In my last piece, I suggested that readers should brace themselves for a possible "Leeb shock", an oil shock, in August or September.
I argued that, based on historical trends, there was a way to use the gold/oil ratio to time hedges against oil shocks, and specifically, Leeb shocks (an 80% or greater rise in year-on-year WTI crude) and that the gold/oil ratio suggested a fairly strong, but hardly conclusive, possibility of an oil shock later this summer. In this article, I would like to reexamine that assessment based on a second and probably interrelated historical pattern.
In this instance, I am going to consider the relationship between the yield spread on Treasuries and Leeb shocks.
This is a chart of the monthly yield spread moved forward sixteen months and flipped upside down compared to year-on-year crude oil from 1968-2009. This comparison has many of the same advantages and disadvantages of my previous comparison, although the conditions are somewhat easier to define in this instance.
Quite simply, except for the 1987 and 2009 oil shocks, the yield spread flattened to below 1.0 sixteen months prior to a Leeb shock. In 1987, it only went as low as 1.24. In 2009, the yield spread was quite high (over 3.0). From the reverse angle, we can also say that on three occasions, once in the early 1980s, once in the late 1990s, and once in the early 2000s, the spread fell below 1.0 without being followed by a Leeb shock. (In the case of the 1990s, oil went up approximately 50%. In the other two cases, the yield spread broke back under 1.0 after an unusually brief period over that level).
This is, I believe, broadly in line with the conventional "yield curve indicator" for recessions, which has used inverted yield curves (i.e., spreads below 0 rather than 1) to predict recessions "four to six quarters" in advance.
If that is true, we have a kind of cluster of historical correlations involving the oil/gold ratio and/or flattening yield curves tending to lead "Leeb shocks", stock market crashes (and, as I hope to show at a later date, gold crashes), and/or recessions--a series of overlaps that might resemble a kind of Venn diagram lacking a single center, to borrow approach from the Indologist Wendy Doniger. In my mind, this is strongly suggestive of a market mechanism, or a series of mechanisms, at work that has not been clearly delineated by traditional economic theories, but so far, it is only suggestive.
In the meantime, let's look at where we stand today, in the light of historical experience.
The only number that really matters here is the 10-year rate, since short term rates have been kept extremely low for so long.
On the face of it, it does not suggest an oil spike. In the spring of 2011, the spread was well over 3.0 and near its highest levels of the last two years. The only time a Leeb shock occurred under such conditions was in the post-crash bounce of 2009. The two oil spikes in the early 2000s that fell just short of Leeb shock levels likewise occurred when the spread had been approaching 3.0 sixteen months prior. Like today, those were also periods of unusually low interest rates. Some may remember the "Greenspan conundrum".
In sum, the rise in the gold/oil ratio last year was distinct but not the most impressive example ever, and the yield spread is silent at best, but its performance since the beginning of the Great Recession and even during the decade before suggests that an oil spike is still possible when interest rates are uncharacteristically weak.
Of course, investors should and will weigh these elements differently. In my mind, the risk of a jump in oil is sufficient, even in the face of slack demand and economic woe, to warrant preparing for such a possibility. The brevity and the relative definiteness of the time-scale, which reduces the risks posed by contango, as well as the fact that oil has been revisiting last year's lows recently, convinces me that this is not an unreasonable precaution to take.
The greatest doubt I have is regarding the trigger for such an oil spike. As many ask, "Where is the demand going to come from?" And, it is definitely hard to imagine how it could materialize in the next three months. But, there is also the possibility of a supply crunch caused by a geopolitical disturbance.
One of the curious things about these two approaches to forecasting oil is that one of them (the yield spread) appeared to 'predict' the Leeb shock that occurred upon the invasion of Kuwait, while both of them appeared to 'predict' the oil crisis of 1973, in each case a year or more in advance. For some, that will likely only confirm the backwardness of this kind of approach, which is certainly understandable, but I am more troubled by the fact that a supply shock has never occurred without the yield curve flattening first.
Considering the risk of a significant rise in oil prices in the short term (if three months is still short term) and the difficulty of determining whether it will stem from the supply or demand sides (or both), seeking direct exposure to oil through an ETF such as USO or oil futures seems more prudent than acquiring exposure through energy stocks such as XOM.
Additional disclosure: I am long September WTI.