Note: In this series, I am attempting to describe the fundamental relationships between and within the yield and price complexes.
Keynes wrote that the old gold standard functioned according to an unwritten set of principles, which he called "the rules of the game." The "rules" primarily referred to gold's role in international trade and finance, but in this article, I am going to refer to them as the broader relationship that gold had with prices and equity and bond yields.
In my previous article, I showed that there appears to have been a revolution in the way that gold relates to these factors. Understanding the nature of that revolution can help investors and citizens understand some of the paradoxes of the modern market.
First, let's review some of what we have learned so far. Or what we think we've learned, at any rate.
1. Real commodity prices have been positively correlated with equity yields (effectively, the inverse of P/E ratios) at least since 1871. (For gold, this became true from the mid-20th Century).
2. All prices, but especially producer and commodity prices, were positively correlated with bond yields from 1730-1913, what Keynes called "Gibson's Paradox."
3. Under the gold standard, all nominal prices were real prices.
4. It is likely that under the gold standard (i.e., up until 1913), bond and equity yields were highly correlated with one another.
5. Therefore, real commodity prices have likely been correlated with equity yields for nearly 300 years, and possibly longer. Adam Smith also seemed to recognize a link between both prices and interest rates and the rate of profit.
6. Under the dollar standard and American hegemony, services and labor-intensive goods have been in a chronic state of inflation.
7. Under the dollar, goods generally have been in a chronic state of deflation. That is to say, the prices of goods have fallen even faster than the value of the dollar for a good half century. (This is true whether you use government figures or alternative sources. Actually, even more so if you use alternative sources).
8. Under the dollar, stocks are negatively correlated with equity yields; in other words, they are positively correlated with P/E ratios. This began in the 1920s, but accelerated after Bretton Woods.
9. Under the dollar standard, precious metals and energy have fared quite well relative to consumer and producer prices and other commodities. The real prices of things like gold and oil have been quite buoyant for the last forty years, once we account for their 'Gibsonian cyclicality.'
10. Under the gold standard, commodity/gold price ratios were positively correlated with equity yields, but under the dollar standard, commodity/gold ratios have tended to be inversely correlated with equity yields.
In sum, there are some surprisingly durable relationships that appear to be a function of the real economy and to operate independently of the monetary system, but these have had to adjust to some equally profound changes since the dollar standard began to displace the gold standard.
In this article, I would like to show how these principles can be combined to indicate major transitions in modern (post-1971) markets.
The great weakness of these rules, however, is twofold, and I think it best to acknowledge and respond to them up front.
Reasons for skepticism
First, on a personal level, I feel quite pleased with the progress I have made in the last year on the level of intermarket analysis. Unfortunately, I have struggled to find almost any "fundamental" reasons for these relationships. I cannot explain why these relationships exist. The easiest problem to solve should be the relationship between commodity prices and yields. It has existed, apparently, for three centuries. As anybody who has ever read any of my previous articles knows, I am essentially reconfiguring "Gibson's Paradox." All one has to do is recognize that it is equity yields that trump bond yields and commodity prices that trump consumer prices to make this connection.
And, yet, that also implies that the profession, the science, the calling of economics has failed to account for the Paradox, simply because it has failed to adequately frame the problem. The economists who have written about it include Tooke, Keynes, Wicksell, Fisher, Friedman and Schwartz, Shiller and Siegel, Hayek, Sargent, Barsky and Summers, and perhaps Adam Smith, as well. In other words, I am not wholly discouraged by a failure to unearth a rationale for these relationships, although if my argument about these relationships are right, it should bring us closer to one.
The second issue is that if we are somewhat at sea with respect to the causality of centuries-old relationships, then it is even harder to say how durable the new trends are. For example, I have shown elsewhere that since 1960 that the equation EY - DY - 10y + 1y - CPI% = 0 and all its permutations hold true ("EY" and "DY" referring to the earnings and dividend yields, respectively; "10y" and "1y" to Treasury yields). Like many of the major changes in the relationships between and among prices and yields, the move towards this equation began soon after the Bretton Woods agreement. Other changes began almost immediately after the establishment of the Fed (e.g., the relationship between stocks and P/E ratios). Others did not emerge until the 1970s. Real wages, for example, soared above trend under Bretton Woods and have collapsed under Bretton Woods II.
In other words, there is no reason why we should view any system, but especially the current one, as static. And, because we have never had a system such as the one we have now, unless we can come up with some absolute and ironclad economic laws or principles to explain future developments, we cannot guarantee that future performance will at all reflect past performance. The complex interrelationship among and between prices and yields and the fact that economists have failed to come up with a solution to something as long-standing as Gibson's Paradox makes it highly unlikely that we will be able to confidently formulate any ironclad economic rules in the immediate future, although that is always the goal, of course.
I have in mind previous instances of historical work done by economists on things like money supply or unemployment. The Phillips curve, for example, demonstrated a long-standing inverse relationship between unemployment and inflation from 1861-1957. If that was the case and everything I said in the ten points above were true, that implies that commodity prices and unemployment were negatively correlated during that time (since commodity prices were the primary determinant in price movements prior to the dollar standard). As I and others have written before, however, real commodity prices (especially energy) have been positively correlated with the rate of unemployment for about half a century now. It is not hard to imagine that this is tied in some fashion to the unprecedented expansion in the size and price of services, that the service sector, like the stock market, does not respond to continuous credit expansion in the same way as goods prices, but it is hard to be sure.
Punctuated Dollar Equilibrium
It is not my purpose here to account for wages, unemployment, or the impact of credit expansion as such, but to assert that over the last three hundred years, particularly in the Anglo-Saxon world, we have been living under two distinct monetary regimes and that the switch from one to the other has transformed any number of fundamental economic relationships. The system we live under now has been in place for a century, but it has also clearly continued to evolve, with major transitions at Bretton Woods I and then Bretton Woods II. An evolutionary biologist would call this "punctuated equilibrium," I think. Many of the relationships that once existed no longer do so, and many of the ones which we take for granted (say, that between stocks and P/E ratios) are actually rather newfangled. It used to be that rising commodity prices were a sign of good times. Today, it means that a storm is brewing…or that it's about to break.
Most of the changes since 1913 have been fairly persistent, however. For example, the positive correlation between stocks and P/E ratios began shortly after the Fed was established and this relationship has grown stronger rather than weaker over time. The yield equation I mentioned above is actually a simple modification of an equation I had found for the gold standard. That modified equation came online around 1960, although it appears to have been working its way towards that new equilibrium since Bretton Woods. When Bretton Woods II replaced I, that equation remained undisturbed. In other words, changes under the dollar system appear to be mostly cumulative rather than destabilizing. By "not destabilizing," I do not mean to say that they are economically or socially benign, but rather that they are not, with respect to one another, anarchical. There is a logic to it all, although it may be a vicious and opaque logic.
I am doing a poor job of saying that I am aware that the amendments to the list above that I am to present below are built on rather slim historical evidence but that the fluidity of the situation we have been in since the establishment of the Fed gives us but little alternative. Human relationships are not like scientific relationships. In the mythic story of the discovery of America, Columbus "proved" the Earth was round, and that ended the debate. The physical world has fairly clear boundaries and possibilities.
Under the gold standard, the world was flat; but once we discovered the glories of central banking and that it could make the world round, the world became round (although the discipline of economics, like Columbus, was not aware that new continents would lie between it and its goal). That does not mean that it can simply remain round as long as we will it, but it is far more nebulous than the rules that have been clearly established in the physical realm. Perhaps some deeper force will make the economic world flat again, or maybe it will morph into some other shape.
Rules of the Dollar Game
In the meantime, the following principles seem to be a fairly stable set of relationships under the Bretton Woods II order:
A. Equity yields are led by real energy prices, especially crude oil.
B. Precious metals lag. More to the point, blow-off tops in precious metals, especially relative to other prices, indicate a top in equity yields.
Combining these two points with the ten points at the beginning of this article, we can use these principles to make additional assertions about market behavior. For example, since the earnings yield is simply the inverse of P/E ratios, and P/E ratios have been positively correlated with equity returns, we can posit that:
C. Counterintuitively, a spike in precious metals relative to other commodities is indicative of an imminent bull market in equities.
D. Similarly, because commodities and inflation tend to fall with equity yields, these spikes in precious metals also foretell a bear market in commodities.
Indeed, I believe that the dramatic spikes in silver and gold in 2011 were the conclusion of a decade-long equity bear/commodity bull market. But, let's review the evidence for 'A' and 'B' first.
Precious metals and equity yields
In my last article, I essentially argued that gold had suddenly become a "commodity's commodity" from the middle of last century. That is, not only had gold begun to behave like other commodities in tracking equity yields, but gold priced in terms of those commodities also correlated with equity yields. There are too many commodities for me to go through each and every possible ratio, but it would appear that this is generally true for the broader precious metals sector, and perhaps also tin.
I also noted that precious metals (but also tin to a certain extent), as well as energy commodities (specifically, coal and oil), also had performed quite well under the inflationary dollar standard, especially relative to other commodities and consumer goods. If the mysterious powers that be are manipulating the precious metals markets downwards, they are doing an awful job of it.
But, let's look at the relationship between some of these commodities and equity yields over the last fifty years. Up until now, I have been using annual data, but now that we can focus on more recent history, we have access to monthly data, which better captures the volatility of commodity prices.
I will begin with the relationship between precious metals and the earnings yield. I have calculated the earnings yield by simply taking S&P monthly earnings and dividing them by the S&P monthly average price as reported by Shiller. In other words, this yield is not "cyclically adjusted" to bring out the trends. The intimacy of using concurrent periods draws this relationship out better than using the inverse of Shiller's P/E10, but since we are talking about long-range trends (secular changes in commodities and equities), the difference is ultimately negligible.
As I said, precious metals tend to lag, and they tend to signal the top of equity yields with a burst. You can see in the chart below, which includes gold, silver, platinum, and tin that the major peaks in the earnings yield (namely, 1974, 1980, 2007, 2011) coincide with blow-off tops in these metals (usually within months), especially gold and silver, those least prone to industrial use. Interestingly, depending on one's perspective of the last five years, one could argue whether gold or silver did a better job of this. Gold completely ignored the 2007 top in equity yields and kept rising up until it had enough in 2011.
As an aside, I have to note the peak in silver prices in 1980 and what I suspect is the myth of how that spike came about. Supposedly, that was the product of a failed attempt by the Hunt brothers to corner the silver market. In light of the historical evidence of the relationship between commodities and equity yields, it would seem that this explanation should be classed with all of the other explanations of commodity prices that we so often see repeated in the media. Even if the Hunt brothers played a role in that spike, it was undoubtedly the underlying market conditions that permitted them to kindle that blaze.
Now to energy.
Energy and Equity Yields
In the chart below, you can see various grades of natural gas and coal next to the earnings yield.
Natural gas appears to lead the earnings yield, although that does not appear to be true for the US natural gas market, which seems to lag. Coal also doesn't do an especially good job of leading equity yields. Generally, however, one can see that natural gas tended to lead equity yields, especially in the upturns of the early 1970s and late 1990s.
This relationship between energy and equity yields is more clearly visible with respect to crude oil. The secular rises in equity yields were preceded by upturns in the real price of oil.
You should note, as well, that oil shocks, like spikes in precious metals, also mark the conclusion of secular rises in the earnings yield, but they tend to lead those spikes in precious metals temporally but not with respect to magnitude. In other words, oil spikes tend to precede precious metals spikes but are not as powerful or as sustained as precious metals spikes. The impact of oil shocks during these secular rises in equity yields have resulted in the most devastating of our post-Bretton Woods recessions (i.e., mid-1970s, early 1980s, and the current global crisis).
Another thing to note about oil is that unlike almost all other prices, including most commodities, oil falls farther than gold when equity yields fall. Looking at nominal prices, one can see these relationships playing out with respect to equity yields.
Even though the price of crude oil was still pegged to the dollar even longer than gold was, it still began to rise from its plateau before the reversal in equity yields had gained any traction. The peaks in yields were marked by much stronger and sustained moves in gold than in oil, although the oil shocks appear to have been more sudden. During much of the 1980s and 1990s, oil then fell much farther than did gold, until the late 1990s, when oil began its march to $150/bl in 2008, which it has conspicuously failed to recover, even though gold and silver marched on.
The oil/gold ratio and equity yields
This narrative also largely plays out in the oil/gold ratio itself, although it can be hard to pick out at first glance. Shiller's P/E10 does a better job of drawing this relationship out, but I will stick with the earnings yield I have been using so far for the sake of consistency. Since we are talking about macro-level events, I hope readers will agree that not everything has to occur simultaneously.
In any case, typically, a rise in equity yields is kicked off by a rise in the oil/gold ratio, and a top in equity yields is signaled by a collapse in that ratio. In previous articles, I have noted that the key levels in this the oil/gold ratio (although in different contexts, which I will get to later) appear to be 0.05, 0.07, and 0.12.
As a rule of thumb, equity yields tend to rise from the point at which the oil/gold ratio trips over the 0.12 mark and fall from the point at which the ratio falls below 0.05. Again, the brave new monetary world we inhabit makes that kind of precision of doubtful reliability, but it has also been remarkably durable despite that. The confirmation of those signals can be found in the behavior of the commodities themselves. The 1980 low in the oil/gold ratio was entirely a product of the explosion in precious metals prices over and above the oil shock.
(Source: St Louis Fed, Robert Shiller)
As I mentioned in points C and D, if all of this is generally true, we can mark the beginning and end of commodity and equity bull and bear markets. Again, elevated levels in the oil/gold ratio coincide with bull markets in commodities and bear markets in stocks, while a low ratio indicates bear commodity markets and bull stock markets.
Moreover, because precious metals, especially gold, lag in their performance and because they exhibit stronger, more stable trending than do other commodities, the oil/gold ratio is a useful way to predict the behavior of the gold market itself. In the chart below, one can see that most of the biggest gold runs have been signaled by spikes in the oil/gold ratios, and the weakest gold markets by exceptionally low ratios.
(Source: St Louis Fed)
I should also point out that where these signals did not work immediately (such as in the early 1970s and after the 2007 crisis) and gold continued to rise "despite" the signal, there was an oil shock within twelve months of that signal (defining "oil shock" as an 80% year-on-year rise in oil a la Leeb). That was what happened in 1973 and 2009-2010.
On the flip side, we can also use the oil/gold ratio as an indicator of the future behavior of stocks.
(Source: St Louis Fed)
As you can see, the Dow tends to be flat when the oil/gold ratio is high, and it tends to be bullish when that ratio is low.
A similar relationship can be seen with respect to the Dow/gold and the oil/gold ratios.
(Source: St Louis Fed)
The bottom in the Dow/gold ratio coincided almost within months of the oil/gold ratio breaking below 0.05 for the first time since before the 1973 oil shock. The top in the Dow/gold ratio coincided with the oil/gold ratio breaking above the 0.12 level for the first time since 1976.
The fundamental thing, however, is that what made these relationships possible is the differences in the way that stocks, energy, and precious metals interact with equity yields.
What makes it all even more curious is that energy and precious metals, and especially oil and gold, are so unlike one another. Oil is the indispensable material foundation of industrial civilization; it is nonrenewable and cannot be recycled. When it is used up, it goes to hydrocarbon heaven. In contrast, gold, as Warren Buffett has pointed out on rather suspect authority, is almost completely useless to both humans and Martians and is simply transferred from one hole in the ground to another. That they should react somewhat differently to equity yields, I suppose, is not surprising. What is startling is that they should interact in what would appear to be a rule-bound fashion, but also that the ratio of their prices should continue to oscillate around an average for a good century and a half while virtually all other commodities have fallen in price relative to gold. The only other commodity that seems to have such a stable relationship with gold is silver (the gold/silver ratio has historically moved between the 10:1 or 100:1 levels), but it remains mysterious.
The explanations for these problems probably rest with why it is that precious metals have behaved more like commodities than have other commodities since the end of Bretton Woods. As I said before, both silver and gold, when deflated by other commodities, have become correlated with equity yields. And, one suspects that the answer to that question rests with the questions raised here before and which have been implicitly asked by economists seeking to explain Gibson's Paradox: why do commodities correlate with equity yields? And what was the role of gold in keeping consumer prices tied to commodity prices during the gold standard?
In the meantime, it is important for traders and investors to remain aware of long-term macro-trends, and the prices of oil and precious metals provide a wealth of information in that regard, provided one is aware of the history of prices and yields.
Additional disclosure: I am long Dow futures and short AUDJPY.