January 25, 2015

Oil, Stocks, Bonds: An Historical Comparison

In a post a few days ago, I noted that the gold:oil ratio had reached an extremely high level of 28 to 1, such that one ounce of gold would suffice to purchase 28 barrels of oil. That has occurred three previous times over the last 25 years, in early 1994, in late 1998, and in mid-February 2009. It's of interest that all of these occurred right around the beginning of the new calendar year. In 1993-94, the high occurred on December 30, but was retested on February 15. Almost exactly the same pattern occurred in 2008-09, with a high at the end of the year and a retest in mid-February. In 1998, the high was reached on December 12; it was not retested.

I was curious what happened in the subsequent relative performance of stocks, bonds, and oil in these three previous episodes (when the gold:oil ratio touches 28). Herewith a series of charts showing the denouement.

For 1993, the chart below shows a ratio between the price of oil and the price of stocks. The second frame shows a ratio between the price of oil and the price of bonds.  As these charts show, oil outperformed bonds handily. It did so for stocks in the ensuing year, but then fell back.

Here's the same chart for the period from 1998 to early 2002.

As previously discussed, a ratio chart comparing the price of stocks with the price of oil , as these charts do, can be misleading in its practical implications. What Jeffrey Gundlach calls "investible commodities" don't correspond with a simple price chart. Still it is impressive that in both these prior instances the relative performance clearly favored oil as against both stocks and bonds over the ensuing 12-15 months. This is so even if we substitute a bond fund for the bond price, as in the following for 1993 to 1996:

For 2008 and 2009, the first chart below shows not the price index of $WTIC but the exchange traded fund USL. And I compare this to SPY (whose price reflects dividends) and AGG (a total bond  fund whose price reflects interest payments received). Just as the price charts for 1994 and 1998 don't reflect dividends and interest, the commodity prices do not reflect the transaction costs of commodity funds. That changes the implied ratio, by how much is the great question.

The above chart looks very unimpressive, especially in comparison with the price charts. The USO chart is even worse, though I forbear to show it here. The price of oil, as the chart below shows, bounced much higher than the exchange traded funds, even one, like USL, supposedly constructed for the very eventuality that transpired.  Why this happened is not clear to me; I suspect is is partly related to the incredible volatility all markets displayed during the financial crisis, greatly increasing transaction costs, not simply to the issues associated with contango and backwardization..

In conclusion: it's a mixed bag. I think oil's bottom in relation to gold is close. Once it bottoms, oil will inexorably rise relative to other financial instruments, but "investible commodities" are a pretty unsatisfactory means of expressing this, as they are handicapped by various deficiencies. Despite these deficiencies, USL looks good here relative to other financial assets; it may go down further, but over the next year I think it likely it will hold its value against other assets and will probably surpass them.

No comments: