In fact, this fall is doubly peculiar, because there's something else that should have raised the gold price but didn't - falling real interest rates. Traditionally, gold prices have risen when real interest rates have fallen. This is because gold has a zero yield so when interest rates are high there's a big opportunity cost of holding it; you miss out on income from cash or bonds. As interest rates fall, however, so does this opportunity cost which should make gold more attractive. This helps explain why gold soared for most of the 2000s. However, real interest rates here and in the US have fallen since the spring and yet gold hasn't risen.
We have, therefore, a double puzzle; neither geopolitical risk nor lower interest rates have helped the gold price. I suspect this is because of three offsetting forces which have depressed the metal.
One is that geopolitical turmoil isn't reflected in other asset prices either. If you look at prices of shares or oil, you would infer that the last few months have been remarkable for their lack of events. This means that two forces for a higher gold price - the search for a safe haven against stock market turmoil or insurance against inflation unleashed by higher oil prices - have been absent.
Secondly, inflation expectations have fallen in the euro area - if not in the UK or US. This has reduced European investor demand for gold as protection against rising consumer prices.
Thirdly, the US dollar has risen; its trade-weighted index is up by more than 2 per cent since the end of May. This matters because gold has tended to fall (in dollar terms) when the dollar has risen; since 1990 the correlation between the dollar's trade-weighted index and the dollar price of gold has been a hefty minus 0.79. The fact that the dollar's strength has held down gold is, therefore, consistent with the historic pattern.
You might think these are mere after-the-fact rationalisations, and perhaps weak ones at that.
Maybe. And herein lies precisely the case for investing in gold. It is the very fact that its movements are hard to explain and even harder to predict that makes gold a good investment. Because there's a fair chance of gold moving in different directions from bonds or shares, the metal is a handy way of spreading risk. If gold loses when share or bond prices rise - as has happened recently - its losses are bearable. And the counterpart of gold sagging when other assets do well is that sometimes it will do well when those other assets don't.
A little maths will illustrate the point. Let's assume that gold (in sterling), the All-Share index and gilts' future volatilities and correlations will be what they have been since 1986. A portfolio split 50:50 between equities and gilts then has an annualised standard deviation of 8.9 percentage points. This implies that, there's a roughly 8 per cent chance of losing 10 per cent or more in a year.
However, a portfolio split 40:40:20 between equities, gilts and gold has a lower standard deviation, of 7.8 percentage points. This implies only a 6.1 per cent chance of a 10 per cent loss, even on the assumption of zero expected returns on gold. This is despite the fact that gold in itself is more volatile even than equities. This apparently odd result comes about because gold has zero correlation with gilts and equities, and so helps diversify risk.
The last few months has corroborated this; the fact that gold has fallen as gilts and equities have done okay confirms its qualities as a diversifier. In this sense, gold's recent lacklustre performance highlights the very case for investing (a little) in it.
