Image: Illustration by Kyle Bean
In Brief
- Sophisticated models used by investment firms to calculate risk contributed to the market crash of 2008.
- Despite their ubiquity, these risk models fail to take into account important forces that affect the market.
- Researchers are building ways to work around these limitations and prevent a repeat market crash.
- Yet these strategies may limit profits, making it unlikely that banks will adopt them without being forced to do so.
The market crash of 2008 that plunged the world into the economic recession from which it is still reeling had many causes. One of them was mathematics. Financial investment firms had developed such complex ways of investing their clients’ money that they came to rely on arcane formulas to judge the risks they were taking on. Yet as we learned so painfully three years ago, those formulas, or models, are only pale reflections of the real world, and sometimes they can be woefully misleading.
The financial world is not alone, of course, in depending on mathematical models that aren’t always reliable for decision-making guidance. Scientists struggle with models in many fields—including climate science, coastal erosion and nuclear safety—in which the phenomena they describe are very complex, or information is hard to come by, or, as is the case with financial models, both. But in no area of human activity is so much faith placed in such flimsy science as finance.
This article was originally published with the title A Formula for Economic Calamity.
Already a Digital subscriber? Sign-in Now
If your institution has site license access, enter here.
See what we're tweeting about
22 Comments
Add CommentFormulas should be banned!
Reply | Report Abuse | Link to thisThank you David, for the article. I would like to see more like this please.
Reply | Report Abuse | Link to thisI want to challenge your statement, "The only real option is not to trust the models, no matter how good the equations seem to be in theory." Unfortunately, many use this as an excuse for inaction and many systems -- like banking and the climate -- bring peril in response to inaction. Rather, the challenge is to recognize that the models produce results with uncertainty that needs to be embraced and understood. Uncertainty, like mistrust, is not another excuse for inaction. And remember, uncertainty cuts both ways -- results can be better or worse than a single output number. The goal is to take responsible actions in light of uncertainty.
I also want to respond to "The price of this reasonable caution, Avellaneda notes, will be a system that doesn’t operate as efficiently—in other words, investors will get less rich off it, on average." Talib reminds us in Black Swan that model-driven, highly leveraged strategies (and banks) are brittle. If you observe performance over a period longer than a quarter and long enough to include the consequences of brittle strategies -- those are the busts -- then investors are already getting less rich returns. My 401K has a ten-year performance hole in it, the so-called lost decade, that sadly demonstrates this phenomenon.
"Yet as we learned so painfully three years ago, those formulas, or models, are only pale reflections of the real world, and sometimes they can be woefully misleading..."
Reply | Report Abuse | Link to this"But in no area of human activity is so much faith placed in such flimsy science as finance."
Is this a parody of Scientific American? The world's climate is easily ten orders of magnitude more complex
than the "financial climate."
And yet all six billion of us are expected to uniformly and without question put all of our faith in these models, which are designed by a relatively small handful of "scientists" with a decidedly leftist political world view.
The article would be laughable if it weren't so tragic.
I disagree. The complex derivatives and such look to me to have been intentionally created to be impossible to understand by anyone except the creators. (It is much easier to create a Gordian Knot than to untie one). As a wonderful little article in S.A. pointed out many years ago, financial markets are zero-sum games: for one person to do better than the overall market, someone else has to do worse. And the only way to win a zero sum game is to know more about it than the other guys. One of the ways is to be more knowledgeable is to create something that only you understand. Another popular way is to generate false information so others make incorrect decisions. Given the motivation some people have to "win" by accumulating more money than their compatriots, it is no wonder that crashes occur regularly.
Reply | Report Abuse | Link to thisHere is an excellent video and paper from MIT that follows the general theme and provides additional specific examples.
Reply | Report Abuse | Link to thisVIDEO: "Warning: Physics Envy May be Hazardous to Your Wealth"
mitworld.mit.edu/video/794
ACADEMIC PAPER: "Warning: Physics Envy May be Hazardous to Your Wealth!" (See "one-click download" at top)
papers.ssrn.com/sol3/papers.cfm?abstract_id=1563882
Excellent examples throughout. My personal favorite is the coin-tossing machine by Diaconis, Holmes, and Montgomery (2007). Coin tossing is random, right? WRONG! Their carefully adjusted machine tosses a coin that lands heads up, 100% of the time. It's an important lesson on checking your assumptions at the door.
ARTICLE: "Overheard at MIT: Why Economics Isn't Like Physics"
sloanreview.mit.edu/the-magazine/2010-fall/52113/why-economics-isnt-like-physics/
Add to this the other trend toward extremely high-speed trading using networked supercomputers and you can see the obvious potential for calamity.
Reply | Report Abuse | Link to thisSo, the market fundamentalism that has advocated lack of regulation and control has had no impact? Just bad mathematics? In other words, is it a bad application of an otherwise good system, or a systemic error, that was bound to happen?
Reply | Report Abuse | Link to thisModels do not cover unpredictable events well. But there are financial experts and mathematicians at companies who calculate risk.
Reply | Report Abuse | Link to thisThese people must have known that the system was set to destruct. Probably, there was no good action that could be taken to prevent destruction.
Could a perceptive AIG dump all its risks at a late date without alerting the market?
As long as finance professionals mainly come out of math, physics, and engineering, these sorts of debacles will continue to occur. In July 1998 I published a paper in <i>Risk</i> in which I wrote "The array of powerful statistical techniques available to the risk manager… are founded on quicksand." One month later, Long Term Capital Management, with two economics Nobelists on its board of directors, collapsed, inducing a liquidity crisis that was a foreshadowing of 2008.
Reply | Report Abuse | Link to thisMarkets are not physical systems. They are complex adaptive human social systems, and finance is most assuredly not physics.
I'd like to read your paper. Do you have a link?
Reply | Report Abuse | Link to thisThere are several (including the Risk paper) at http://itrac.com/overview.htm
Reply | Report Abuse | Link to thisThis one is perhaps most representative of my views about physics-based models of complex adaptive systems:
http://itrac.com/paper/PhysicsResponse.doc
There are several, including the Risk paper, here:
Reply | Report Abuse | Link to thishttp://itrac.com/overview.htm
Most representative of my general view of using physics-based models in finance is this unpublished piece:
http://itrac.com/paper/PhysicsResponse.doc
(And I may have screwed up a posting so this may appear twice.)
The probability of a force 5 hurricane hitting New Orleans is different if a force 4 hurricane is in the gulf then if there are no hurricanes in the area.
Reply | Report Abuse | Link to thisThe risk manager needs to consider a strategy to follow if there is an extreme hurricane pointed in the general direction of New Orleans.
Conditional probabilities change as conditions change.
When economic conditions changed, risk managers would see that the likelihood of financial doom increased.
Yet there was no warning.
I liked your two papers that I read.
It is naive and crazily intrusive to blithely suggest that each and every transaction be reported to some central data collectors.
Reply | Report Abuse | Link to thisFor one... this is like trying to control the flow of a river by barcoding and tracking each drop of water. You have to be wholly insane to think this a practical method.. even if, by some abstract form or reasoning it seems rational.
Additionally it is insanely naive to think that this info would remain confidential. The IRS constantly is bagged for failures in data security and there is far less money to be made from those leaks than there would be from transaction data collection as suggested in the complete article.
One of the simplest solutions would be to simply get the "damned fools" out of the business of money-handling. If you had access to the inner workings of banks by informants deep inside them (as I have in the past) you would know it is a nightmare of people who almost know what they are doing... but not quite.
So much of the problem would simply disappear via market forces if true morons were not so ready to think they can gain something-for-nothing ... and this is not just the bankers but those who invest in banks.
we also might wonder about the term "collapse" as if this were some cataclysm. Is it really a big problem if a Ponzi scheme finally comes to an end? What is a problem is if it continues and grows.... then the inevitable is worse.
Let the fools fail... quit rescuing them... and understand the the real physical 3-D wealth, will remain exactly the same before and after a collapse of the illusory paper economy and its imaginary value.
"Perceived wealth" unbacked by any physical assets (which assets includes 'humanpower' and human skills in the workforce)is pure baloney, nonsense and is less concrete than the water vapor that makes up fog.
So long as fools continue to envy and greedily collect "perceived wealth".. wealth backed by less asset value than the vacuum of interstellar space.... we will have an economy that is as ready to disintegrate as any other pipe dream.
You can't usefully evaluate a model without knowing what it will be used for. Freedman seems to assume that if the models were accurate enough they would be used to ensure the stability of the financial system. I doubt that. Most stockholders want higher short term profits, to raise the stock price so they can sell their stock for more than they paid to buy it. But it is usually the managers, not the stockholders, who control the firms, and most of them want to get as much as possible in salary and bonuses before the firm collapses. None of the people who control the investment firms are concerned with the long term viability of the firm, let alone the productivity of the economic system.
Reply | Report Abuse | Link to thisBefore starting a new venture, a company will want to know what are the risk / rewards. A model is one way to do that analysis.
Reply | Report Abuse | Link to thisAfter the venture is started, the risk / rewards will change. Some of the assumptions will prove to be right and some will prove to be wrong. Even so, the venture will continue to a point past the original "go" criteria because the managers are committed to the venture and adverse to telling their supporters / customers that the venture is no longer favorable.
Managers may ask for many analyses hoping to find one that confirms their original decision. They are very adverse to telling their supporters / customers that they made a terrible mistake.
Some managers will delay giving people the bad news because the bad news can end their jobs. These people go into hiding and play dumb.
An independent auditor should not have a bias. He will provide an accurate picture. The independent auditor keeps management from hiding bad news.
Unfortunately, some auditors are not independent and the dynamics in the financial system discourage unbiased risk analyses.
Models are not "predictions" though, at least not Bayesian models. They produce scads of potential outcomes, providing a view of the risk landscape. And models *can* be tuned to improve accuracy; NASA revised their risk models after Challenger, bringing their estimates of loss of life or injury much closer to what we would today say "feels" right. (http://science.ksc.nasa.gov/shuttle/missions/51-l/docs/rogers-commission/Appendix-F.txt) But basing *any* decision on the outcome of a model alone, especially without understanding what went in and came out of said model, is a bad move.
Reply | Report Abuse | Link to thisAnd I've seen (but cannot cite) papers identifying the fact that some models *did* correctly estimate the risk of portfolios that were too heavily invested in [flimsy] mortgage-backed securities. Some of the companies who found those results chose to either take the risk or believed their model to be wrong. (!!) Either way, the much bigger issue that damages the market and risk management much more is the fact that the US government was put in a position where they were obligated to bail out these major banks, and failed to incur sufficient penalty to the shareholders and executives of those banks commensurate with the catastrophic loss they would have felt otherwise. Now we need to include government intervention and rescue in our models...
The disaster was not the fault of the models but of how they were used and by whom. The decisions to take the risks were not made by the owners of the companies, that is, the stockholders whose equity was at risk, but by the managers, who risked only the discontinuance of their stream of bonuses. Given a remote chance of an acceptable loss, and a high probability of continuing a favorable income stream, taking the risk was a rational choice.
Reply | Report Abuse | Link to thisWhoever said that the rich have the right to become forever richer? Economic growth means the end of planetary ressources. The best we can hope for is stability, but only if ressources are fully recycled, including CO2. For this to happen, the ton of CO2 needs to be at least $28 per ton. That would launch the Biomass Pyrolysis industry, providing hydrogen and sequestered biochar to remove carbon from the atmosphere.
Reply | Report Abuse | Link to thisThe author says " these strategies may limit profits" . In the wider scheme of things and considering their role in society, is there any good reason why these speculative activities should not have their profit expectations limited? As to modelling the markets, is that not trying to model when human behaviour switches from orderly and predictable to panic and irrational? In a panic people can overrule their judgement by wondering what others know that they don't, and feel the safest action is to follow the crowd - a stampede mentality. If the checks and balances against extreme behaviour are not in place and working, these exctremes will occur, especially if those "playing" are divorced from the consequences of their activities
Reply | Report Abuse | Link to thisThere was a stampede, but was the stampede irrational or founded on an economic collapse?
Reply | Report Abuse | Link to thisI think the latter and I also think that the collapse was predictable at a much earlier date.
David H. Freedman’s article “A Formula For Economic Calamity” Scientific American, November 2011 page 79, is an economic example of Wolfgang Pauli’s phrase "not even wrong," meaning that they were so incomplete that they could not even be used to make predictions to compare with observations to see whether they were wrong or not” (http://www.amazon.com/gp/product/0465092756).
Reply | Report Abuse | Link to thisDavid Freedman clearly does not understand how financial professionals use models. Certainly he did not bother to ask anyone who uses models to make a living in the market place.
So what is wrong with the article? Financial mathematics and engineering had very little to do with the collapse of 2008. The economics of mortgage origination quality was the driver of the catastrophe. Mortgages that were thought of being worth 100 yesterday became worth 40 today if not zero. Financial mathematics has nothing to say about absolute valuations, only relative valuations. No financial modeling complex or simple could have predicted nor prevented the financial meltdown. And good business practice would have avoided contagion.
The article is correct in describing the boom in demand for financial mathematical talent. What is article does not describe is the need for coupling the knowledge of microeconomic dynamics; the static world of financial mathematical simultaneous equations and Bayesian statistical analysis. If such combination of knowledge had been viewing market conditions long before 2008, possibly the market would never have reached such a treacherous and inevitable condition.
2008 is a harbinger of a changing of the guard. Pre financial mathematical senior executives are no longer effective in capital market financial management. The new executives will have to have the knowledge of economics, mathematics and statistics to effective guide a modern capital markets firm into the future.
Who is a capable financial market journalist whose work is informed: Nick Dunbar http://nickdunbar.net/