- Shopping Bag ( 0 items )
Positive feedback—when A produces B, which in turn produces even more A—drives not only abrupt climate changes, but also the most important and disruptive events in economics and finance, from asset bubbles to debt crises, bank runs, even corporate corruption. But economists, with few exceptions, have ignored this reality for fifty years, holding onto the unreasonable belief in the wisdom of the market. It’s past time to be asking how do markets really work? Can we replace economic magical thinking with a better ...
Positive feedback—when A produces B, which in turn produces even more A—drives not only abrupt climate changes, but also the most important and disruptive events in economics and finance, from asset bubbles to debt crises, bank runs, even corporate corruption. But economists, with few exceptions, have ignored this reality for fifty years, holding onto the unreasonable belief in the wisdom of the market. It’s past time to be asking how do markets really work? Can we replace economic magical thinking with a better means of predicting what the financial future holds, in order to prepare for, or even avoid the next extreme economic event?
In Forecast, physicist and acclaimed science writer Mark Buchanan answers these questions and more in building a new model for economics, one that accepts that markets act much like the weather does. While centuries of classical financial thought has trained us to understand “the market” as something that always returns to equilibrium, economies work more like our atmosphere—a loose surface balance riding on a deeper torrent of fluctuation. Market instability is as natural—and dangerous—as a prairie twister. With Buchanan’s help, we can better govern the markets and weather their storms.
"The 2008 crash created not just a crisis in the economy, but also a crisis in economic thinking. Ideas of market stability and efficiency lost their value faster than a sub-prime mortgage. In this compelling and lively account, Mark Buchanan tells the story of the new ideas that are revolutionizing the field. We may not be able to perfectly forecast the economic weather, but we can be better prepared for storms to come."—Eric Beinhocker, Executive Director of the Institute for New Economic Thinking at Oxford University and author of The Origin of Wealth
"A lucid, absorbing story"—New Scientist
“The author's stimulating deconstruction of contemporary economic theory parallels a treatment of major positive developments in physical sciences and pays due respect to the functions of government and law.”—Kirkus Reviews
The Equilibrium Delusion
Economics is a discipline for quiet times. The profession, it turns out, ... has no grip on understanding how the abnormal grows out of the normal and what happens next, its practitioners like weather forecasters who don't understand storms.
—Will Hutton, journalist The Observer, London
I think physicists are the Peter Pans of the human race. They never grow up and they keep their curiosity.
—Isidor Rabi, physicist
Like many towns sprawled across the great flat expanses of the mid-western United States, Overland Park, Kansas, is no stranger to extreme weather. Each year, in spring and early summer, warm, moist air sweeping in from the Gulf of Mexico slips under sheets of colder air tumbling in over the Rockies. "Warm air rises," that banal phrase of folk physics, here takes on life: gravity drives vast plumes of the warmer, lighter air to penetrate the colder layer above and billow upward to a height of ten miles. This is the first ingredient in a recipe for atmospheric violence. The second is surface winds, blowing to the North and from the West, which drive this incipient protostorm to rotate like a turnstile. The result—several billion tons of wet, warm air whipped into an unstable, spinning tower—creates near-perfect conditions for powerful tornadoes, some as much as a mile in diameter, with winds churning to 400 m.p.h.
To most of us, tornadoes seem freakish, unnatural, definitely abnormal. Yet the state of Kansas alone sees hundreds of them every year, and they come about through perfectly ordinary atmospheric processes. Broadly speaking, it's all just part of what happens in the atmosphere: one event builds on another, and then another, and soon an ordinary gray sky becomes a violent, memorable twister. More technically, we could say the atmosphere is prone to what scientists call "positive feedbacks," the consequences of which our human minds find hard to imagine.
Maybe you've heard the term elsewhere. Positive feedbacks are a longstanding concept in science—the process by which small variations in a given system can become increasingly large. They are commonplace in discussions of global warming. Melting glaciers turn white ice to blue water, reducing how much sunlight gets reflected back into the atmosphere; the process could accelerate planetary warming. Positive feedbacks arise in psychology, biology, electronics, physics, computer science, and many other disciplines. Yet even though many of us recognize this notion, we are dreadful when it comes to estimating its impact.
Take $1,000 and invest it in something that earns interest at a rate of, say, 10 percent each year. Leave it there for thirty years with the interest earned feeding back into the account. How much will you have? Well, 10 percent of $1,000 is $100, so you might think the amount should increase by $100 or so each year. Thirty years makes the total gain about $3,000, giving a total of $4,000. Of course, as the amount grows, you'll be getting 10 percent of a growing number each year, so you'd expect something a little more than $4,000. Maybe, without a calculator, you'd guess $5,000 or $6,000? It already starts to feel hopeful to consider $10,000. But human intuition is no match for the mathematics. The actual total after thirty years is a little over $20,000. The amount feeds on itself and grows faster than anyone would expect.
There's more here than a lesson about money; it's a lesson in human thinking and why the world so often surprises us.
As humans, we're terrible at imagining the likely consequences of positive feedback. Take a piece of paper and fold it, and then take that doubled paper and fold it again, and then again, thirty times in all. Actually, don't waste your time. You'll find you can't do it because the result, if you did, would be about seventy miles thick. Ask a friend to give you an apple today, two apples tomorrow, four the third day, and so on, for one whole month (thirty-one days). You'd better rent a large ware house because on the final day alone you'll get more than 2 billion apples. This is the power of positive feedback: each step not only makes things bigger, but also gears up the process itself, accelerating how fast things get bigger in a way that leads to consequences far beyond our expectations.
Positive feedback matters a lot more than we think, because in one form or another it lies behind almost everything that makes our world rich and surprising, changeable and dynamic, lively and unpredictable. It makes seeds sprout and grow into trees, matches burst into flame, and single cells divide and proliferate into living, thinking human beings. It drives political revolutions and new religions, and it makes perfectly peaceful blue skies give rise, with little warning, to storms of terrifying violence, like those tornado-spawning storms in Kansas. Our brains lack intuition for all this. In meteorology, and in the rest of science, it's taken years of learning from mistakes to recognize how and why positive feedbacks play such a crucial role in causing events we might not otherwise expect.
Yet outside of these areas, an intellectual blind spot to the power of positive feedbacks still holds us back. Nowhere is this truer than in the science of human systems, in social science, and especially economics and finance. Consider what happened, for example, on May 6, 2010.
Four Minutes of Mayhem
In addition to being a frequent site of tornado activity, Overland Park, Kansas, is also home to the headquarters of an important investment company named Waddell and Reed Financial, Inc. Founded on a shoestring by financiers Cameron Reed and Chauncey Waddell back in 1937, the firm started out with offices in department stores, but today it has grown to handle funds totaling more than $60 billion. It lacks a famous name, but it's big enough to make investment decisions that—with the help of positive feedbacks—can threaten the stability of the entire global economy. In less than five minutes.
In the spring of 2010, a mutual fund run by Waddell and Reed had invested heavily in futures contracts for the Standard and Poor's Stock Index, one of the most widely traded stock futures. Buying such a futures contract means that you agree to buy the S&P 500, not now, but on a certain fixed date in the future. The price you pay, however, is fixed now. They're among the simplest "derivatives" products, which "derive" their values from the value of something else, in this case the S&P 500. If that index rises in value, the value of the future also rises, as it raises the likely future value of the index, too. Waddell and Reed were deeply into these futures as a hedge or balance against other investments they had, and their strategy seemed sound until the early days of May, when financial authorities in Greece admitted (under pressure) that levels of government debt were far in excess of the limits set by the Europe an Central Bank. Suddenly, as Europe an and international bankers met to find ways to keep Greece from defaulting, the future of the Europe an monetary union was called into question. Investors worried. Between the market's opening and noon on the sixth of May, stocks of the Dow Jones Industrials slid downward by 2.5 percent.
At two thirty-two p.m., concerned that troubles in Eu rope were spreading to the United States, Waddell and Reed requested a broker at Barclays bank to get out of the stock index futures market. Using a computer program to make the trade, the broker began trying to sell $4.1 billion in so-called E-mini stock index futures. Selling has the effect of driving prices down, so the program was designed to work cautiously, selling a little at a time, spreading it over the day. For ten minutes or so, things went smoothly. But then, at two forty-one p.m., something kicked off an explosive chain of events. High-frequency traders, whose computers conduct thousands of trades per second, had been buying most of the contracts as Waddell and Reed sold them. Many of these traders make money as "market makers": their computer programs stand ready to buy or sell at any moment, and they profit from slight differences in the prices they set for buying and selling. But according to the trade-by-trade data for the day, these programs seem to have purchased too many futures contracts, accruing an inventory larger than desired. At two forty-one p.m., one of these programs decided to bail out of market making and started selling aggressively, causing futures prices to drop off a cliff.
The result was a spectacular plunge as positive feedback involving completely automated trading clipped the value of E-mini futures by more than 3 percent in only four minutes. But this was only act 1 of the unfolding drama.
This tumult in the futures market then acted like a detonator to set off act 2—an implosion in the stock market itself. Stock traders soon noticed the sudden drop in futures prices and jumped in to profit by buying up these cheap futures, meanwhile selling an equivalent amount of the same stock. In effect, the S&P 500 five months in the future could be had for a fraction of the index price now; with no particular reason to think the index would rise or fall strongly over those months, investors dumped the current stock in favor of the future. In a few minutes, the sheer volume of trading soared so violently that automatic protection rules for computers trading on the New York Stock Exchange and other exchanges kicked in to shut down trading. Stock prices went into free fall as sellers could find no one willing to buy. The blue-chip stock of Procter and Gamble lost a third of its value in three and one-half minutes; Accenture stock went all the way to less than a penny per share. In all, the Dow Jones Industrial Average lost 9.2 percent of its value in a few minutes—the largest drop in such a short time ever.
Almost as quickly, the markets recovered, closing the day with the Dow and the stocks of most companies back within a few percentage points of where they started. It was as if the markets had, like a jet liner, flown into a pocket of devastating turbulence and plunged several thousand feet in a terrifying death spiral, before recovering, climbing, and carrying on as before.
What happened? In the days and weeks after this "flash crash," popular speculation pointed to a computer error, or perhaps a "fat finger"—a trader hitting the wrong key and initiating a mistaken trade. An article on CNBC's website reported a rumor that someone had typed b for billions rather than m for millions when making a trade on Procter and Gamble (although one blogger noted that it must have been a very fat and strangely shaped finger to miss b and hit m without touching the n that lies between). Others worried that some powerful financial genius had manipulated the market for his or her own gain, luring the computer traders to cause the crash so they could profit.
Investigations by the Securities and Exchange Commission over the next five months found no evidence to support any of these ideas. What it did find, after sifting through mountains of data detailing trades in many markets during the day, was that the Waddell and Reed trade seemed to be the primary trigger for the turmoil, which then took off with an energy all its own. The SEC-CFTC final report on the event—entitled "Findings Regarding the Market Events of May 6, 2010"—documents a very complicated set of happenings that are nevertheless fairly simple in conceptual outline, if you think in terms of positive feedbacks. The event had two or maybe three principal stages, and, like all positive feedbacks, culminated in consequences far beyond what anyone might have expected.
But perhaps the most interesting thing about the SEC-CFTC report is that it never uses the words "feedback" or "instability," makes no analogy to storms in the atmosphere, or indeed to self-reinforcing processes of any kind. Like a food critic describing a sumptuous meal without mentioning flavor, texture, or color, the SEC report struggles to describe positive feedbacks running through the markets without ever mentioning the basic idea. It's a peculiar but telling contortion, which points to the singular weakness in most current thinking about markets and other economic systems.
There are two reasons why the SEC report doesn't use the natural language of positive feedback. First, doing so would immediately undermine one of the report's principal aims, which was to reassure rattled investors that this event was an unusual and exceptional episode, a one-off freak occurrence unlikely ever to happen again. (In the time since the SEC report, there have been further flash crashes, albeit on a somewhat smaller scale; to take just one example, the stock for the computing giant Apple dropped by 9 percent in five minutes on March 23, 2012. Its share price had already substantially recovered by close of business that day.) Investors would like to hear that an evil perpetrator had orchestrated the whole thing, made millions out of it, but has since been identified and arrested. Or at least they want to hear that it all came out of one big trade and a conspiracy of other events that probably won't come together again. Problem solved. Talk about natural feedbacks and instabilities, explain how the event emerged out of quite ordinary happenings in the market, and you're not likely to do much reassuring, even if you do strike much more closely to the truth.
The financial press for the most part played along with this view. They dutifully reported that the big Waddell and Reed trade had kicked off the crash, and in some vague way had probably caused it, though without ever explaining quite how. LONE $4.1 BILLION SALE, as the New York Times reported, LED TO "FLASH CRASH" IN MAY.
The second reason the SEC report failed to mention positive feedback is more fundamental, and so holds more importance for the long run. Economics in general—and finance in particular—have long been based on notions of balance and equilibrium, on the idea that the economy at large and financial markets in particular naturally tend toward a state of balance. Any disturbance or shock is thought to stir forces—negative feedbacks—that will bring the system back into balance. Negative feedback is what makes the water stirred in a cup settle back to rest, as the initial action stirs forces that act against it, slowing the flow and restoring stillness. Markets, economists insist, work similarly. Through a long and circuitous path, the idea goes back to Adam Smith's famous notion of the invisible hand, which allegedly leads markets to good outcomes despite the manifold conflicting and typically selfish motivations of the people involved. The notion of positive feedbacks is rarely, if ever, considered in potential explanations of economic happenings.
The SEC report on the flash crash doesn't use the most natural language to talk about the event because that language goes against the core concepts of mainstream economic thinking, especially as it has developed over the past few de cades.
Instead, it follows a very different approach, not really trying to explain at all. The report documents how the Waddell and Reed trade led to the flash crash through a plausible chain of events, one causing the next, never once asking why such an explosive chain of events was possible on this particular day, why this one large but hardly unusual trade could have sparked it off, or why anyone should think it was unlikely to happen again, possibly with much more severe consequences. After all, had the trade been made a couple hours later, the market might have closed at the end of the day down one thousand points, triggering similar crashes all around the globe.
Equilibrium thinking is so deeply ingrained for most economists that they find it difficult to think in other terms. Standard economic analyses always begin by identifying the aims of the relevant parties involved in some situation, and what they stand to gain or lose from taking various possible actions. The economist then works out mathematically—and the mathematics are basically always the same4—the details of the balance point or equilibrium that will supposedly emerge out of the competition between all these parties. Almost all economic policy advice comes from such exercises, as economists analyze how the properties of the supposed equilibrium would change if the government raises taxes, for example, or changes regulations on CO2 emissions—or how the financial markets would quickly adapt to a new stable equilibrium if a trader in Kansas suddenly starts selling lots of futures. But what if, rather than settling back into a nice equilibrium, the markets instead fall into a condition more akin to a rag soaked in gasoline or like the atmosphere in summertime Kansas, inherently prone to stormy chaos? There is nothing in all the usual calculations to account for that.
Belief in equilibrium confers a sense of safety and predictability, and reflects a firm belief in the triumph of human logic over nature. The assumption of equilibrium sharply restricts the kinds of things economists can imagine happening in the world. This limitation most likely explains why the SEC for months sought to explain the flash crash as the result of something exceptional and abnormal, from a mistake in the market, a fat finger, or computer glitch, because it just couldn't have come from the ordinary, internal workings of the market, which are always supposed to remain in stable equilibrium.
Excerpted from FORECAST by MARK BUCHANAN. Copyright © 2013 by Mark Buchanan. Excerpted by permission of BLOOMSBURY.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
1. The Equilibrium Delusion.................... 1
2. A Marvelous Machine.................... 24
3. Notable Exceptions.................... 45
4. Natural Rhythms.................... 70
5. Models of Man.................... 89
6. Ecologies of Belief.................... 112
7. Perils of Efficiency.................... 133
8. Trading at the Speed of Light.................... 156
9. Twilight of the Idols.................... 178
10. Forecast.................... 203