DRI-494: Wall Street’s Pretense of Knowledge

From the beginning, primitive man observed regularities in the physical world. Indeed, the conduct of daily life is virtually impossible without the implicit formulation of theories of cause and effect. Eventually, man began to record his observations, draw inferences from them and test them against reality, thus marking the formal start of science. Mankind’s early material progress was tied to a grasp of nature’s physical laws.

Many centuries elapsed before formal laws of social behavior were developed. Fortunately, human institutions outpaced the development of the human mind. The market developed through a process of spontaneous evolution, not conscious design; otherwise, economic growth and development would have lagged centuries behind actual historical fact.

At first, the social sciences utilized methods quite unlike those of the natural sciences. Midway through the 20th century, this changed fairly quickly. The break was made in the study of economics, where theory became highly mathematical and abstract and empirical work adopted methods of statistical inference pioneered mostly by British statistician Ronald Fisher in the 1920s. These changes were made in a conscious effort to imitate the theory and practice of the natural sciences.

This change has had far-reaching effects. Unlike the tremendous gains made in the physical sciences, however, the results of abstract mathematics and statistical inference employed in economics and other social sciences have not benefitted mankind. A recent article in The Wall Street Journal provides concrete support for this contention.

The ‘Relationship,’ ‘Correlation,’ and ‘Bedrock Principle’ Linking the Foreign-Exchange Value of the Dollar and the Price of Crude Oil

The March 7, 2012 edition of the Journal contained an article entitled “Ties That Bind Oil and Dollar Snap.” From the subheading, we learned that “traditionally, crude fell when greenback rose, but global influences have changed relationship.” According to author Christian Berthelsen, it has been “a bedrock principle that has dominated the oil market for nearly a decade” that “oil prices move in the opposite direction of the dollar.” But this lodestar is now being dimmed by “rising tensions between Iran and the West.”

The dollar is a monetary unit that has no “direction” in and of itself. The context reveals that the author is referring to the foreign-exchange value of the dollar; that is, its value when exchanged for the currencies or monetary units of other nations. This is not self-evident, by the way – the “value of the dollar” could just as easily refer to its purchasing power in domestic goods and services at current prices. (Indeed, the author uses the ICE foreign-currency index to measure the dollar’s value against a market basket of foreign currencies, much as the Consumer Price Index is used to gauge the dollar’s domestic purchasing power.) Why make an issue of careless terminology? This kind of semantic and analytical sloppiness is significant because it permeates the article and infiltrates the author’s central premise.

For over a decade, dollar depreciation against foreign currencies has tended to coincide with increases in the price of crude oil. Mr. Berthelsen’s account of this is long on assertion and short on explanation; the heavy lifting is assigned to a statistical tool called “correlation,” which has “often reached as much as minus 0.9.” The “minus” means that the dollar’s foreign-exchange value and crude-oil prices have tended to move in opposite directions; the “0.9” means that the movements nearly always follow this pattern. (1.0 would denote a perfect negative correlation.)

But “U.S. oil prices have climbed 12% since early November,” presumably because of a report by U.N. inspectors who suspected Iran of trying to weaponize its nuclear capability. Meanwhile, “the dollar has also strengthened over that period, gaining 4.8%,” thus violating the observed inverse relationship. “Right now, the correlation is 0.3,” denoting a positive relationship.

Noting the existence of a pattern does not, ipso facto, explain – let alone justify – the reason for it. Out of 610 words in the article, Mr. Berthelsen spares 17 for this purpose: “…a higher dollar makes dollar-denominated oil more expensive for those buying in other currencies, reducing demand.” Although vague as well as sloppy, this explanation at least has the merit of correctness. An example will clarify it.

Longtime commercial tradition quotes the per-barrel price of crude oil in U.S. dollars. Suppose, for purposes of exposition, that we assume a crude-oil price of $60 per barrel. Suppose the hypothetical exchange rate between dollars and English pounds sterling is $1.5 per pound sterling. That means that the English price of oil is 40 pounds sterling.

Now suppose that the dollar were to depreciate in value, so that it required $2 to purchase 1 pound sterling. The world price of oil, quoted in dollars, would not immediately change; it would still be $60 per barrel. But that $60 would now be worth only 30 pounds sterling, so that crude oil would be much cheaper in England. In fact, oil would be cheaper in all foreign countries against whose currencies the dollar had depreciated, as it did against the pound sterling.

Buyers in those countries would rush to buy oil in the spot market and to renegotiate contracts for future purchases of oil. Today, the non-U.S. (or non-dollar-pegged) demand for crude oil constitutes a much higher fraction of total world demand than has ever been the case. Thus, the increase in non-dollar-pegged demand for oil would drive up the price of crude oil. Voila! Dollar’s foreign-exchange down, crude-oil price up – an inverse relationship, just as predicted and observed, at least until recently.

What Happened? And Who Cares?

Since November, this so-called “bedrock principle” of the oil market has been overturned. The dollar has gained 4.8% against an index of foreign currencies, which suggests that the price of crude oil should have fallen – but it rose by 12%. Oh, no! What went wrong? Why does The Wall Street Journal seem to think this is a calamity of some sort? And, most important of all, why does this episode represent a microcosm of what is wrong with economics and the public perception of it?

The second of the three questions is the easiest to answer. The article’s author, Mr. Berthelsen, notes that the perversity of the dollar-exchange-value/crude-oil-price relationship “has thrown some trading strategies into disarray.” This is a delicate way of saying that some Wall Street traders and firms are losing money. “It’s disconcerting, when a trade like that works until it doesn’t,” is the reaction of John Kilduff of hedge fund Again Capital. “The correlation has substantially broken down,” moans Alan Ruskin, foreign-exchange strategist for Deutsche Bank. When Wall Street bawls, The Wall Street Journal starts leafing through Dr. Spock.

One might sense the makings of a mystery worthy of Sherlock Holmes and Adam Smith combined. In fact, the logic involved is so basic that the solution might be the subject of an essay question for college undergraduates in intermediate microeconomics. The answer to the first question – what went wrong? – involves little more than an exercise in the economic technique called comparative statics.

Return to our hypothetical example above. Picture the British demand for oil in your mind’s eye as the summation of many million individual demands. At any one point in time, many parameters affect those demands. They include British incomes, tastes and the existence of substitutes for crude oil. Another one of those parameters is the exchange rate between dollars and pounds sterling. When the pound’s foreign-exchange value depreciates, that makes crude oil more expensive and British citizens want to buy less of it at each possible dollar price; e. g., their demand decreases. But – when the threat of future war and oil-supply disruptions in the Middle East raise the specter of future higher oil prices, British citizens want to buy more oil now; e.g., their demand increases. So we have two parameter changes, each moving in opposite directions, pushing against each other. What will the net effect on British crude oil demand be – will it increase or decrease?

Without further information, we don’t know. But we do know this: the exchange-rate effect (the demand-decreasing parameter) operates only in some countries, the ones that don’t use U.S. dollars or a currency pegged to the dollar. The fear of future higher oil prices, though, operates in all countries, including even the U.S. Thus, it rates to be the stronger one worldwide. This means that the demand increase will almost certainly win out overall, which will produce an increase in the world price of oil.

As noted above, this analysis is not exactly rocket science. The technique simply compares the “before” and “after” effect of parametric changes on equilibrium price, without saying much if anything about the path taken by price along the way. This is called “comparative statics.” Economists teach it to freshmen and sophomores in college and expect them to learn it by the time they are juniors and seniors.

Which brings us to question three – why should we care if neither Wall Street hedge fund honchos nor Wall Street Journal reporters seem to have mastered the economic principles underlying what they do for a living?

If You’re So Smart, Why Aren’t You Rich?

The short answer is that there’s no reason why we should care whether hedge funds go broke or not, as long as we don’t happen to be invested in them ourselves. Unfortunately, the trend of events has moved decisively away from free competitive financial markets – in which failure is accepted as the necessary counterpoint to success and wealth – and towards the socialization of risk. The doctrine of “systemic risk” and “too big to fail” is interpreted not merely to allow, but to positively demand, the rescue of failing financial firms. This tends to create an unhealthy interest by taxpayers in imprudent behavior such as that described by the Journal article.

That short answer is woefully inadequate, however. It implies that regulatory scrutiny can substitute for marketplace competition in disciplining financial market participants. As Jeffrey Friedman and Wladimir Kraus point out in their minute examination of systemic risk and regulatory failure, Engineering the Financial Crisis, the regulatory agencies who failed so miserably to anticipate and curb the housing bubble and foresee the financial crisis were staffed mostly by economists.

If any economist worth his or her salt could have produced a comparative statics analysis of oil prices and exchange rates, does it then follow that economists are capable of successfully trading oil futures for hedge-fund clients? We can supply the answer to that by posing what the former Donald McCloskey aptly called “the American question – ‘If you’re so smart, why aren’t you rich?”

Very few economists have attained great wealth, none by trading financial assets. Keynes did well in foreign-exchange markets but poorly in stocks. The Nobel-winning founders of Long Term Capital Management got off to a rip-roaring start as asset managers – until they lost their shirts and had to be bailed out by the Federal Reserve. Milton Friedman let Peter Bernstein manage his money. Paul Samuelson’s successes in financial markets mostly came similarly, via the management of others.

Rather than paving the royal road to riches, knowledge of economics teaches us what not to do.

The Pretense of Knowledge

F. A. Hayek titled his Nobel Prize-acceptance speech “The Pretense of Knowledge.” It was a criticism aimed at his fellow economists, largely for their promise to maintain full employment without inflation by manipulating the money supply and government expenditures. But Hayek’s sentiments also apply to economists’ quantitative pretensions that are, if anything, more prevalent now than when he wrote.

For decades, economists have taught the subject of macroeconomics to undergraduate students using the simple Keynesian model. That model treats it as a mere diagrammatic exercise, using either geometry or algebra, for governments to remedy a condition of high unemployment by increasing government spending and/or reducing taxes. In fact, teachers commonly supplied hypothetical numbers to their freshman or sophomore students and made them reproduce the calculations necessary to reach “full employment.

Anybody conversant with reality knows that governments cannot really do this – and this is not for want of trying, either. Economists in and out of government are popularly expected to provide forecasts of economic variables ranging from gross domestic product to interest rates to unemployment. They cannot do this, either – at least not any reliable degree of accuracy. It is no excuse to say that the public is demanding the impossible of economists; after all, we don’t hear them refusing the commission, do we?

Hayek was even more vitriolic in his denunciation of economists’ use of statistics. He claimed that the natural and biological phenomena to which the laws of large numbers validly applied were not comparable to the social phenomena that comprised the subject matter of economics. He did not live to see the partial vindication of his position in the work of economic historian Deirdre (formerly Donald) McCloskey and Stephen Ziliak. They documented the blatant misuse of the concept of “statistical significance” in the vast majority of articles appearing in the American Economic Review during the 1980s and 1990s. The term is widely thought by the general public to refer to the quantitative impact of a study’s results, but it actually refers only to the representativeness of the sample used in the study. Yet 70% of published articles failed to make this distinction in the 1980s; in the 1990s, 79% failed.

What can economists do? Well, the Law of Demand is widely viewed as the most universal of all economic propositions. It posits an inverse relationship between the price of an economic good and the quantity of that good that people will wish to purchase. (More than a few cynics, including those who support abolition of the Nobel Prize in economics, claim this is the only economic proposition of any value or significance.)

Yet even economists, choking in their own hubris, would never dream of following the Wall Streeters’ example. For example, no economist would calculate a simple correlation coefficient – apparently what was quoted in the Journal article – between the price of oil and purchases, and then use it as a forecasting tool. For one thing, it would tend to show a positive or direct relationship between oil prices and purchases, not an inverse one. Why? Over time, prices, real incomes and technological innovation tend to increase together, thus allowing us to purchase more oil at higher prices.

The Law of Demand is a ceteris paribus proposition; its strictures hold all other parameters constant. Economists invoke it when the time period is sufficiently short or the scope sufficiently narrow so that variables like the money supply, consumer incomes and changes in related markets do not affect the result. Hayek doubted whether actual quantitative prediction was even possible in economics – he preferred to confine himself to “pattern predictions,” in which only directions of change and not quantitative magnitudes were sought.

A Good Alibi Is Almost as Good as a Good Investment

Why has Wall Street gone where even economists fear to tread? On Wall Street, money managers are salesmen first and investors second. In selling, nothing succeeds like a good pitch. In a bad market, a good alibi can be a substitute for a good investment. Wall Street has watched economists work for years and has their rap down cold. “Those trades should have worked – just look at this beautiful correlation!” It hardly matters that Wall Street apparently doesn’t understand its own sales talk, since it is the public and the news media it is trying to convince.

“Some traders say they expect the historical relationship [between the dollar’s exchange value and oil prices] to revert to the mean as soon as Iranian tensions cool,” remarked Mr. Berthelsen in the Journal. Taken at face value, this is gibberish. The phrase “revert to the mean” applies to a true statistical relationship, such as the heights of large numbers of human beings or repetitive flips of a true coin. It is nonsense in this case. But it sounds impressive, as pseudo-science often does.

The fact that economists promise to get results and are not punished for failure is reassuring to would-be imitators on Wall Street. And the fact that failure is recompensed with bailouts creates a “heads I win, tails I break even” mentality. Since we can’t be sure that the hedge funds understand what they’re doing, we also can’t be sure they really believe what they’re saying. On the one hand, the article reports, Mr. Kilduff of Again Capital sometimes traded daily based entirely on changes in the dollar’s exchange value. On the other hand, he changed his strategy in December when oil prices began to rise – presumably because he started losing money. Whether this bespeaks cynicism or naïveté, the implications are discouraging.

If the folly of economists were contained within the profession, it would hold little more than intramural interest. But the spillover to Wall Street soon engulfs Main Street, too. This tends to discredit financial markets and free markets generally.

When that happens, we have reached the terminus of F. A. Hayek’s “road to serfdom.”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s