Monday, September 2, 2013

The problem with gold - an academic researcher's view

Claudio Erb and Campbell Harvey in their paper "The Gold Dilemma", published in the 2013 July/Aug Financial Analysts Journal, analyzed all of the reasons for holding gold and conclude that you would be hard pressed to find one that stands up to empirical review. Makes you wonder what all the fuss about gold has been over the last few years. The real price of gold can have significant changes, and there is a positive momentum effect, but the fundamental arguments do not hold together.

Gold is not an inflation hedge. There is not a strong link with either actual inflation or unanticipated inflation. There is a positive correlation, but real price of gold as measured by the gold price to CPI ratio is very volatile. The price of gold is driven by real changes and not the inflation rate. It is the real price of gold that has been the main driver of gold volatility over the last 30 years. The real price of gold also shows strong mean reversion. High real gold prices will see subsequent real returns that are negative. There is a link between fiat money and gold but when compared with other monetary regimes, it is less clear what the link between a monetary regime and gold prices will be. 

Gold is not a currency hedge. Gold and currencies have the right correlation for a hedge, but regression analysis shows there is little explanatory power between the two. Again, the real price of gold dominates any currency relationship.

Gold as an alternative to assets with low real rates. There is a negative relationship between the real yield on TIPS and the real price of gold, but the line of causality is not clear. A time trend can better predict real gold prices than real yields. 

Gold as a safe haven. Simply put, there is not a strong relationship between gold returns being high when stock returns are low. Gold does not seems to have the characteristics of a safe haven. Gold does not have high returns when there is hyperinflation. 

Gold is not a substitute money. Given the fluctuations in the real price are not associated with monetary regime or inflation, the gold as money argument does not seem to work. 

Gold is under owned argument. Investment demand for gold increases when the price rises. The allocation in a global portfolio seems to be currently consistent with a diversified portfolio. The correlation with equities and bonds has fluctuated from negative to positive values, so the allocation that should be held in a diversified portfolio will fluctuate. However, there could be an argument that gold is under invested by some central banks which could lead to higher prices. 

There just has been too much emphasis on holding gold for reasons that cannot be supported without selected analysis of the data. You can hold gold if the real price is low as an asset that will mean revert, but the other arguments are misplaced. There may be some under investing in emerging markets, but the other key arguments are harder to justify.

No comments: