Saturday, October 17, 2015



This post continues the theme of the last post. So I recommend you read that before reading this.

I’m going to try to get a quantitative estimate of the returns we might expect from commodities as an asset class. Actually I’m going to try to predict the value of an index of commodity prices in ten years. Why ten years? Because I don’t have a clue of how to predict at short horizons. With the financialization of commodity assets (we will get to that in a future post) the shorter term is basically a matter of psychology and the progress of world economic growth. Who knows that? Over ten years fundamental models do work rather well, as we will see.

Awhile back I read a study by Vanguard that used a number of methods to forecast equity prices. Interestingly it came to the same conclusion: you have to go out ten years or so to have reasonably good results. It also found that the best predictor was valuation. So if you buy when stocks sell at a low price earnings ratio, the odds are pretty good that you will do well over the next decade. We are going to do more or less the same thing with commodities.

Here is a chart that illustrates what I mean. I’m going to use this type of chart a fair bit, and you may not be familiar with it. So here is what I am doing: The horizontal axis is the price of the commodity index in 2015$. The vertical axis is the forward ten year return. So a dot on the graph, say from June 1999, is the price on June 1999 and the change in price from June 1999 to June 2009. Obviously this stops in October 2005, since I don’t know forward ten year returns from then on.

The graph looks pretty much like it should. As the commodity price index goes down, forward return goes up. It’s not a straight line; at very low levels of the price index, the return gets much higher. If you plot it on a double log scale, it is straight. The major outliers on the graph are during the period 1984 till 1992, when the forward return was lower than expected. My guess is that’s because the lingering effects of the great inflation of the late 70s took more than ten years to wear off.

Right now we are at 102 on the horizontal axis. This give an expected return of about 40% over the next ten years. Note that this is a constant dollar return. So if you expect inflation to run 2% a year, the total return would be 64%. From that you would have to subtract carrying cost (contango) or add in backwardation (not likely). The total return would be positive but not outstanding, maybe 3% per year. That’s what I mean when I said in the first post that commodities were somewhat cheap, but not a screaming buy.

If you want to do a little advanced statistics, you can get a somewhat better fit. I have done it although this is not a place to go into the technical gore. The forecasted return gets a little better, from 64% to 74%.

Next Post: The emerging market middle class and the myth of an agriculture shortage.

No comments:

Post a Comment

Comments are welcome, although I can't promise to answer every question.