The price of gold relative to silver has swung dramatically over the past 50 years, ranging from as low as 17x at the 1980 silver bubble peak to as high as 119 at the height of the Covid-driven liquidity crisis. Short-term movements in the ratio are extremely difficult to predict, but the technical picture strongly favors silver outperformance. Over the long term, the high substitutability of the metals suggests the ratio should mean revert in silver's favor, allowing it to outperform by a few percent a year.
Looking at the charts of gold and silver, silver looks like the better bet. While the two metals are highly correlated, silver has much more support around current levels and has bounced off support at around $22 level to post what looks like a higher low. In contrast, gold has posted a series of lower highs after failing to hold onto the $2,000 level, and no meaningful support is seen until around $1,800.
Silver also looks the most likely to take out key resistance which currently comes in at around $25 marking the down trendline from the 2021 peak. While such a move would likely also require gold to move higher, we should see silver outperform. The chart below shows the silver price alongside the silver/gold ratio, clearly showing that silver breakouts tend to drive outperformance despite the long-term trend of underperformance.
Theoretically, gold should outperform during periods of low interest rates as it is a closer substitute for fiat currency. However, as silver has much higher volatility it tends to outperform gold during bull markets. Silver should theoretically outperform during periods of strong economic growth as is much more commonly used in industry, get there is little evidence in the data. From 2008 to 2020 the gold/silver ratio was highly negatively correlated with industrial prices as one would expect, but this correlation has broken down since, which is surprising given the ongoing boom in semiconductor demand.
It is not at all clear to me why silver is trading at such low levels relative to the price of gold. Rising gold holdings by central banks may explain part of it, but central bank gold holdings have had little bearing on the gold/silver ratio in the past. As the two metals are extremely close substitutes, the ratio has a tendency to mean revert. In fact, this mean reversion has historically been extremely reliable. As the chart below shows, the gold-silver ratio is closely correlated with subsequent gold underperformance over the following decade, with an R-squared of 0.77 over the past 50 years. The current ratio is consistent with annual gold underperformance of around 3% annually over the next 10 years.
As silver is more volatile than gold, anyone looking to play the ratio should keep in mind that in the short term, it is highly likely to be driven by the silver price. Investors may therefore wish to short more gold relative to their silver longs in nominal terms if they are less bullish on silver. The following chart shows the performance of the gold/silver ratio rebased to 2003 alongside the performance of the ratio if we reduce silver's weekly volatility by half to match that of gold.
This effectively shows the performance seen if investors had held a long position on silver and a notional short position on gold that is two times larger. As you can see, the reduction of silver's volatility has actually worsened its relative outperformance which I believe highlights silver's relative undervaluation.