| |
Metals shares as portfolio diversifiers
Warwick Lucas
Posted: Mon, 28 Jan 2008
[miningmx.com] -- Yes, I have a confession. I have a relatively unusual habit for a stockbroker, namely I am interested in and read a lot of academic finance study.
In particular on the JSE, I keep an eye on the work of Professor Paul van Rensburg of UCT, who has made considerable ground in attacking the efficient market hypothesis on the JSE.
Although the efficient market hypothesis has had a long standing for some time, its mathematical manifestation, namely the capital asset pricing model, seems to be violated when using the all share index as a single market proxy. Perhaps it shouldn't be a surprise, after all the fortunes of the mining companies depend on mineral prices that are established from international political and economic events that are often substantially divorced from developments in the local South African economy.
This means that it is not
unreasonable to assume that returns on mining shares and industrial shares will be influenced at times by different underlying factors. At the end of a 1997 study, Van Rensburg and Slaney determined that the performance of the JSE is far better represented using an arbitrage price theory (APT) framework, with a two factor model employing proxies of both industrial and resources shares.
In extending my
reading, I noted a 2007 paper by Katt and Oomen, focusing on the return properties of a variety of commodities within an international context. Unlike equities and bonds, commodity futures are positively correlated with unexpected inflation, a result that we would intuitively expect given the "hard" nature of commodities.
Within commodities there are big differences between likes of energy, metals and soft commodities. Overall though, it was hard to escape the conclusion that a balanced commodities portfolio could add worthwhile diversification to investments. This negative relationship with stocks and bonds, and positive correlation with inflation clearly makes for a great portfolio diversifier, but not all commodities are likely to be equally good inflation hedges.
In fact, in terms of mutual dependence, returns show low dependence between groups, but high levels of dependence within. For example there might be a low correlation between gasoline and gas, and a
high correlation between say gas and crude oil. The implication is that diversification across commodity groups is preferred over diversification within the group. This resource didn't seem to matter or change, whether daily or monthly correlations were applied to the problem.
My interpretation is that this strange kind of behaviour probably also gives rise to explain the relatively "otherwise" behaviour of South African equities. The suggestion is that the correlation would be mostly completely different those of other equity markets.
In addition, correlations with equities can vary over different parts of the business cycle, particularly in relation to energy and metals. At the end of economic expansions, correlations for energy increase and for metals decrease. At the start of recession, correlations with energy are lower and for metals are significantly higher. The reasons for this are not discussed in the paper, but I suspect one could make a food vs. durables
comparison with retail stocks (with perhaps less difference in cyclicality)!
When equities are volatile, commodities and bonds tend to correlate together, but when commodities are volatile, this is not the case. When commodity markets are volatile, equity correlation increase in metals and decrease in energy. Upshot is, not all commodities are equal when it comes to diversifying, nor do they all work in the same way at the same time. Energy is a particularly good diversifier in recessions.
The behaviour of gold recently has also been fascinating. Enormous levels of monetary reflation have seen the gold price steadily moving higher in recent years. In the past few days, matters have been trickier as with on the one hand a surge of saleable assets following sub-prime fallout and on the other an awareness of the risks posed by bank frailties and South African power cuts, the gold price has traded wildly.
At the end of the day, commodities relate to inflation
in that the investor's bottom line is not just to make money but again purchasing power i.e. outperform inflation. Whereas stocks and bonds tend to discount future cash flows, commodity prices do not and it follows that the relationship between commodity returns and inflation is quite different.
In strong economic growth, upward pressure is applied to commodities, producer and consumer prices and interest rates. Then such high prices reduce growth potential of company profits by crushing margins and in time reduce the present value of future cash flows. So stocks and bonds will drop and commodities could keep going.
It is also notable that the strongest relationship between inflation and commodities stems from the spot market. The correlating of rolling returns with inflation seems to be largely insignificant. Nevertheless the astute trader will look at the slope of the curve and how it is changing in order to give a bearing on the next direction in a commodity
price.
Best of all, there is no change in this behaviour in a turbulent market period i.e. unlike equities across different countries, there is limited evidence of such a contagion effect in times of adversity.
1. P van Rensburg and Slaney “Market segmentation on the JSE” J.Stud.econometrics 1997
2. Harry M. Kat and Roel C.A. Oomen “What Every Investor Should Know About Commodities” Journal of Investment Management 2007
| |
