In our last post, Outlook for the 2020s, I outlined the big picture for the coming decade as I see it. I argued on valuation grounds that both nonUS stocks and physical commodities will outperform US stocks. Moreover, while US Treasuries are more attractive than most other sovereign bonds, especially European and Japanese sovereign bonds, they should not be expected to turn in anything like the stellar performance of the first twenty years of this millennium. But what does that mean in terms of portfolio construction? Is it a good idea, for example, to avoid US Treasuries?
No. A good investment portfolio is more than just a collection of good investments. How they work together determines the level of risk involved. Even if US Treasuries turn out to be the mediocre performers I anticipate, they nevertheless are still likely to have a low to negative correlation with stocks. So a portfolio that includes stocks is likely to be less volatile if it also includes Treasuries than if it does not. This in turn means that if you have to tap your portfolio at some point, you’re less likely to be forced to sell an asset that’s in the dumper than if everything in it tends to rise and fall together. For the same reason, gold is an excellent asset to include in a portfolio with stocks, even when it doesn’t sport especially high returns itself.
As it turns out, of the physical commodities I track – copper, gold, silver and platinum – gold probably has the least upside potential over the coming decade. It looks better than US stocks, and due to its low correlation with stocks, should be part of every investment portfolio. But over the coming years, I nevertheless expect silver to outperform gold. As of the beginning of this decade, copper is a bit cheaper yet, and therefore has more room to mean revert to the upside. Finally, of these four physical commodities, platinum is historically the cheapest, and consequently has the most room to rise.
But since none of these commodities has a cash flow, what can we use as a basis for valuation? This is based on an analysis of their current prices relative to their historical ratios over the first two decades of the 2000s. The below chart gives a visual indication of this. Specifically, it is of the log prices of each since the beginning of 2000, normalized to their mean levels, and plotted relative to the mean of the four. The result indicates how the price of each has fared relative to the group, and eliminates the dollar from the picture.
As you can see, the chart indicates that the last decade ended with, and this decade opened with, gold as the most richly valued of the four, with platinum as the least richly valued of the four.
Bear in mind however that of the four commodities, gold is also the least correlated with stocks, so in a portfolio that includes a significant weighting in stocks, gold continues to merit a significant weighting as well. And its status as the most richly valued of the four commodities doesn’t mean it’s richly valued compared to US stocks; to the contrary it still looks cheap in comparison to US stocks.
What about other commodities? In reply to JK’s question about oil in my last post, I pointed out that oil is highly correlated with copper. The same applies to some degree to virtually every physical commodity, since we price them all in the same currency. While week to week correlations may be merely significant, over longer time frames this common denominator has typically been the overwhelming factor in commodity prices. Over the past several decades, for instance, most physical commodity prices have risen in dollar terms by about an order of magnitude, if only because the value of the dollar has fallen by about an order of magnitude. Given recent trends in US monetary and fiscal policy, and the likelihood for more in response to a deflating financial asset bubble and political attitudes towards debt and spending, this phenomenon should be obvious in spades over the coming years.