As neighbors on the periodic table of elements, gold and silver also share a close connection in global commodity markets. Both metals are used in investment, jewelry, dentistry and industry. They also exhibit a strong correlation: +0.76, on a daily basis since the beginning of 2015.
In our research, we have been able to ascertain that only two factors consistently influence both metals: 1) Interest rate expectations, and 2) mining supply. The factors, however, have important differences. Changes in interest rate expectations exert a short-term, day-to-day influence that is exogenous to the metals market. Mining supply, however, has a long-term, year-to-year influence that is endogenous: Gold and silver prices influence mining supply, often with long lags, and mining supply in turn impacts the price of both metals. Like binary stars orbiting a common center of gravity, gold and silver exert a strong influence on one another.
Precious metals prices exert little to no influence on the U.S. Federal Reserve’s (Fed) monetary policy, but changes in market expectations for Fed rates have a strong impact on two of the four precious metals, which include platinum and palladium. If one compares the day-to-day correlation of the changes in the Fed Fund Futures rate (100 minus the Fed Fund Futures price) with the changes in precious metals prices, one finds a strong and persistently negative correlation for gold and silver. Since the Fed hiked rates in December 2015, those correlations have grown stronger.
Gold is more sensitive to changes in interest rate policy than silver. This is probably because gold is the truly precious of the two: It is widely used in investment and jewelry, which can also be seen as an investment. While silver is also used in investment and jewelry, it is more widely used in industrial and other applications. By contrast, Platinum Group Metals such as platinum and palladium show little response to changing expectations on Fed rate hikes (Figure 1). Platinum and palladium are considered rare and expensive industrial metals.
The reason why gold and silver prices react negatively to higher interest rates is simple. The metals do not pay interests during the duration they are held. As such, higher interest rates tend to make fiat currencies like the U.S. dollar appear more attractive to investors than gold or silver. And, lower interest rates increase the relative attractiveness of the metals vis-à-vis paper currencies.
Silver and gold, especially the latter, have benefitted from the diminished expectations for further Fed rate hikes since the increase in December 2015 (Figure 2). However, expectations have begun to rebound. In August, Fed Chair Janet Yellen held open the possibility of a rate rise at the September 20-21, 2016 meeting of the Federal Open Market Committee (FOMC), and Fed Funds Futures market reacted by positioning for higher interest rates.
Growing expectations for an interest rate hike appear to have halted the rally in the gold and silver for the time being. Gold peaked in early July, and silver in early August, but the subsequent declines have been modest. One factor that might be limiting the declines and is also one of the most overlooked is: Supply. According to the GFMS Gold Survey Q2 report, gold mining supply fell by 2.2% in the second quarter of 2016 from a year earlier after having barely grown during the first quarter. While interest rate expectations strongly influence gold from a day-to-day perspective, mining supply tends to have a longer-term impact, and the fact that mining supply has stopped growing is good news for gold and silver, at least in the short run.
The total amount of global gold stocks stands at around 167,000 metric tons. About 3,000 metric tons will be mined this year, expanding the total outstanding stock by 1.6%. The total outstanding stock of silver is 1.6 million metric tons, almost 10 times that of gold. About 24,400 metric tons of silver will be mined this year – about eight times that of gold -- which will expand total outstanding supply by 1.5%.
The rate at which gold and silver stocks grow appears to have a strong and under-appreciated impact on prices. Figure 3 shows that some of the major gold bull markets took place during periods of declining growth rates in the existing stock of gold. Growth in gold stocks was painfully slow during the 1920s and early 1930s, helping to create the deflationary environment that plagued the United Kingdom in the 1920s and the United States during the early 1930s. In 1933, the Roosevelt Administration devalued the U.S. dollar versus gold by nearly one-third in order to expand U.S. money supply.
Devaluing the U.S. dollar raised the price of gold in dollar terms and led to a boom in mining production. Post-World-War-Two gold production rose steadily until it peaked in 1965. It then began to decline sharply, from 2.5% growth in outstanding stocks during the mid-1960s to barely a 1% increase in 1980. In 1971, the United States was forced to de-peg the dollar from gold and float its currency. Gold prices soared from $35 per ounce in 1970 to as high as $800 in 1980.
The consensus view is that the 1970s gold bull market occurred primarily because the Fed held interest rates below the rate of inflation for much of the decade, leading to a sharp acceleration in the rate of inflation (Figure 4). Easy monetary policy undoubtedly played a role. However, the sharp drop off in mining supply as a percentage of outstanding stock of gold should not be overlooked as a cause.
Likewise, tight monetary policy in the 1980s with very high real interest rates is commonly assumed to have caused the collapse of gold prices. Here too, positive real interest rates undoubtedly made the Fed’s fiat currency relatively more attractive versus gold during the 1980s and 1990s than during the negative real interest rates of the 1970s. But interest rates aren’t the entire story. Soaring gold prices during the 1970s incentivized an investment boom in gold mining, which over the course of the 1980s and 1990s succeeded in nearly doubling mining output as a percentage of existing gold supplies. Silver mining underwent a similar expansion. Increasing mining supply also contributed to the collapse of gold and silver prices during the 1980s and 1990s.
At the turn of the century, gold mining supplies began falling again, reducing the pace of the annual increase in the existing gold stock from over 2% per year in 1998 to just 1.5% per year by 2009. This decline in mining supply coincided with a massive bull run in gold prices, and also with a transition from positive real interest rates during the 1990s to negative real interest rates by 2009.
Since 2009, gold mining supplies have been growing again, at least up until this year. The growth may have contributed to the downward pressure on gold prices since 2011 despite Fed Funds rates being persistently lower than the rate of inflation. Of course, one could argue that the Fed’s ending of quantitative easing in 2014 and its December 2015 rate hike constituted a monetary policy tightening, which is undoubtedly true. Nevertheless, we doubt that the change in the real rate of interest is the only factor holding gold prices well below their 2011 highs. The 37% growth in annual gold mining supply between 2008 and 2016 as well as the surge in silver production (Figure 5) probably also put the prices of both metals under downward pressure.
Our analysis shows that gold and silver prices are negatively influenced by mining supplies – their own, and each other’s. That is, an increase in gold mining supply negatively influences the price of both gold and silver. Similarly, an increase in silver mining supply negatively influences the price of gold as well as silver (Figure 6).
The reason for the influence is fairly simple. Gold and silver are partially substitutable. In the gold-silver ecosystem, higher prices lead to an increase in recycling (secondary supply), bringing scrap back onto the market. Secondary supply does not negatively impact price – it is public knowledge that that amount of gold or silver has already been mined. Therefore, while higher prices incentivize the return of secondary supply to the market, it doesn’t affect prices. However, the industrial and dental use of gold and silver both respond negatively to changes in price.
Where the two metals differ a great deal is in jewelry. When gold prices soar, consumers dramatically cut back their purchases of gold jewelry (Figure 7). The same is not true for silver. Even if silver prices rise more than gold in percentage terms, consumers can still largely afford to buy silver jewelry because it costs only about 1/75th the price of gold. As such, if gold mining supply increases, it will likely lower the price of gold and of silver. This makes gold the dominant force in the gold-silver ecosystem.
Interest rates: The U.S. economy is doing reasonably well. Despite low productivity growth and declining corporate profit growth, the labor market is quite strong. The total number of people working has risen by 1.9% during the past year and average hourly earnings have increased by 2.6%. All told, this equates to a 4.5% growth in total labor income. GDP growth has been stagnant mainly due to a decline in corporate profits (from very high levels), falling inventories (a good sign going forward) and deterioration in the U.S. trade balance. Inflation has been subdued as a result of falling energy prices but this won’t last forever, and the core rate of inflation has been perking up. As such, our view is that the Fed will gradually adjust interest rates higher, and that the boost gold and silver prices got from diminished expectations for a rate hike is mainly behind them.
Mining supply: Primary supplies of gold and silver have stagnated or declined slightly, probably because the drop in prices in 2014 and 2015 led some marginal producers to cut back on their mining operations. Now, gold and silver prices are both well above their cost of production. Silver has been trading around $18.50 per ounce, giving silver producers very strong profit margins (Figure 8). Gold producers aren’t nearly as profitable on an all-in basis but gold has at least rebounded above its cost of production. As such, we think there will be a strong growth in silver mining output later this year or in 2017, whereas gold mining output should, at the very least, stabilize and, perhaps, begin to rebound.