Statistics and Individuality Jude Wanniski May 28, 1999

Last week's lesson on "Macroeconomics: the Enemy Within," invited an interesting discussion among several SSU students in the TalkShop. It led in directions that became more interesting, even provocative. I'd written a number of pieces in recent years about the philosophy of markets, but when I walked through the discussion below, I found myself considering new ways to think about markets -- what they are, how they happen, and how they work. The string began with the thoughts of James Prather, a physicist. The entries have been edited only to make them flow:

James Prather: The basis for Jude's debunking of Nobel Prize winning economic science is that there is no basis for applying "statistics" to economics. [In my realm], statistics applicable to physics and chemistry were largely developed at the beginning of the 20th century to deal with 'classical' [e.g. ideal gas laws] and 'quantum-mechanical' [e.g. radioactive decay] (1) 'ensembles' of (2) identical 'particles' faced with making (3) 'decisions' from an (4) identical 'menu' of 'choices.' The critical requirements for the application of 'statistics' to any 'science' are that (a) the decision makers be indistinguishable from each other and (b) the decision-makers menus be identical. NEITHER of these critical REQUIREMENTS are met for any microeconomic system and hence no macroeconomic system based upon 'statistical' arguments can possibly be correct.

All 'statistical' laws are ex post facto derived and can be applied to 'predict' future 'behavior' of a system if and only if [A] that new system consists of decision makers Indistinguishable [1] not only from each other but [2] from the decision makers whose previous behavior resulted in the development of the statistical 'law' and [B] the Menu for the subsequent set of decision makers is identical to the menu of 'choices' of the original system.

Scott Robinson: Very interesting. Has anybody created a computer universe where the above preconditions are not met, but yet behave according to statistical models? Would be an interesting Ph.D. Thesis. (Where would one find an advisor for this?)

Donald Luskin: The Santa Fe Institute is the center for studying economics in the context of "complexity." This approach, at its best, allows the techniques of statistics and computer-modeling to be applied to economics and other behavioral phenomena -- without the arbitrary constraints of classical approaches that suffer from all the flaws that James mentions.

Dick Fox: Determinism has become the pop religion of our time. Scores of scientists are looking for genes that determine behavior. Then every day you can hear in the media claims that a new survey has proven this or that. Experimental science has taken a back-seat to "survey" science. This trend is not unique to economics. With all this pseudo-science going on, is it surprising that economics is treated the same way?

James Prather: I am well aware of the myriad of things that SFI worthies have had their fingers in. There is an interesting piece in the April 26 New Yorker about SFI Alumnae Farmer and Packard and their SBC chaotic market venture. [Author Bass also wrote an entertaining book about their earlier roulette venture called something like The Eudomenaic Pie.] I had supposed that Farmer-Packard were treating markets as 'self-organizing systems' a la Ilya Prigogine -- systems that are far from equilibrium and that can 'bifurcate' -- and have not had the sort of success over the past 8-9 years that they and their backers had hoped for. But that may be because markets are somewhat like the weather in that, no matter where you are, the weather tomorrow will be very much like the weather today at least 2/3 of the time. [An example of 'self-organization' is the development of a tornado and the weather conditions that allow for such 'bifurcations' are fairly well known. But that doesn't mean that tornados are predictable. Nor would identifying the basic conditions necessary for a market 'bifurcation' mean that you could predict one.] Enrico Fermi is reported to have won lots of bets because he realized that most complicated systems -- those not far from equilibrium -- do not and can not change as rapidly as most people expect. That's one reason why Trend Analysis works, and there are probably traders who have made more money following trends in the past 8-9 years than those expecting order out of chaos.

Lan Nguyen: Martin Armstrong of PEI has a very interesting model based on 2 laws -- the conservation of energy and the 2nd law of thermodynamic which generates more correct calls on financial markets than anything else I ever see. I have ventured myself in these two areas to broader my understanding of his model and I think nothing can escape these 2 laws, markets included if we treat the globe as a closed system until we found another planet to trade with.

Jude Wanniski: There are folks at CalTech who track IPO prices and find them following a curve that rises and falls before rising again. The suggestion is that you wait for the dip before you buy, to catch the second wave. Of course, if the first rise is humongous and the dip is slight, you have missed out on most of the asset's appreciation. Because I think of the market as an aggregation of individuals, I think of asset prices as being the result of myriad CONVERSATIONS. The value of a stock can rise or fall because some single new individual has decided to join the conversation, after looking over the enterprise behind the paper asset.

My friend Michael Milken is such a person, who can see value long before anyone else does, because he has a gift for that sort of thing. The feds did not believe he had the gift so they put him in jail for insider trading when it is ridiculous to think that after amassing $2 billion by seeing the future, he would try to make a few extra dollars with Ivan Boesky. I told Milken he is the only man I know who gets The Wall Street Journal delivered on his doorstep the day before anyone else gets it. My point: When you ask Milken what causes economic growth generally, he more or less says that HE does. That is, he is able to tell in advance what enterprises will succeed with what managers with what capital structure. He does not need a lower capital-gains tax as an incentive. He works because he loves to work and create.

In the same way, Ted Forstmann of Forstmann, Little & Co. has been a client and a friend for almost 20 years, since he first heard about what I do and called and asked to be a client. Ted knows nothing about the stock market, how it works, or really what it does. He would not think of reading the papers, looking for a stock to buy for his portfolio. He looks at entire companies, studying them by the dozens until he finds one that is obviously screwed up in some way that nobody else has noticed. Like Milken, he is a financial genius, because he can see a pile of manure and know that underneath is a perfectly good pony. Where Milken is superior at understanding how to finance a person and his enterprise idea that nobody else sees value in -- junk companies, junk people -- Forstmann buys the whole company and reorganizes it himself, then sells it at 10, 50 or 100 times what he paid for it.

Alas, there are not enough Milkens and Forstmanns to go around, so society has to operate on rules and regs that invite masses of people to individually take risks, with most of them losing at least the first time around. (If at first you don't succeed, etc.) In other words, there is no substitute for insight. I've become a modestly wealthy man, from a standing start of almost zero net worth 25 years ago, by giving my little bits of capital to those asset managers who were my clients who were paying most attention to what I was saying about the macro political economy. I sell Polyconomics as if it were a weather forecaster, and over the years I've done better than all the competition. A better mousetrap and the world beats a path to your door.

Scott Robinson: I had a conversation with a friend I reacquainted twenty years after playing soccer together at Berkeley. We both have daughters the same age going to the same school. He is retiring as an accountant to go into farming full time. His notion of the successful business person is the one who makes the correct decision (first) on the least amount of information. I liked that one.

Steven Piraino: I don't understand how Jude's supply model is any less mechanistic than Keynesian macroeconomics. The fact that we take account of people's responses to marginal incentives does not make us less mechanistic, only differently mechanistic. I think it is fair to say we put much less faith in the statistical aggregates upon which the study of macroeconomics is based, but this wasn't stressed in "Macroeconomics: the Enemy Within."

* * * * *

Next, we will read Rich Karlgaard's column in the May 27 issue of Forbes. The new editor of Forbes, Karlgaard replaced the legendary Jim Michaels, who retired last year after several decades at the helm. Unlike Fortune and BusinessWeek, where the culture revolves around the corporation, Forbes has always focused on the individual, the entrepreneur. I think Karlgaard's biweekly column, which always is available online at [www.Forbes.com/Karlgaard], is the best in the public prints. In this, he refers to two previous columns that are equally stimulating, but this one fits nicely with our theme of the value of the individual. David still can defeat Goliath. No matter how thoroughly you dominate your market, or how many billions you have in the bank, you still are vulnerable to the new kid on the block. The bigger they are, the harder they fall.

THE HUMAN ROLE By Rich Karlgaard

MY LAST TWO EDITORIALS squinted into the bright new business dawn now upon us, glowing with the light of microchips and digital bandwidth. These twin forces rise at exponential rates toward infinite capability and zero cost. I call this phenomenon and its likely economic effects, the Law of Infinite and Zero.

The law, of course, is not a real law, such as Newton's laws of gravity and motion. It is not E=mc2. A real law is perfectly predictive; mine is not. Why? Because to talk about economics, you must talk about man—the creator, manager and consumer of all economic goods.

It appears that silicon chips are fated to evolve like crazy. They get twice as fast every 18 months and will continue merrily along that path. Then there's communications bandwidth, once a crawler in comparison. But during this decade the old slug discovered fiber optics, coaxial cable, spread spectrum (its digital wings!) and the key to slipping out of the regulatory cocoon. Now bandwidth progression roars along two to ten times faster than silicon chips. But when you talk about man—compared to chips and bandwidth, man is still mired in the primordial ooze. Has man's nature evolved so much as a jot or a tittle in 5,000 years? Doubtful. I can never read about the Enlightenment without thinking its leaders of arts, letters, music and politics were superior to ours. Now there's a sour thought.

But I digress. The point is that Infinite and Zero is not a real law. Nor is it a Marxist historical inevitability, devoid of man's bumbling or heroics. Rather, it is a force, a prevailing wind blowing mankind into the next millennium.

Character and Leadership

Man is yet at the wheel; he still counts. A few weeks ago Compaq's chairman, Ben Rosen, sacked his CEO, Eckhard Pfeiffer. The news shocked and roiled. Pfeiffer was a business superstar, by any measurement. FORBES had named Compaq its company of the year for 1997. That was Pfeiffer's doing. He was mythical—a big, world-class CEO.

But a year ago the careful Compaq watcher began to see disturbing signs. Trouble began after Compaq acquired Digital Equipment Corp. and Tandem Computers. In theory, the deals made good sense: Digital would propel Compaq upmarket into higher margins; Tandem would give Compaq chops in the lucrative telecom sector. Together, the acquisitions would beef up Compaq to attack IBM and Hewlett-Packard. But Compaq goofed by leaving its PC flank uncovered. Michael Dell and others strode in with their Web direct-sales model and changed the rules. Compaq looked stunned, paralyzed by its channel conflicts, distracted by IBM and HP. It could not react fast enough.

Out of Touch?

Compaq had not lost its touch, but something more profound—its credibility with Wall Street. Last year, when the hard work of folding Digital into its operations was proving tougher and costlier than expected, the Street felt that Compaq started playing cute with its financial reporting. A brand-name analyst warned me then that Compaq was stuffing the sales channel—recording shipments to middlemen as sales. Compaq had not done anything illegal, but it had relaxed its once-higher standards. The Street and Compaq were estranged. That's fatal for a company (or a country) today. Chairman Rosen, wisely, stepped in.

The Law of Infinite and Zero ordained that the Web would emerge as a lower-cost sales channel. When it happened, Compaq was trapped in a typical 1999 incumbent's dilemma, beholden to the old sales channel. I think Pfeiffer could have worked it out. He just needed Wall Street's patience. But cute accounting had killed patience. That's a human error, not an inevitability.

By all rights, IBM shouldn't hold together as an $84 billion colossus against the centrifugal forces of Infinite and Zero. Such analysis ignores the greater force of Louis Gerstner, the CEO. Sun Microsystems should have fallen long ago to Intel and Microsoft and the "inevitable" volume economies of PCS. But that discounts Scott McNealy, the most foxy guerrilla leader since Ho Chi Minh. For years McNealy held Sun together seemingly by loud remarks and spitballs. He hung on until the prevailing winds changed. They now blow at Sun's back.

The Law of Infinite and Zero is at work, and winds do blow, but nothing is inevitable. Leadership and character still matter greatly.

* * * * *

Finally, I had another thought about markets as I reviewed the TalkShop material. It is that it does take bigger and broader and more open markets to determine better values. An auction by itself only tells you the consensus of the people who attended the auction. You don't even need an auction to determine value. That is, money does not have to change hands for you to know that something is worth more, or less, than it was a minute ago or a week ago.

Imagine a hundred people in a room waiting to bid on a Van Gogh, for example, the numbers in their minds jumping around in the millions, perhaps. It only takes one other person coming into the room to look at the picture and pronounce it a copy -- and as long as those in the room recognize his or her authority -- the millions in their heads dissolve to hundreds, if that. Most people on Wall Street don't understand that simple concept, believing that prices are determined by cash flow. The financial press is sodden with comments from experts saying that the bond market went down because too many people were selling bonds. Or that stocks went up because people are putting extra cash into mutual funds. These are all silly arguments to most supply-side analysts. When I wrote about the stock market crash of 1929 as being the result of the growing likelihood of the passage of the Smoot-Hawley Tariff Act, I had eminent Ph.D. economists tell me it could not have been the case, because nobody was quoted in the papers as having identified Smoot-Hawley as something that would cause such a decline. Most people who sold in that avalanche did not know why they were selling and almost none of them could tell you what the tariff act was all about. Someone saw it before anyone else did and sold his shares because of it, and the others in the market conversation sold until the value of the nation's capital stock had reached a new equilibrium. In other words, someone saw a pile of glitter and realized there was manure underneath.

Indeed, years after I presented my hypothesis, someone sent me copies of pages from an old autobiography of E.F. (Bud) Hutton, who recalled that as a young broker in Seattle (I think), he read the Dow Jones ticker in his office referring to the tariff act, weeks before the crash, went to his phone, and advised all his clients to sell. He didn't really explain his decision in terms of a macroeconomic model. It was simply a bad feeling he had about it. Thereafter, when E.F. Hutton spoke, people listened.