Markets Are Neither Efficient Nor Irrational (Part 3)

| About: SPDR S&P (SPY)

It is part of human nature to deny all that is unfamiliar. If a man had never beheld a snake walking upon its belly like a flash of lightning he would deny the possibility of anything walking without a foot…So beware denying any of the wonders of the Day because they fail to accord with the measure of mundane things.

-- al-Ghazali

The contest between notions of efficiency and irrationality in markets throws the veracity and utility of both economics as practiced by academics and fundamental analysis as practiced by investors into doubt. Curiously, the theory of behavioral finance most famously espoused by Robert Shiller is built around the basic tenets of fundamental analysis, which itself assumes that markets must ultimately be rational in the long run. At the same time, fundamental analysis is entirely lost at sea in the pricing of alternative asset classes such as commodities while economics pretends that goods prices are easily understood. These two fields tend to agree on one thing, though: that goods and financial assets are priced differently from one another. One cannot approach the commodity market with the same techniques one uses in the stock market. Here, however, I assert that they are not priced differently and that attempts to prove the efficiency or the irrationality of the market based on a decontextualized analysis will almost inevitably produce distorted interpretations.

The alternative I propose is that all markets are organically linked in ways that are not apparent to either observers of or participants in markets, especially at the secular level of analysis, where there is such a dearth understanding. In part, this article is an attempt to provide a little theoretical breathing room for the future introduction of a tripartite, modal analysis, where the secular mode constitutes an autonomous dimension between the structural and cyclical modes that are the bread and butter of analysis today. The current forms of analysis do a terrible job of differentiating between structural and cyclical factors, largely because the secular dimension is regarded as nothing more than a trend of a certain extension (thus, the name, of course) or a "supercycle" rather than as a state.

If we think about the basic relationships that we take for granted in markets - such as between the supply-demand equilibrium and price, stocks and earnings, and inflation and interest rates - and add a little historical context, we will see how flimsy our ready-at-hand explanations of economic relationships are and where we should begin looking for deeper insight.

Irrationality and Efficiency In Social Systems

The Irrational Market Hypothesis, if you will, is a response to the weaknesses of the Efficient Market Hypothesis, but what justification might there be for starting out with an assumption of efficiency? Ascribing efficiency to any social system is a difficult task; the argument must always revolve around what "efficiency" might mean. Every such system looks efficient from within and inefficient from without. Shiller uses psychology to explain why the simple equation of stocks = profits does not work. Anthropology might have been more useful.

The function of a single type of unfamiliar behavior (say, human sacrifice) may seem irrational when taken out of context, but when put in context, one can see that it was not merely irrational. It had a purpose and a function. The purpose and function may not have been for the good of any individual but rather for that of the group, or more precisely for the existence of the social system above and beyond the group. And, indeed, when a given group is asked to explain its most basic beliefs and practices, those explanations almost always sound laughable from an outside perspective (like a musician explaining the meaning of his/her music).

Another way of putting it is that economics assumes that markets work at the level of an aggregate of individually conscious rationality or a subconscious emotionality and the discipline cannot bring itself to consider the possibility that markets have a logic and dynamic of their own whose priorities are different from (rather than either beneficial or detrimental to) the interests and theories of any particular class, school, individual, nation, civilization, or species. (Who knows what the Invisible Hand is really up to? Does efficiency imply beneficence?)

Economists, whether subscribing to EMH or not, take it for granted that economic relationships should be linear and mechanical, and EMHers will trot out ever more complex equations to force the issue. Shiller's attempt to disprove efficiency is built on the demonstration that the system fails a key linear test (in the relationship between stocks, dividends, and total returns); he then leaps to the conclusion that markets must be driven by sentiment.

There is no logical reason for any spontaneous, complex system to behave as the participants think it should. Our logical, rationalist assumptions tell us that first we have individuals, then we have social systems (societies, families, markets, etc), but a brief reflection on the facts of life shows that this is impossible. It would be at least equally true to say that systems precede individuals.

To put it more bluntly, whether you look to God or to Darwin, it's his world, and we're just living in it. The deeper truths of a complex order are not likely to be immediately apparent either to experience or logic. Shiller's critique of efficiency seems fair, but his solution lacks context. I will try to point to a different order of possibilities.

Simple Questions About Goods Prices & Asset Valuations

If stock valuations operate according to a discrete logic of their own, separate from the supply/demand regime that allegedly governs the world of goods and services, why is it that both commodity, producer and consumer prices have always shown a distinct inverse relationship with financial asset valuations?

Inflation and stock and bond P/Es 1875-2012

(Source: Calculations from Robert Shiller data).

UK consol yield vs consumer and producer prices 1730-1914

(Source: Roy Jastram's The Golden Constant; Bank of England)

In a certain respect, this is a relationship we take for granted. Everybody "knows" that there is a correspondence between the price of goods and interest rates, and that is no different from saying that there is an inverse relationship between bonds (NYSEARCA:UST) and goods prices.

Any number of prominent economists have tried their hand at this problem. Wicksell, Fisher, Keynes, Hayek, Friedman, Sargent, Barsky and Summers, as well as Shiller and Siegel, have each attempted to explain the precise nature of a relationship that goes back centuries. Generally speaking, the explanations vary a good deal, but there doesn't seem to be much disagreement about the inevitability of the connection.

In other words, we have a fairly well-developed sense of how goods prices interact with at least one class of financial assets (bonds).

Shiller's attitude with respect to bonds, however, is somewhat confusing. On the one hand, he argues that there is a long-term historical relationship between consumer prices and bond prices (as he argued in the article linked to in the previous paragraph), but this seems to contradict the notion that financial assets are priced irrationally. Although he also wrote a paper (both of these papers can be found in Market Volatility) discussing excess volatility in the long-term bond market, it certainly lacks the power and conviction of his critique of the stock market. And, if an occasional mumbled quote from Shiller about a bond bubble can be found on the internet, he generally seems much more reticent on this point. In Irrational Exuberance, he seems to suggest that stocks are much more prone to sentimental excess than are bonds.

In the end, Shiller seems to believe that the relationship between goods prices and bond prices means that the latter are more rational than not. In other words, this seems to be an exception both to the notion that financial markets are necessarily "irrational" and that financial markets are priced differently from goods markets.

Regardless of what Shiller may or may not think about this particular question, if this connection between consumer prices and interest rates is as persistent as orthodoxy seems to promise, then we have at least one clear example of a breakdown in these oft-quoted claims about financial assets.

The question that the first chart in this article poses is whether the connection is not between stocks and inflation rather than bonds and inflation. Is it possible that we are wrong about the bond-inflation connection? I think that is probable, but one then has to ask how such beliefs can become so authoritative and for so long. Are the oppositions we organize our thoughts around (irrational vs. efficient, supply vs. demand, monetarist vs. Keynesian, left vs. right, bear vs. bull) relevant to reality, or are they just social rituals that warm us with the sense that we have a pretty good handle on what's what?

Goods Market Fundamentals

As I pointed out in part one of this series, the idea that prices for goods and services represent an equilibrium between supply and demand is as indemonstrable as are the theories of efficient asset markets. Anybody who has traded commodities is surely familiar with the difficulties of getting those markets to be any more "rational" than the stock market. Many commodity partisans (those who hold ideological positions on commodity prices) will insist that the markets are being manipulated by powerful special interests, of course. For precious metals, the manipulation is imagined to be downwards; for oil, it is upwards.

The key issue is not real or imagined short-term distortions in prices, but long-term price behavior, and there are very few, if any, commodities that evince this relationship between supply and demand.

One doesn't have to take my word for it. The relationship has never been demonstrated. But, the supply-demand equilibrium model of price determination is uttered like an incantation at the beginning of every discussion of markets. It is part of the English idiom: "It's simple supply and demand."

In the commodities market, where we have relatively uniform products that go back centuries, academics still cannot agree on how supply and demand interaction determines price, only that it does.

Stock Market Fundamentals

As for asset prices, in part two, I showed that while equity prices are hard to account for, since valuations can move to rather extreme levels without apparent reason, the notion of "irrationality" proposed by Shiller does not fit the historical facts very well. Not only does the behavior of the stock market that we term "irrational" (where stocks seem to have no interest in earnings and dividends) seem to have specific patterns under modern monetary arrangements, but under the gold standard, that behavior seems to have been far more conventionally rational. Not only did stocks track earnings very tightly, but P/E ratios appear to have been less volatile.

Correlation between 5yr stock performance and earnings 1871-2010

(Source: Calculations based on Shiller's data)

In fairness to Shiller, near the end of Irrational Exuberance, he does mention the episodes in 1899 and 1922 that do not conform to his generalization about P/E ratios and stock market returns, and he acknowledges at the beginning of the book that even if one had bought the market at the peak of CAPE in 1901 that it would have given investors more than a decade of steadily positive returns. On balance, though, he downplays these outliers, and he does not draw attention to their chronological proximity.

Inflation and Financial Asset Valuations

Shiller also notices that high P/E ratios in the 1960s and 1990s coincided with low inflation, and he argues that this low inflation fed the exuberance, because investors assumed that a "new era" existed. This is an odd interpretation on Shiller's part, because deflation accompanied the 1929 and 1894 peaks, as well. For the 1890s, it is very hard to argue that low inflation was coinciding with "irrational exuberance" about the economy. In fact, that period has been regarded as a "depression."

Today, we are confronting just such another (theoretically unhappy) episode of disinflation, just as mainstream observers start to accept that this is a secular bull market with room for multiple expansion - even though up until very recently, the entire market had apparently been waiting for reflation to occur in order to get the markets moving again.

We do not have data for P/E ratios prior to 1871, but we know that goods prices, especially producer goods, were highly correlated with bond yields (therefore, inversely correlated with bond prices) up until 1914, when interest rates became subject to the manipulation by the Fed. And, from 1871 up until today, goods prices, especially producer goods and primary commodities, have been inversely correlated with equity valuations as measured by the P/E ratio.

Commodity producer prices vs price/earnings price/dividend 1871-2010

(Source: Shiller data; Jastram; Stephan Pfaffenzeller's updated Grilli-Yang Commodity Price Index)

In light of how much has changed since the time of the British-led gold standard, is it possible that the constancy of this secular coincidence of disinflation and high P/E ratios points to a deep structural fact about markets and economics?

An EMH Account of Link Between Inflation and P/E

The EMH authors of my desktop reference on investing note a connection between P/E ratios and consumer inflation, arguing that the latter factor "reduces the quality" of earnings which investors then appropriately discount in the stock price. As with much in EMH and economics generally, every particular explanation of this sort has a certain knowingness about it, but why would investors be all that fastidious about the quality of earnings at the level of the broad market over decade-long periods? Does it really make sense that they would deliberately avoid earning a cheap buck?

That textbook was published in 2002, before the episode in the following decade in which P/Es fell without a secular rise in inflation but with a sharp rise in primary commodity prices. More to the point, one would have to ask why stocks have thrived in a period of historically unprecedented structural inflation (read, for example, David Hackett Fischer's The Great Wave for a highly readable historical overview of inflation in the last millennium) if this "only quality earnings, please" notion were true. Since World War II, there has been a clear upward bias in the price/dividend ratio that has coincided with a relentlessly inflationary trajectory in consumer prices.

Price/dividend ratio vs CPI 1871-2010

(Source: Shiller; Jastram and Federal Reserve)

My suggestion is that a structurally inflationary credit regime has the rather non-linear effect of boosting the structural-level performance of stocks while inverting that performance at the secular level, which would tend to be more likely if the relationship between stocks and inflation were interdicted by third factors.

The popular notion that commodities represent a safe haven is unlikely, because under the gold standard, stock prices and commodities were more likely to positively correlate with one another than to negatively correlate. In other words, the link is clearly between equity valuations and goods prices rather than the stock market itself.

stocks vs commodities vs earnings 1871-1914

(Source: Shiller; Jastram)

Causality or Coincidence?

If the connection is, in fact, primarily between equity valuations (rather than bonds) and goods prices, the theories of efficiency and irrationality and the discipline of economics generally have a lot of explaining to do. Can it possibly be that the "irrational exuberance" in the equity market drove down the price of tin in the 1980s and '90s while a lack of confidence drove up the price of urea in the 2000s? But, then, what is the nature of the connection between goods and P/Es, and why are commodities and producer goods more sensitive to that connection than consumer goods and services? (However improbably, the relationship between consumer inflation since 1960 seems to be equally balanced between bond and equity yields, as outlined here).

The historical relationship between goods prices and asset prices suggests that both categories are influenced by a similar set of forces rather than "directly" impacting one another. Both asset valuations and goods prices seem to be symptomatic of much deeper forces at work in the economy than either economic theory or the day-to-day experience of businesspeople and traders have ever been conscious of.

An Uncomfortable Choice

We are forced, then, to choose between two uncomfortable worldviews:

1. Either the economic world (markets collectively) is built on arithmetical relationships (e.g., demand/supply = price) that can only be demonstrated by arcane, geometrical models of reality (academic equations and models),

2. Or, that world is built on an unknown and unsuspected geometry that somehow produces simple arithmetical relationships (e.g., prices = E/P).

Let's call the first the Efficient Market Hypothesis (where efficiency is a function of a given market's ability to integrate information), and the second, tentatively, the Improbable Market Hypothesis (where improbability is an inverse function of our estimate of total market insight).

Corresponding with these two points would be the following perspectives on the trajectory of economic thought:

1. There used to be a world without EMH, but now we have much greater total insight into how markets function. EMH has contributed to our total capacity to process information correctly, permitting us to stop wasting resources on the pursuit of alpha, thereby increasing efficiency. On the other hand, efficiency is derived from everyone pursuing alpha.

2. We are always forced to act in the twilight between "known" and "unknown." It is improbable that any totalistic account about markets is final. Relative to the totality of the markets, the individual or the group is bound to feel that the whole exhibits a certain "lumpiness," where things not only happen randomly but patterns emerge that seem inexplicable and willful.

I think that if one reflects on any other area of life, something like the latter condition almost always tends to be the case rather than the former: reality is unspeakably complicated, even to the point of absurdity, but it is not undifferentiated noise. This is also the path of fundamental change, the overturning of initial, logical assumptions by ones that must, as a matter of course, be initially improbable. It is only in hindsight that relatively primitive belief-systems seem irrational and ridiculous. When they are in the ascendant, they seem perfectly self-evident.

Birds of a Feather Flock Together

Let me try to describe the conundrum in less abstract terms. In Irrational Exuberance, Shiller uses the analogy of a microphone to illustrate the notion of feedback. In Edward Leamer's book, Macroeconomic Patterns and Stories, he suggests that one use a biological metaphor rather than a physical one to make a point about markets, so I will reluctantly attempt something like that:

Back when I was a kid living in the country, I wanted to be an ornithologist (when I was free from quarterbacking for the Miami Dolphins and driving tractor trailers). Naturally, I developed a certain appreciation for some species of birds while disliking others. One of the ones I could never learn to admire was the European starling. They were obnoxious and ubiquitous, the Eurotrash of American birds. They had a talent for mimicry but without the charisma of a mockingbird or a catbird. Neither individually nor in gangs did they seem to have anything useful to offer. But, if you have ever seen them in the mass, in flocks, they produce a quite different effect. The complex displays give the impression of being willful but without obvious purpose, and beyond the ken of rational explanation. You can see one of these displays on YouTube, and if you read the comments beneath these kinds of videos, these pests, within this different context, tend to provoke aesthetic rapture and theological debate. Now, if you go to a website like Northwestern University's NetLogo, their models contain a series of basic variables that allow you to recreate the spontaneous behavior of bird flocks. For some, that may or may not say something about the ultimate divine origins of the natural world, but it would not be appropriate to simply assume that since individual starlings failed to indicate this ability to produce spectacular "murmurations," as they are called, that this therefore suggested a non-rational explanation for their collective behavior. In other words, just because you couldn't see anything special about starlings when face-to-face with one, nor could you see any purpose or cause to their elaborate formations, that wouldn't mean that the behavior must originate outside of the nature of the starlings themselves, whatever the ultimate origins of that nature is. That is, it does not exclude an expectation that there is an ornithological or zoological or scientific, rational explanation.

The same is true for financial assets and goods prices. Neither everyday observation of individual goods or securities (or classes of goods and securities) or their metrics nor logical abstraction suggest the deeper connections and possibilities of the whole, but that doesn't mean we need to chuck out the very idea of rationalism at the first sign of trouble. It means, rather, that something is fundamentally wrong with our prejudices.

The End of the Beginning or the Beginning of the End?

The impact of Shiller's overthrow of strictly economic analysis and the new respectability his theory must have deeper reverberations throughout the academic and investing worlds as it becomes part of orthodoxy. It is only a matter of time before it dawns on these communities the full implication of the idea that the failure of the markets to live according to the basic principles of analysis can be accounted for with non-fundamental narratives. At the moment, Shiller's fame and praise is for his demonstration that markets have too often been "wrong." That is an essential development, but his answer of "irrationality" suggests that there are no economic explanations for deviations from the models and that markets cannot really be understood.

Methodologically, it would have been better to question the principles of efficiency rather than to accept them as correct before discarding them for being useless, which is what Shiller's critique of the efficient market hypothesis boils down to.

The connection between goods and asset prices strongly suggests that our most basic explanations about markets and economic principle are almost unbelievably ill-constructed. In one sense, in this article, I have merely taken the connection between inflation and interest rates that most people take for granted, turned it into an asset pricing question, and then elaborated upon it to demonstrate even more stable relationships between prices (especially producer prices and commodities) and equity valuations. But, these as yet unexplained connections rend the fabric of our assumptions about how markets function.

In future articles, I hope to lay out a few, provisional principles of macro-analysis to help us navigate this woeful state of affairs. In the meantime, the historical relationships between commodity prices and equity yields and, over a shorter time-frame, between equity yields and stock prices suggests that stocks (SPY, DIA) are in a secular bull mode while commodities (GSC, GLD, SLV) are in a secular bear mode. This state is confirmed by the post-crisis behavior of interest rates and earnings, as well.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.