The rolling implosion of commodity prices since 2008 has been a massively bullish secular signal for rising P/E ratios and therefore rising equities (DIA), just as it was in the early 1980s, and it is far more important to identify the secular trend rather than every cyclical mood swing or market hiccup. Nevertheless, cyclical shifts offer opportunities to those who are more sensitive to short-term risk.
In April, I wrote that both secular and cyclical signals looked bullish over the next twelve months but that there was a slight risk of a cyclical bump this fall. In that article, I focused on the secular story, but as October approaches, I would like to talk a little about cyclical signals and what they are saying about the market.
For now, secular conditions look great, but I sense that there is a slight increase in the possibility of cyclical trouble. Although these cyclical signals remain bullish, they also suggest that the greatest likelihood of a recession would be in the fourth quarter of 2013. They are slightly too close to negative levels to wholly ignore them. I will present my evidence, and readers can judge for themselves, as always.
What causes recessions?
Let's begin with the 1970s and '80s. What caused those nasty recessions? Most people would, I assume, blame oil shocks. What happened before and after those oil shocks might cause more disagreement. Were the oil shocks themselves responsible for the recessions, or was it the inflation those shocks provoked that sent the economy into a tailspin? Ask conventional wisdom what caused the inflation of the 1970s, and people would seem to split between blaming the oil shocks themselves and monetary and fiscal policy.
In the end, I am somewhat unclear on what conventional wisdom's narrative of the 1970s is. It seems to be that oil shocks that were geopolitical in nature provoked inflation, but the inflation was stopped by monetary policy in the persona of Paul Volcker, the elder statesman of central banking. This is not an impossible chain of causation, but there does seem to be some tension built into it. Why not just assume that inflation died down because of a reduction in oil shocks?
In short, we seem to be of two minds when it comes to the problems of the 1970s and 1980s. In 1990, we had another oil shock, again geopolitical apparently, that allegedly caused a recession, but did not induce an inflationary spiral.
But, we also have a widely accepted general theory about the role of the yield curve in provoking recessions (see PDF), although not every Nobel laureate agrees. According to this theory, an inversion of the Treasury yield curve--so that the spread between 10-year and 3-month rates falls below zero--typically precedes a recession a few quarters later. This has been true for every recession for the last forty years, including those of the 1970s.
To turn back to oil, Stephen Leeb has taken the conventional wisdom about the role of oil shocks and reduced it to a formula, whereby a rise in year-on-year oil prices of 80% initiates a stock market slump until that rate of increase drops to 20%. Every recession since the Nixon Shock has followed or coincided with one of these "Leeb shocks," as I like to call them, including those of the supposedly post-oil shock economy of the 2000s.
Stages of a Downturn
It is not my purpose to make sense of what the precise nature of this connection is, but it suggests that there is some sort of relationship between the yield curve, oil prices, stocks, and recessions. I am going to argue that a typical downturn should be marked by five symptoms which come in two waves.
Ideally, it should go something like this:
1. The yield curve spread should be between 0 and 1 (rather than negative).
2. A few months later, the gold/oil ratio should spike.
3. A year later, oil should spike.
4. Stocks slump shortly after that.
5. Recession hits.
As an initial illustration, in the following chart, you can see points 1, 3, and 5. The yield curve spread falls (or rises) to between one and zero, a Leeb oil shock occurs about a year later, and the economy goes into recession. (Note: in order to draw out the relationship better, I have subtracted the 10-year yield from the 3-month yield, thereby inverting the signs of the traditional spread).
So, that brings us to the first indicator. Last summer, when long-term yields bottomed out, the yield curve spread got down into the neighborhood of 1.3. That is too high to trigger a recession signal if we wanted to interpret it legalistically, but it is a bit too low for comfort. The slump tends to follow about sixteen months (five quarters) later, which would be, by my math, this fall.
In order for a Leeb oil shock to occur in October, oil would have to go to about $150, another $45 above where it is now. If you need to find a geopolitical trigger for such a move, there are plenty of candidates. Talk of Israeli red lines and a resumption of talks about peace talks between Israelis and Palestinians should make one nervous.
The gold/oil ratio as a cyclical lead indicator
The second lead indicator is a spike in the gold/oil ratio, a much less orthodox metric than either the yield curve or oil shocks.
I have not found a wholly satisfying way to demonstrate or define a spike in the gold/oil ratio. Typically, it involves a 'sudden' spike above the 20 level, but that is not always the case. At any rate, I will use a few different ways to attempt to convey this relationship. I hope the reader will bear with me.
Below is a scatter-plot of the detrended gold/oil ratio against the performance of WTI (OIL) crude over the next twelve months. That is somewhat difficult to use in real-time, but I think it suggests the strength of the connection.
Next is a scatter-plot of the gold/oil ratio against the return on crude over the next year. A somewhat weaker connection, but still noticeable.
And, here is the year-on-year change in the gold/oil ratio followed by the subsequent performance of crude.
Although it is an imperfect tool, for whatever reason, it brings us greater timing precision than using the yield curve alone.
Anatomy of a Modern Recession
I do not want to overplay this, however. In past episodes, the procession of symptoms is rarely mechanical. Apart from the bear stock markets of 1977-1978, 1998, and 2011, here are the major downturns of the last forty years. Although timelines are somewhat difficult to generate with Excel, and I have had to simplify the duration of the yield curve spread indicator somewhat, this should give readers a sense of the patterns.
First, in 1972, the yield curve spread fell below one, and the gold/oil ratio spiked due solely to the rise in gold. As an aside, I would like to draw the reader's attention to this latter point: the relationship between the gold/oil ratio and the subsequent performance of oil is not necessarily due to a fall in oil prior to its rise. Either way, a year later things turned nasty.
(Note: The yield curve bars in these chronological charts are approximations for purposes of illustration).
The late 1970s saw a similar set up: a flattening of the yield curve and a spike in gold in 1978 signaled trouble in '79 and '80 and then immediately again in the early 1980s, although less uniformly.
In 1987, we had the first "flash crash," in the absence of any obvious culprits, blamed by some on new trading techniques. A year prior, however, the yield curve had flattened (this time due to falling long-term rates) without inverting and a spike in the gold/oil ratio (due to falling oil prices). Recession did not follow, however.
In 1990, Iraq's surprise invasion of Kuwait "caused" an oil shock that brought about a slump in equities and a recession. A year prior to that, however, the yield curve had flattened, and the gold/oil ratio fell from its high plateau of previous years.
At the end of the 1990s, the tech bubble and irrational exuberance all came to a crashing halt about a year after a typical yield curve and gold/oil set-up.
In the late 2000s, although there was a gold/oil bump, it is hard to look back on it as a spike, but the yield curve did steepen before all hell broke loose.
Finally, in 2010, we had another "flash crash," preceded by a Leeb oil shock, which in turn was preceded by a huge spike in the gold/oil ratio. No recession was to come of it, at least in the US.
That brings us to today.
Last fall, the gold/oil ratio, which has remained fairly high since the onset of the credit crisis, somewhat like the late 1980s, tipped above the 20 level and has slid pretty steeply since, but it still does not look like a classic set-up for an oil shock.
So, we have two faint signals of trouble in the fourth quarter of 2013. Oil prices are now residing above $100 again, and geopolitical trouble is always lurking about in the Middle East.
But, how should we weigh this? It is definitely possible that we are looking at the stock market's highs for 2013 right now, but the secular signals I talked about before remain lit "green" and the cyclical signals are too weak to allow them to trump the secular posture of the market on even a short-term basis this summer. The slide in the gold/oil ratio this year slightly raises the likelihood of a problem in the fall, but without a much higher oil price (at least $150), I see no reason to turn bearish on equities or bullish on commodities. There is still a greater likelihood that equities will continue to surprise to the upside.