I say this for a simple reason. There's a massive negative correlation between the gold price and yields on five-year US Treasury inflation-proofed securities (Tips). It's been minus 0.89 in monthly data since January 2005, which is as high as you get in financial economics. The long fall in real bond yields between 2006 and 2011 was accompanied by rising gold prices and the sell-off in gold in 2011-13 came as real yields rose. Since 2005, each percentage point move in real yields has been associated with a $286 an ounce move in the opposite direction in the gold price.
There's a simple reason for this. There's a cost to holding gold. It's the interest you could earn if you held financial assets instead. This means that when interest rates are high, so too is the cost of holding gold. The metal must therefore be lowly priced to reflect this high cost - in the same way that a car is cheap if it needs regular trips to a mechanic. As interest rates fall, the cost of holding gold falls, and so its price should rise.
This is consistent with Hotelling's rule, which says that the expected path of a commodity should be determined by the level of interest rates: low interest rates mean low expected price rises, and high rates mean high expected price appreciation. This rule implies that when interest rates rise, commodity prices must fall to a level from which they are subsequently expected to rise.
All this implies that if interest rates are steady we should see a slight trend rise in the gold price, albeit with volatility around this trend. This is just what we've had since mid-2013.
Equally, this warns us of the big danger to gold - that real interest rates could rise further. In the near term this would happen if the Fed's continued withdrawal of quantitative easing reduces demand for bonds or if a steady economic recovery increases investors' appetite for risk and so reduces demand for safer assets. In the longer term, it could happen if a rebalancing of China's economy towards consumer spending reduces the global savings glut that has contributed to low bond yields in the west.
Conversely, though, there are two reasons to expect that real rates could stay low and hence gold prices stay high. One would be if western economies really have entered a phase of 'secular stagnation' with lower long-term trend growth; generally speaking, lower real growth should mean lower real interest rates. The other would be if there continues to be what MIT's Ricardo Caballero calls a shortage of safe assets. This needn't be merely because investors remain risk-averse. It could also happen if Asian economies (other than Japan) continue to produce fewer safe assets such as high-quality bonds or secure bank deposits than they generate in savings.
Now, it is of course risky to rely entirely upon past correlations continuing, and you can tell stories in which gold prices rise as bond yields do so or in which both fall. Nevertheless, it's likely that gold will, barring day-to-day noise, continue to be related to interest rates. And this has simple implications for investors. It tells us that the more you expect the global economy to return to 'normal' - with decent growth in the west, less risk aversion than now and less of a savings glut - the more pessimistic you should be about gold. Conversely, gold could do well if we remain in the new world of ultra-low rates.
In this sense, gold might have a place in investors' portfolios - as a form of insurance against secular stagnation and weak long-term growth.
