened to be close to an extreme. Here is what a simple ratio plan would have done, with an entry at an extreme and an exit at the 20-year weekly median of 72.32. The plan was not optimised or tested; it is the most basic version of the idea, and the timing of trade A is partly luck (our data starts there).

Trade A: ratio high, long silver and short gold.

• Entry, 23 October 2025: ratio 85.51, in the top 15% of the 20-year weekly history. Gold (December futures) 4,145.6, silver 48.482. $10,000 long silver, $10,000 short gold.
• Exit, 9 December 2025 (47 days later): the ratio closed at 70.41, below the median. Gold 4,236.2, silver 60.169.
• Result: silver +24.1%, gold +1.3% (adjusted for the late-November contract switch). The silver leg made far more than the gold short lost: about +$2,282 before costs and financing.

What happened next is the real lesson. The ratio did not stop at the median. Silver kept running and on 26 January 2026 the ratio hit 44.16, its lowest in our daily file, with silver at 115.08 and gold at 5,082.5. The highest daily reading in the file had been 86.33 on 27 October 2025, so in three months the ratio had nearly halved. Then it reversed: by 3 November 2026 the ratio was back at 79.46, with silver 49.587 and gold 3,940.4. Holding trade A from 23 October 2025 to 3 November 2026 would have left only about +$1,121 (silver +2.3%, gold -8.9%). The exit rule mattered more than the entry.

Trade B: ratio low, long gold and short silver.

• Entry, 23 January 2026: ratio 49.34, below 50 for the first time since 2012. Gold 4,979.7, silver 100.925.
• Exit, 14 October 2026 (264 days later): the ratio first closed back above the median at 73.95. Gold 4,107.9, silver 55.550.
• Result: gold -20.3%, silver -45.0%. Both metals fell hard, but silver fell further, so long gold/short silver made about +$2,470 on $10,000 legs, before costs. On the way, the trade first went the wrong way: silver jumped to 115.08 three days after entry, a loss of roughly 14% on the short silver leg alone.

What the two trades teach:

1. Both extremes in the year came back to the median, as the 20-year episodes suggest they usually do.
2. One took 47 days, the other 264. Ratio trades need patience and capital that can sit through it.
3. The profit came almost entirely from silver’s moves (up 24% in trade A, down 45% in trade B). Silver is the engine of the ratio, and the risk.
4. Two trades in one year prove nothing about the future. They show the mechanics and the size of the swings.

Sizing the two legs

Over 259 trading days from 23 October 2025 to 3 November 2026, gold’s daily moves had a standard deviation of 1.77% and silver’s 4.29%: silver moved about 2.4 times as much. That has a big consequence: equal dollars is not equal risk.

• Dollar-neutral: the same dollar amount in each leg. Simple, and it is what trades A and B above used. But the silver leg carries most of the day-to-day risk, so the trade behaves partly like a plain silver position.
• Volatility-weighted: put fewer dollars into silver, scaled by the volatility ratio. With silver about 2.4 times as volatile, $10,000 of gold would be matched by roughly $4,100 of silver ($10,000 ÷ 2.4). This makes each leg’s daily swings similar, so the trade responds more to the ratio itself.

Worked example (hypothetical): gold $4,000, silver $50.00, ratio 80.00. A trader expects the ratio to fall and wants $10,000 of gold exposure.

• Short 2.5 ounces of gold ($10,000). On a CFD that is 0.025 lots at 100 oz per lot, which most platforms can’t do, so in practice 0.03 lots (3 oz, $12,000) or 0.02 lots (2 oz, $8,000).
• Dollar-neutral long silver: $10,000 ÷ $50 = 200 ounces.
• Volatility-weighted long silver: about $4,100 ÷ $50 = 82 ounces.
• If the ratio falls from 80 to 70 because silver rises to $57.14 and gold stays at $4,000, the dollar-neutral trade makes 200 × $7.14 = about $1,429, the weighted trade about $586. If instead gold rises 5% and silver 5%, the ratio is unchanged and both trades make roughly nothing on the gap (the dollar-neutral one is flat; the weighted one loses about $295, because its gold short is bigger in dollars than its silver long).

Contract sizes differ by product and platform (silver CFDs are often 5,000 ounces per lot), so check the specification and position size before trading. Also budget for costs on both legs: two spreads in, two out, and overnight financing on both.

Building the ratio on your own chart

You don’t need a special feed. Most charting platforms let you divide one symbol by another: on TradingView, for example, typing TVC:GOLD/TVC:SILVER draws the spot ratio. A few habits make the chart useful:

• Use daily or weekly candles. Intraday ratio charts are noisy and the two markets’ liquidity differs hour by hour.
• Mark the percentiles, not just the line. Draw horizontal lines at roughly the 10th, 50th and 90th percentiles of the period you trust (from our data: 51.53, 72.32 and 87.54 since 2006).
• Keep the components on screen. Put gold and silver in two panels below the ratio so every ratio move can be traced to its cause.
• Note the source. Spot ratios and futures ratios differ slightly, because each metal’s futures carry its own basis.

A ratio trading plan, step by step

A complete plan has the same parts as any other trading plan, applied to two legs:

1. Thesis. “The ratio is in the top 15% of its 20-year range during an economic upswing; I expect silver to outperform gold over the next six months.”
2. Trigger. Not the first extreme reading. For example, a weekly close below the previous week’s low after a peak, so the ratio has started to turn.
3. Instruments and size. Both legs on the same platform if possible, sized dollar-neutral or volatility-weighted, with the total risk written in dollars.
4. Exit for profit. A ratio level (for example the median), a time limit (for example six months), or both.
5. Exit for loss. A ratio level beyond the extreme (for example five points above the entry) where the thesis is wrong, plus a plan if one leg gaps.
6. Costs. Spread on both legs, and the overnight financing of both legs for the expected holding time, added up before entry.
7. Review. Log the trade in ratio terms and in dollars. Note which leg made or lost the money.

The plan matters because a ratio trade can look right for months while losing money, or look wrong for months before coming good, as trade B above did.

Where to trade each leg

Ratio trades can be built from several kinds of instruments, each with its own costs:

• Spot CFDs (XAUUSD and XAGUSD). Flexible sizes and both directions, but overnight financing on both legs every night. Check each metal’s lot size: they usually differ.
• Futures. Gold and silver futures on the same exchange settle daily and carry their own roll costs and calendars; our gold futures lesson explains the mechanics for gold.
• Physical metal or funds for long-term switching. No leverage and no financing, but wider dealer spreads; suited to slow metal switching rather than trading.

Whichever you use, make sure both legs can be opened and closed at the same time; a ratio trade with one leg filled and the other not is just a naked metal position.

Limits of the ratio

• No fixed mean. The 2006-2015 and 2016-2025 averages differ by about 20 points. A “cheap” reading by one decade’s standard was normal in the next.
• Small samples. Seven episodes above 90 and four below 50 with a year of follow-up. Any percentage from that is fragile.
• Slow and rough. Extremes can last a year or more, and the path back can include large moves against you.
• Silver’s own events. Industrial demand, mine supply and speculative squeezes move silver without moving gold. The January 2026 spike in our data is an example.
• Futures data quirks. The site’s ratio uses front-month futures closes. Contract switches in either metal can nudge a single day’s reading; we adjust gold for its switches where we measure changes, but our file doesn’t let us date silver’s switches.
• Costs on two legs. Every ratio trade pays twice: two spreads, two financing charges.
• It is not a gold forecast. The ratio can tell you which metal is stretched relative to the other, not where gold goes next.

Base rates in one place

• Typical range since 2006: the middle half of weekly readings sat between 61.10 and 81.62; the middle 80% between 51.53 and 87.54.
• High readings: at or above 90 in 7 episodes since 2019; lower a year later in 6, often by only a few points.
• Low readings: at or below 50 in 5 episodes; higher a year later in all 4 that had a full year of data.
• Silver’s extra movement: about 2.4 times gold’s daily volatility in the last year of data.
• Two real extremes in one year: back to the median in 47 days and 264 days.

Treat these as a description of 20 years of one data source, not a forecast. Re-check them on the ratio tool before using them.

Pro nuance

• Look at the components. A ratio move driven by gold rising is a different market from one driven by silver falling. Always chart both metals beside the ratio.
• Regime first. In a recession or a crisis, expect the ratio to rise; in a reflation or industrial boom, to fall. Fading the ratio against the regime is fighting the main driver.
• Use the ratio’s own trend. Some traders only take mean-reversion ratio trades after the ratio has turned (for example a lower weekly high after a peak), rather than at the first extreme reading. That costs some of the move but avoids standing in front of the 2020-style spikes.
• Think in ounces for long-term switching. The goal of a metal switch is more ounces, so measure results in ounces, after dealer spreads.
• Watch silver’s liquidity. Silver markets are thinner; spreads widen faster in stress, exactly when ratio extremes happen.

Common mistakes

• Reading the ratio as a gold signal. A falling ratio does not mean gold is falling.
• Treating one number as “normal”. Use percentiles and the regime, not a fixed mean.
• Sizing both legs by dollars without checking volatility. The silver leg will dominate.
• No exit plan. Trade A shows the exit rule mattered more than the entry.
• Ignoring two sets of costs and financing. Ratio trades can take months, and financing adds up on both legs.
• Assuming the past episodes are a law. Seven and four cases cannot carry that weight.

Quick checklist

1. What is the ratio now, and where does it sit in the 20-year percentiles?
2. Is the move driven by gold, by silver, or both? Have I charted both?
3. What regime are we in: fear, recession, reflation, boom?
4. Am I trading the ratio (two legs) or using it as context for a gold trade?
5. How have I sized each leg: dollars or volatility-weighted?
6. What is my exit: a ratio level, a time limit, or both, and my stop if the ratio keeps stretching?
7. Have I counted spreads and financing on both legs over the expected holding time?

Sources

• Coinage Act of 1792, section 11 (the US legal ratio of 15 to 1 between gold and silver)
• CME Group: Silver futures (SI) and Gold futures (GC) (the contracts behind the site’s ratio data)
• Our own data: the site’s gold/silver ratio file (weekly from 10 November 2006 and daily from 23 October 2025 to 4 November 2026, computed from COMEX GC=F and SI=F daily closes via Yahoo Finance), the site’s silver and gold daily closes, and the site’s 250-day gold-silver correlation (daily log returns, to 4 November 2026). Gold changes use closes adjusted for the daily contract switches; silver closes are unadjusted. Results before costs and financing.

Educational content only, not financial advice. Trading gold and silver with leverage carries a high risk of loss.

Frequently asked questions