The Gold/Silver Ratio: How to Read It and How a Gold-Silver Spread Works

The gold/silver ratio is the price of one ounce of gold divided by the price of one ounce of silver. It tells you how many ounces of silver it takes to buy one ounce of gold. Traders use it to judge whether silver is cheap or expensive relative to gold, and some open a two-legged spread to bet on the ratio changing. This article covers the mechanics, how to read the ratio, why it does not have to return to its average, and works through a spread step by step.

How the ratio is calculated

Gold/silver ratio = gold price per ounce ÷ silver price per ounce

Throughout this article we use assumed prices, not market quotes or forecasts: gold at 3,400 dollars and silver at 40 dollars. The ratio = 3,400 ÷ 40 = 85, so 85 ounces of silver are worth one ounce of gold.

The ratio can change in three ways, and it pays to keep them apart:

• Silver rises faster than gold, or falls more slowly: the ratio falls
• Gold rises faster than silver, or falls more slowly: the ratio rises
• Both move by the same percentage: the ratio does not change at all, however far prices travel

So the ratio is a measure of relative value. It does not say whether gold or silver is cheap or expensive in dollars. The ratio can fall while both metals rally or while both crash.

Why silver is more volatile than gold

Gold and silver usually move in the same direction, but silver tends to swing harder both ways. There are three main reasons:

1. A much smaller market. The silver market is many times smaller than gold’s, so the same amount of money flowing in or out moves silver’s price more.
2. Much of the demand is industrial. A large share of silver goes into electronics, solar panels and other industrial uses, so its price is tied to the economic cycle too. Gold’s big holders are investors and central banks, who tend to treat it as a safe haven.
3. Supply reacts slowly to price. Most silver is mined as a by-product of lead, zinc, copper and gold mines, so a higher silver price doesn’t quickly bring new silver mines.

On the chart this means that when markets panic (see risk-on, risk-off), silver often falls with industrial metals and harder than gold, so the ratio tends to jump. When the economy is busy and investors are happy to take risk, silver has room to outrun gold and the ratio falls. These are general tendencies, not fixed rules.

Reading a high or a low ratio

• A high ratio compared with the past = silver is cheap relative to gold. People who expect the ratio to fall see silver as having the edge over gold.
• A low ratio compared with the past = silver is expensive relative to gold. People who expect the ratio to rise take the opposite view.

“High” and “low” always need a reference. A common one is a moving average of the ratio with a band at ± 2 standard deviations (σ), the same idea as Bollinger Bands. Or you compute a z-score = (today’s ratio − average) ÷ σ. A z-score above +2 means the ratio is higher than usual for the period you measured.

The catch is that the answer depends on how far back you measure. Different look-back windows give different answers, and the ratio has moved over a very wide range in the past. A level that looked high in one era can be normal in another.

Why “the ratio will return to its average” is not guaranteed

The core idea behind trading the ratio is mean reversion: once it swings far enough, it comes back to the average. But nothing anchors the gold/silver ratio.

• No law or peg holds it. In the 19th century many countries used both gold and silver as monetary standards and fixed the exchange rate between them at around 15–16 to 1. When countries dropped silver as a monetary standard, that anchor disappeared. Since then the ratio has simply been the result of supply and demand in two separate markets.
• Market structure changes. New industrial demand for silver, central-bank gold buying, or funds that hold physical metal can shift the “normal level” of the ratio for years. This is a market regime change.
• A ±2σ band moves with it. The average and σ come from past data. If the normal level moves, the band follows it, and the point that looks “too high” may just be the start of a new level.

The ratio’s average is not a magnet. It is just a summary of the past. If the normal level shifts, an “unusually high” signal walks you into a trade that keeps losing.

To test whether a pair of prices really tends to revert statistically, see pairs trading and cointegration, which covers the tests and their limits.

How a ratio spread works

If you expect the ratio to fall, you buy silver and sell gold at the same time. If you expect it to rise, you do the reverse: sell silver, buy gold. Both legs are opened together and closed together.

The aim is to strip out as much of the “whole metals complex moving” as possible and keep only the effect of the ratio changing. There are two ways to size it:

• By value (dollar-neutral). Both legs have the same value when opened. This makes the profit and loss track the ratio most closely.
• By volatility (volatility-neutral). The silver leg is smaller in value because silver swings harder, so both legs carry similar risk. But the P&L no longer follows the ratio exactly.

Check contract sizes before you calculate

At many brokers, 1 lot of gold (XAUUSD) = 100 ounces and 1 lot of silver (XAGUSD) = 5,000 ounces, matching the standard futures contracts. Some use other sizes, for example 1,000 ounces of silver per lot. Before calculating, right-click the symbol in MT5’s Market Watch, choose Specification and read the Contract size of both symbols. Get it wrong and your position size is off by a multiple. The example below uses 100 and 5,000 ounces.

Example: long silver, short gold, opened at a ratio of 85

Assumed prices: gold 3,400, silver 40, ratio 85. You expect the ratio to fall, so you buy silver and sell gold with equal value on each side.

1. Sell 0.10 lot of gold = 0.10 × 100 = 10 ounces, worth 10 × 3,400 = 34,000 dollars
2. Silver needs the same value: 34,000 ÷ 40 = 850 ounces
3. In lots: 850 ÷ 5,000 = 0.17 lot (an exact fit, worth 850 × 40 = 34,000 dollars)

Later the ratio falls from 85 to 80. Here are three ways that could happen:

Scenario	Gold	Silver	Silver leg (long 850 oz)	Gold leg (short 10 oz)	Total
A. Gold flat, silver up	3,400	42.50	850 × 2.50 = +2,125	0	+2,125
B. Silver flat, gold down	3,200	40.00	0	10 × 200 = +2,000	+2,000
C. Both fall, gold −10%	3,060	38.25	850 × (−1.75) = −1,487.50	10 × 340 = +3,400	+1,912.50

Check: 3,400 ÷ 42.50 = 80, 3,200 ÷ 40 = 80 and 3,060 ÷ 38.25 = 80. Every row closes at a ratio of 80.

Look at row C: silver’s price fell, yet the spread still made money because gold fell more. That is the heart of a spread. You are betting on the ratio, not on the direction of prices.

Why do the three rows differ? For an equal-value spread:

P&L = value per leg × (gold price at close ÷ gold price at open) × (ratio at open ÷ ratio at close − 1)

Row A: 34,000 × 1 × (85 ÷ 80 − 1) = 34,000 × 0.0625 = 2,125, matching the table. Row C: 34,000 × 0.9 × 0.0625 = 1,912.50, also a match. The ratio fell 5.9% but the spread made 6.25% of one leg’s value, and the dollar amount still depends on where gold is when you close.

If you are wrong and the ratio rises from 85 to 90 with gold flat, silver drops to 3,400 ÷ 90 = 37.78 and you lose 34,000 × (85 ÷ 90 − 1) = −1,888.89 dollars. The same 5-point move loses slightly less than it would have made, because the formula divides.

Sizing by volatility instead

Assume gold’s daily volatility is 1.5% and silver’s is 2.4% (made-up figures; measure your own from real data, for example with ATR). The silver leg is then worth 34,000 × 1.5 ÷ 2.4 = 21,250 dollars = 531.25 ounces = 0.106 lot, rounded to 0.11 lot (550 ounces).

Run scenario A again (gold flat, silver to 42.50): you make 550 × 2.50 = 1,375 dollars. Scenario B (silver flat, gold to 3,200) still makes 2,000 dollars. The ratio moved by the same amount but the results differ, which means this version still hides a bet on gold’s direction. Know that before you choose it.

Costs and risks the formula doesn’t show

• Spread on two legs, when opening and when closing. Check the spread of both symbols on your own account (see spread cost per trade).
• Swap on two legs. The ratio moves slowly, so these spreads are often held for weeks or months, and the swap on both the long and the short leg can be negative (see swap and overnight fees).
• Margin on two legs. Most brokers charge margin per symbol and don’t net it the way they might for opposite positions in one symbol (see leverage, margin and stop out).
• Lot sizes that don’t round cleanly. The minimum 0.01 lot of silver is 50 ounces, or 2,000 dollars at a price of 40. If you sell only 0.01 lot of gold (worth 3,400), you would need 0.017 lot of silver, which you can’t open. Rounding to 0.02 lot gives 4,000 dollars, so the silver leg is 17.6% too heavy.
• The relationship isn’t constant. Gold and silver are highly correlated, but not perfectly. The stretches where they part ways are where spreads lose most (see correlation and portfolio risk).

Set the stop loss as a ratio

Pick the ratio level where you admit you’re wrong, then calculate what that costs. In the example, giving up at 90 costs about 1,889 dollars (with gold flat). To keep the trade within 1% of the account, the account would need about 1,889 ÷ 0.01 = 188,889 dollars. With a smaller account you have to shrink both legs, and then you run into the rounding problem above.

Our test: silver is not cheap gold

Many people assume silver follows gold anyway, so a system that works on gold should work on silver. In our backtest, we took one trading system developed on gold and ran it on 12 symbols in the MT5 Strategy Tester with real tick data from January to September 2026, adjusting lot size and the maximum allowed spread for each symbol.

Symbol	Profit factor	Maximum equity drawdown	Trades
Gold (XAUUSD)	2.01	4.79%	1,907
Silver (XAGUSD)	0.96	41.25%	2,346

Profit factor is gross profit divided by gross loss; below 1 means a net loss. On silver the same system lost money overall, and its drawdown was almost 9 times deeper than on gold. Part of the reason is that silver swings differently from gold, in both size and timing. Part may be that the rules were tuned to gold specifically (see overfitting). This is one short backtest period, and past results do not predict future results. But it does show you shouldn’t assume silver behaves like gold without testing it.

The ratio alongside other information

The ratio is only one angle. Both metals are priced in dollars, so interest rates and the dollar affect both (see how interest rates move gold and currencies and the dollar index DXY). If gold is your main market, the gold XAUUSD trading guide covers the basics, and gold CFDs, futures, spot and ETFs explains how the different forms of gold differ.

Common mistakes

1. Reading a high ratio as “silver will rise”. The ratio can fall because gold drops while silver drops too. Someone who only bought silver loses money despite calling the ratio correctly.
2. Opening equal lots on both legs. Selling 0.10 lot of gold (34,000 dollars) against buying 0.10 lot of silver (500 ounces = 20,000 dollars) leaves the gold side 14,000 dollars heavier. That is a bet on gold falling, not a bet on the ratio.
3. Not checking contract size. Silver at 1,000 ounces per lot versus 5,000 ounces per lot makes a fivefold difference in position size.
4. Believing the average must come back, then holding with no exit point, or adding as the ratio runs away, which works like averaging down.
5. Forgetting swap on two legs. A spread held for months can pay enough swap to eat the whole gain from the ratio.
6. Moving a gold system to silver without testing. The test above shows how different the results can be.