Posted on 09/29/2009 3:59:54 AM PDT by TigerLikesRooster
You Don't Know The Math
Michael Maiello, 09.29.09, 06:00 AM EDT
The complexities of the financial markets are beyond your broker's abilities.
Four years retired from Yale, Benoit Mandelbrot, the inventor of fractal geometry, is still trying to teach the essential lesson of his life's work--nature and markets defy easy description. Anyone who tells you otherwise is either trying to sell you something, doesn't know the math or both.
Last week, Mandelbrot, 84, spoke to a gathering of diplomats, business leaders and philanthropists at the Louise Blouin Foundation's Creative Leadership Summit. Mandelbrot has also served as mentor to economist Nassim Nicholas Taleb. Mandelbrot's work on fractals and complex systems is inextricably tied to Taleb's concept of the "Black Swan," or those once in a lifetime events that seem to happen far more often than once in a lifetime.
In 2004, Mandelbrot and Richard L. Hudson wrote The (Mis)behavior of Markets: A Fractal View Of Financial Turbulence. It was thankfully reissued with an update about the credit crisis, but it would have been better had people read and heeded the warnings in the first edition. Though we often speak about markets reverting to the mean (and they do) we frequently forget that the road to the mean is quite chaotic.
(Excerpt) Read more at forbes.com ...
Ping!
Excellent, but funny how these lessons need to be relearned every decade.
This is a bit like a self-fulfilling prophecy -- nearly tautological. The "mean" is always going to be somewhere within the data's range, and so a volitile series of data will always "revert" towards it.
Anyway, the sooner people realize (again) that markets are *people* in action, the less blind faith they'll be inclined to put in complex analysis.
That said, they can be fun tools to play with (the tools of analysis, I mean).
The same would hold true for a retail investment portfolio. In 2008, none of the traditional hedges worked. Foreign stocks fell along with U.S. stocks. Commodities spiked but then fell. The stock market's volatility spread to other markets, proving that they are interconnected rather than discreet.
I think Mandelbrot is missing a key point. The main reasons there was a systemic failure was a lack of risk management, a lack of self-restraint and a lack of transparency. Entities were leveraged to the gills, from investment banks to homeowners - which meant the impact of any downturn would be grossly magnified. Systemic risk was thought to be a thing of the past - it wasn't. And ratings agencies had issued pure bunkum - so suddenly, all around the world, banks holding what they thought were AAA mortgage-backed securities all of a sudden had no idea what they were worth once they figured that the ratings were meaningless and done only to keep the issuers happy.
Capitalism didn't fail - men did. A very, very modest amount of regulation and a modicum of self-restraint, like what happened in Canada, would have helped keep the correction fairly modest. Instead, we let Wall Street completely off the leash, not realizing that they had 'their guys' positioned to cover the risk on the taxpayer's dime.
Moves and countermoves to outdo one another will lead to such a destabilization. It is a temporal process.
When you see it from the angle of stochastic time-series, you may term it black swan or fat tail.
When it gets to find out what makes black swan or fat tail, you can argue institutions or players contribute to increasing moral hazard.
Also:
— Many of these complex models were based on relatively simple assumptions that were just wrong.
— Many of the models users relied on the model instead of tempering the results with their own judgement
— Garbage In, Garbage Out
- You just don't walk away from a mortgage - banks can go after them and take assets.
- Government regulation is centered in the risk management department - leaving executives the ability to engage in basic operations with minimal government involvement.
- Bankers willfully walked away from riskier investments.
A sound mix of self-restraint, self-responsibility and focused, practical regulation.
What are the Dems proposing here? Typical liberal micromanagement. We'll go from one extreme to another, as we typically do in this country.
Mandelbrot and the other practitioners of chaos theory would say that even simple models will lead to chaotic behavior.
The problem really isn't oversimplification, the problem is that the mechanisms are inherently non-linear and chaotic. A good introduction to the history of chaos theory is Gleick's "Chaos".
However, the mathematics that were used by the quants to predict the market are similar to those used to predict global warming. That should give you warm fuzzies.
I build financial models for a living (for a single corporation) and have been doing so for nearly 20 years. Two thing are absolute certainties:
1) The models are only as good as the assumptions that drive them.
and
2) The model is wrong. It might be close but you will ALWAYS be wrong.
Another absolute, the vast, vast majority of models will never get the inflection points (i.e. changes in the market from growth to decline) correct. Everyone always assumes that the current trend will keep going.
They are also subject to individual/political biases in the assumptions. i.e. make the models output the answer you wish to see.
Forecasting is an art, not a science.
That's a great book.
I think about chaos whenever morons start talking about "climate models" and "global warming".
Sensitive dependence upon initial conditions is apparently a fundamental law of the universe that God designed to keep us suitably reverent.
I don't think he missed the point at all. In any nonlinear system, when everything gets lashed together so nothing can fail unless everything fails, then eventually, one day, everything will fail. Anyone who has studied the theory of nonlinear systems knows that. Of course that includes some graduate level engineers and physicists. Those are not the guys who end up running the financial system.
But it wasn't about a function of hedges failing. Instead, it was a function of effective hedges not even being in place. Credit Default Swaps were not backed by adequate capital. 40-1 leverage ratios meant a wicked feedback cycle in a downturn.
The results were quite foreseeable to those not intoxicated by all the money that was being made as the bubble inflated.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.