Bruce Bartlett used his New York Times Economix blog post today to argue that “Gold’s Declining Price Is a Reversion to the Mean”. He buttresses his argument by pointing out that,
In a recent paper, the economists Claude B. Erb and Campbell R. Harvey present strong evidence that the gold market was severely overbought. The increase in gold prices did not represent a change in the trend of inflation. As the chart indicates, even with the sell-off, the price of gold is still high and has a long ways to fall to get back to the “golden constant” that gold-standard advocates cite as proof that the dollar should be pegged to gold.
Bartlett thinks this picture proves mean reversion. But it doesn’t. More importantly, he thinks that the authors of the paper with this figure claim this picture proves mean reversion. But they never actually take a stand. The paper doesn’t really test the proposition. In fact, the paper is careful not to take a stand, and never to clearly assert anything one way or another.
The paper is a nice talky piece that can be used for a fun conversation with potential investment clients who demand to talk about gold. The important thing is to touch on all the popular angles, whether or not you put any stock in them. “You want to talk about mean reversion? Here, let me show you a graph that suggests mean reversion.”
There is lots of statistical evidence that the gold price does not mean revert. In particular, there is lots of evidence from futures prices that there is no predictable reversion.
Those people who want to claim otherwise have a minimal burden of proof to meet before we spend too much time listening. Maybe the gold price does mean revert. But show me some data with some real analysis. And be clear about what’s what in the analysis. Smooth talk is a waste of everyone’s time.