Archive for the ‘Modeling’ Category

Economics and MAS

Wednesday, September 9th, 2009

Some recent articles by economists about the need for new [tag]economic theories[/tag] to replace neo-classical (aka mainstream, aka autistic) economics:

From Krugman’s article:

“As I see it, the economics profession went astray because economists, as a group, mistook beauty, clad in impressive-looking mathematics, for truth. Until the Great Depression, most economists clung to a vision of capitalism as a perfect or nearly perfect system. That vision wasn’t sustainable in the face of mass unemployment, but as memories of the Depression faded, economists fell back in love with the old, idealized vision of an economy in which rational individuals interact in perfect markets, this time gussied up with fancy equations. The renewed romance with the idealized market was, to be sure, partly a response to shifting political winds, partly a response to financial incentives. But while sabbaticals at the Hoover Institution and job opportunities on Wall Street are nothing to sneeze at, the central cause of the profession’s failure was the desire for an all-encompassing, intellectually elegant approach that also gave economists a chance to show off their mathematical prowess.

Unfortunately, this romanticized and sanitized vision of the economy led most economists to ignore all the things that can go wrong. They turned a blind eye to the limitations of human rationality that often lead to bubbles and busts; to the problems of institutions that run amok; to the imperfections of markets — especially financial markets — that can cause the economy’s operating system to undergo sudden, unpredictable crashes; and to the dangers created when regulators don’t believe in regulation.

It’s much harder to say where the economics profession goes from here. But what’s almost certain is that economists will have to learn to live with messiness. That is, they will have to acknowledge the importance of irrational and often unpredictable behavior, face up to the often idiosyncratic imperfections of markets and accept that an elegant economic “theory of everything” is a long way off. In practical terms, this will translate into more cautious policy advice — and a reduced willingness to dismantle economic safeguards in the faith that markets will solve all problems.

 . . . .

Such Keynesian thinking underlies the Obama administration’s economic policies — and the freshwater economists are furious. For 25 or so years they tolerated the Fed’s efforts to manage the economy, but a full-blown Keynesian resurgence was something entirely different. Back in 1980, Lucas, of the University of Chicago, wrote that Keynesian economics was so ludicrous that “at research seminars, people don’t take Keynesian theorizing seriously anymore; the audience starts to whisper and giggle to one another.” Admitting that Keynes was largely right, after all, would be too humiliating a comedown.

And so Chicago’s Cochrane, outraged at the idea that government spending could mitigate the latest recession, declared: “It’s not part of what anybody has taught graduate students since the 1960s. They [Keynesian ideas] are fairy tales that have been proved false. It is very comforting in times of stress to go back to the fairy tales we heard as children, but it doesn’t make them less false.” (It’s a mark of how deep the division between saltwater and freshwater runs that Cochrane doesn’t believe that “anybody” teaches ideas that are, in fact, taught in places like Princeton, M.I.T. and Harvard.)

Meanwhile, saltwater economists, who had comforted themselves with the belief that the great divide in macroeconomics was narrowing, were shocked to realize that freshwater economists hadn’t been listening at all. Freshwater economists who inveighed against the stimulus didn’t sound like scholars who had weighed Keynesian arguments and found them wanting. Rather, they sounded like people who had no idea what Keynesian economics was about, who were resurrecting pre-1930 fallacies in the belief that they were saying something new and profound.

And it wasn’t just Keynes whose ideas seemed to have been forgotten. As Brad DeLong of the University of California, Berkeley, has pointed out in his laments about the Chicago school’s “intellectual collapse,” the school’s current stance amounts to a wholesale rejection of Milton Friedman’s ideas, as well. Friedman believed that Fed policy rather than changes in government spending should be used to stabilize the economy, but he never asserted that an increase in government spending cannot, under any circumstances, increase employment. In fact, rereading Friedman’s 1970 summary of his ideas, “A Theoretical Framework for Monetary Analysis,” what’s striking is how Keynesian it seems.

And Friedman certainly never bought into the idea that mass unemployment represents a voluntary reduction in work effort or the idea that recessions are actually good for the economy. Yet the current generation of freshwater economists has been making both arguments. Thus Chicago’s Casey Mulligan suggests that unemployment is so high because many workers are choosing not to take jobs: “Employees face financial incentives that encourage them not to work . . . decreased employment is explained more by reductions in the supply of labor (the willingness of people to work) and less by the demand for labor (the number of workers that employers need to hire).” Mulligan has suggested, in particular, that workers are choosing to remain unemployed because that improves their odds of receiving mortgage relief. And Cochrane declares that high unemployment is actually good: “We should have a recession. People who spend their lives pounding nails in Nevada need something else to do.”

Personally, I think this is crazy. Why should it take mass unemployment across the whole nation to get carpenters to move out of Nevada? Can anyone seriously claim that we’ve lost 6.7 million jobs because fewer Americans want to work? But it was inevitable that freshwater economists would find themselves trapped in this cul-de-sac: if you start from the assumption that people are perfectly rational and markets are perfectly efficient, you have to conclude that unemployment is voluntary and recessions are desirable.

Yet if the crisis has pushed freshwater economists into absurdity, it has also created a lot of soul-searching among saltwater economists. Their framework, unlike that of the Chicago School, both allows for the possibility of involuntary unemployment and considers it a bad thing. But the New Keynesian models that have come to dominate teaching and research assume that people are perfectly rational and financial markets are perfectly efficient. To get anything like the current slump into their models, New Keynesians are forced to introduce some kind of fudge factor that for reasons unspecified temporarily depresses private spending. (I’ve done exactly that in some of my own work.) And if the analysis of where we are now rests on this fudge factor, how much confidence can we have in the models’ predictions about where we are going?”

 

 

The importance of conceptual models

Monday, November 24th, 2008

J. Doyne Farmer of the Santa Fe Institute has a book review in a recent issue of Nature (“The two cultures of Wall Street“, volume 456: 173-174, 13 November 2008), of Jeremy Bernstein’s book:  “Physicists on Wall Street and Other Essays on Science and Society” (Springer 2008).

“Quantitative [tag]hedge funds[/tag] tend to divide into those run by economists and those run by scientists from other disciplines, such as physics, maths or computer science. For example, perhaps the most successful hedge fund, the New York-based firm Renaissance, has some 70 researchers, none of whom is an economist. By contrast, LTCM’s only connection to physics is that [physicist-turned-trader Emanuel] Derman once applied for a job there and did not get it.

This distinction is not just a matter of professional pride and disciplinary boundaries. Bernstein misses an important point: economists and physicists traditionally approach the problem of [tag]risk control[/tag] in different ways. Risk control is the art of determining the likelihood of large and unexpected price changes happening in the future. It is well known that extremely large changes, and financial crashes in particular, are more frequent than would be expected from a ‘normal’ statistical distribution. Physicists tend to favour a ‘power law’ mathematical description to model the [tag]heavy tails[/tag] of these distributions, giving a pessimistic view of the likelihood of large price movements. By contrast, the economists who led LTCM spoke about price movements in terms of standard deviations, a terminology that is only relevant for [tag]normal distributions[/tag]. This demonstrates that they were not thinking about the problem in the right way.”

 

“Economics needs a scientific revolution”

Saturday, November 8th, 2008

That is the title of a recent essay in Nature magazine by econo-physicist Jean-Philippe Bouchaud, available here (Nature, 455: 1181, 30 October 2008)     The essay is superb and repeats what people have been saying about economics since – oh, since at least the time of Karl Marx.    One of the jokes of marketers is that marketing only exists to the extent that the assumptions of economics are false:  an entire professional discipline exists in the space which economics ignores.  Perhaps the [tag]financial crisis[/tag] will cause some economists to jettison their assumptions, and allow mainstream [tag]economics[/tag] to rejoin the reality-based community.  However, it must be said that economists seem strongly impervious to any fundamentally-critical ideas.

What follows is an excerpt from Bouchaud’s essay.   I only disagree with him on one issue:  the role that models from [tag]statistical physics[/tag] may play in fixing mainstream economics.   Although economies can be viewed as collections of large numbers of interacting particles, as if they were clouds of gas, there is an important difference between the two:  in economies, unlike in matter, the particles are intelligent, and their interactions subject to socially-constructed (and hence dynamic) patterns, rules and cultural tropes.  Are there models in statistical physics that assume the so-called laws of nature are dynamic and that the atomic particles intelligent? If not (and I think there are not), we need new models and new approaches to modeling.  Bouchaud is right about the questions, but not about the answers. 

“Classical economics is built on very strong assumptions that quickly become axioms: the rationality of economic agents (the premise that every economic agent, be that a person or a company, acts to maximize his profits), the ‘invisible hand’ (that agents, in the pursuit of their own profit, are led to do what is best for society as a whole) and market efficiency (that market prices faithfully reflect all known information about assets), for example. An economist once told me, to my bewilderment: “These concepts are so strong that they supersede any empirical observation.” As economist Robert Nelson argued in his book, Economics as Religion (Pennsylvania State Univ. Press, 2002), the marketplace has been deified.

Physicists, on the other hand, have learned to be suspicious of axioms. If empirical observation is incompatible with a model, the model must be trashed or amended, even if it is conceptually beautiful or mathematically convenient. So many accepted ideas have been proven wrong in the history of physics that physicists have grown to be critical and queasy about their own models.

Unfortunately, such healthy scientific revolutions have not yet taken hold in economics, where ideas have solidified into dogmas. These are perpetuated through the education system: students don’t question formulas they can use without thinking. Although numerous physicists have been recruited by financial institutions over the past few decades, they seem to have forgotten the methodology of the natural sciences as they absorbed and regurgitated the existing economic lore.

The supposed omniscience and perfect efficacy of a free market stems from economic work done in the 1950s and 1960s, which with hindsight looks more like propaganda against communism than plausible science. In reality, markets are not efficient, humans tend to be over-focused in the short-term and blind in the long-term, and errors get amplified, ultimately leading to collective irrationality, panic and crashes. Free markets are wild markets.

Reliance on models based on incorrect axioms has clear and large effects. The [tag]Black–Scholes[/tag] model, for example, which was invented in 1973 to price options, is still used extensively. But it assumes that the probability of extreme price changes is negligible, when in reality, stock prices are much jerkier than this. Twenty years ago, unwarranted use of the model spiralled into the worldwide October 1987 crash; the Dow Jones index dropped 23% in a single day, dwarfing recent market hiccups. Ironically, it was the very use of a crash-free model that helped to trigger a crash.

This time, the problem lies, in part, in the development of structured financial products that packaged subprime risk into seemingly respectable high-yield investments. The models used to price them were fundamentally flawed: they underestimated the probability that multiple borrowers would default on their loans simultaneously. These models again neglected the very possibility of a global crisis, even as they contributed to triggering one.

Surprisingly, classical economics has no framework through which to understand ‘wild’ markets, even though their existence is so obvious to the layman. Physics, on the other hand, has developed several models that explain how small perturbations can lead to wild effects. The theory of complexity shows that although a system may have an optimum state, it is sometimes so hard to identify that the system never settles there. This optimum state is not only elusive, it is also hyper-fragile to small changes in the environment, and therefore often irrelevant to understanding what is going on. There are good reasons to believe that this paradigm should apply to economic systems in general and financial markets in particular. We need to break away from classical economics and develop completely different tools. Some behavioural economists and econo-physicists are attempting to do this now, in a patchy way, but their fringe endeavour is not taken seriously by mainstream economics.

While work is done to enhance models, regulation also needs to improve. Innovations in financial products should be scrutinized, crash-tested against extreme scenarios outside the realm of current models and approved by independent agencies, just as we have done with other potentially lethal industries (chemical, pharmaceutical, aerospace, nuclear energy).

Crucially, the mindset of those working in economics and financial engineering needs to change. Economics curricula need to include more natural science. The prerequisites for more stability in the long run are the development of a more pragmatic and realistic representation of what is going on in financial markets, and to focus on data, which should always supersede perfect equations and aesthetic axioms.”

 

The Emperor’s new credit default swaps

Friday, November 7th, 2008

Not only are we are in the middle of a global financial meltdown, but we are inundated with people telling us that they knew all along that it was inevitable.

Two recent articles in the New York Times are nice examples of this.

Robert Shiller talks about the problems of making such predictions from a groupthink perspective, arguing that it is hard to be the lone voice in the crowd, not least because pointing out that the market is in a speculative bubble isn’t a good career move:

We all want to associate ourselves with dignified people and dignified ideas. Speculative bubbles, and those who study them, have been deemed undignified.

Shiller’s own predictions of impending doom were based on ideas from [tag]behavioral economics[/tag] which suggests, surprise surprise, that people aren’t perfetly rational, and so there are psychological aspects to economic behavior. In other words, if models had taken these “irrational” aspects into account, the problems in the market would have been obvious.

A related point is made here, which argues that the models being used to compute the risk of the collapse of the subprime mortgage market just weren’t being applied right:

A recent paper by four Federal Reserve economists … concluded that the risk models used by Wall Street analysts correctly predicted that a drop in real estate prices of 10 or 20 percent would imperil the market for subprime mortgage-backed securities. But the analysts themselves assigned a very low probability to that happening.

Garbage in, garbage out, right? But, of course, nobody was pointing out that is was garbage when it still might have helped.

Quantifying the Landing Reaction of Cockroaches

Monday, July 7th, 2008

From a call for research proposals just issued by the [tag]European Space Agency[/tag]:

“Planetary exploration missions include an entry-descent-landing phase, where the spacecraft descends on a relatively steep trajectory to the surface. In order to minimize the loads during landing, a deceleration system is employed, consisting essentially of some or a combination of parachutes, retrorockets and airbags. Due to the communication round-trip delay time between spacecraft and ground control, external supervision of the descent is limited. As a consequence, a landing device is desired that can autonomously stabilize the descent and guide the spacecraft to a safe landing place.

In the research field of unmanned autonomous vehicles, animal models have become intensively studied models for potential technological transfer. Inspiration is drawn from various aspects such as neuronal control, aerodynamics, material properties, and actuators. Studying the neuronal control (biocybernetics) of insect flight is appealing for a number of reasons. First, in insects neurons can be addressed individually and hence, their function can be well determined. Secondly, flight control is realized in a fly-by-wire manner: Sensory data is acquired and processed in a way that only one individual steering signal for each flight situation is generated and sent downstream to the motor control. From this single signal, the appropriate motor reactions are generated. The present study aims at a ‘technological transfer’ of the neuronal architecture involved in triggering the landing reaction of steeply descending cockroaches.

Cockroaches obtain flying capacities of robustly designed wings but are rarely observed to fly. From observations of scientists we know that cockroaches use their wings mostly in emergency situations, i.e. when forced to jump off elevated spots. Once air born, the cockroaches quickly deploy their wings and use them to glide to the ground, controlling their trajectory and choosing a landing site. This task requires fast reactions and quick decisional strategies. The cockroaches’ flight system is therefore assumed to be tuned specifically towards fast landing and not to e.g. flying or take-off. In consequence, it is assumed that the cockroach model – compared to other potentially flying insects – is simpler in its neuronal architecture and hence faster in its reaction and hence displays a potential model for a biomimetic transfer to an engineered landing system. Proposing universities and research centres are encouraged to include in their proposals relevant additional scientific information or a critical analysis of these assumptions.”

 

Before electricity, we had to use water

Friday, May 9th, 2008

The New Zealand-born economist, Bill Philiips, is best known for identifying an empirical relationship between a country’s inflation rate and its unemployment, the so-called [tag]Phillips curve[/tag].  However, before becoming an economist, Phillips had been an engineer, and in 1949 he built one of the first models of a national economy, the MONIAC.  MONIAC used flows of coloured water to represent money flows through an economy, and perhaps explains (or is a reflection of) traditional economics’ obsession with distinguishing stocks from flows.  

In the 1970s, the Australian cartoonist Bruce Petty also built a physical model of a national [tag]economy[/tag], but this time with seats for several human operators, each representing The Government, The Unions, Big Business, etc.   Instead of the hydraulic flows used by Phillips, Petty’s model used mechanical levers and pulleys, which impacted in convoluted ways on the machine and on the other operators.   This model looked something built by Heath Robinson or Rube Goldberg, and was immense fun to watch it at work.   I’ve not yet been able to find a video of Petty’s model at work.