Saturday, July 13, 2013

Economists and Econophysics

I'm not sure what to make of Didier Sornette. As I said in "UPDATED--Commodities and High Frequency Trading: Prices Being Driven By Price Moves Rather Than Fundamentals":
He can be a bit of a flake but I'm guessing he's smarter than I....
More after the jump.

One thing I do know, economists suffer from physics envy more than practitioners of other soft sciences.
Anthropologists, sociologists, psychologists et al seem to have accepted that they aren't really scientists, at least deep, deep down they do. Economists? They seem to believe that because they use the tools of science (maths) that their area of focus is akin to chemistry or physics. It's not.

At best they are modelers (no disparagement, it's something humans are pretty good at) and because the economy, like weather, is both a complex system and a chaotic system, it is probably going to take the advent of quantum computers to get to the next level of deep understanding.

For more insight into why economics is not a hard science read the work of author, raconteur, juggler and bongo player Richard Feynman:
Gates Puts Feynman Lectures Online
All that being said, I am reminded of a quote that describes similarities between both physics and economics, along with another human activity:
Physics is like sex. Sure, it may give some practical results, but that's not why we do it.
-Richard Feynman, Nobel Laureate, physics, 1965

From the Wall Street Journal's FX Horizons at MoneyBeat:
Let’s face it, economists make lousy economists.

After decades in which consensus estimates routinely failed to foresee recessions and underestimated recoveries, the blindness of most economists to the looming financial crisis in 2006 and 2007 came as a big blow to the profession’s reputation. It’s perhaps no surprise, then, that a growing number of investors and policy advocates are turning to a renegade band of physicists for an alternative way to understand how economies and markets work.

The “econophysicists,” as some call them, have been around for a couple of decades but with the aid of newly powerful computers and the rise of algorithmic investing strategies, their insights are gaining more weight in the post-crisis world. By mapping human behavior in a social context, these scientists have found that it often mirrors the forces of change occurring in the physical world, allowing them to generate a more accurate picture of how markets function than most mainstream economists.

Mark Buchanan in his new book, Forecast: What Physics, Meteorology, and the Natural Sciences Can Teach us About Economics, describes the central plank of the theory as being a belief in the universe’s dynamic tendency toward disequilibrium and instability. It’s a direct challenge to the prevailing economic theory that markets are inherently self-correcting and always reverting to equilibrium.
Take events such as the 1987 stock market crash, the dot.com bubble, and the flash crash of 2011. The econophysicists will tell you that the market at those times displayed patterns akin to the interaction of water molecules before boiling point or like the tectonic plate activity before an earthquake. They see positive feedback loops, where new market participants are repeatedly influenced by the actions of those before them, moving prices within the entire system inexorably toward an eventual breaking point: a financial bubble leading to a bust.

Even outside of the most newsworthy bubble events, markets constantly go through this dynamic process, says Mr. Buchanan. Using a metaphor from the physical world, he describes it as “avalanche” of common behavior that “sweeps through the system and, as it goes on, fosters a large, coherent change. Then, suddenly, that idea falls out of favor, and there is correction in the other direction.”

The econophysicists have amassed a decent track record. Some like French physicist-economist Didier Sornette have accurately predicted the turning points in certain financial bubbles. Others, such as the inventors of the Santa Fe Stock Market Model, have come up with computer programs that have been shown through regression analysis to behave similarly to the way that erratic, bubble-prone markets behave in the real world. A few are even finding ways to make money from these insights....MORE
HT:  The Physics of Finance whose Mark Buchanan is quoted above.

Sornette made an accurate and timely call on the Shanghai Exchange but since then has seemed to have to do quite a bit of explaining of why things don't pan out the way he forecasts. This isn't really a criticism, what looks like complete or partial failures may just be an artifact of the fact he's best at bubbles, which we haven't had much of the last couple years.

We've visited with Sornette a few times most recently in Feb. 2013's "Innovation as a Social Bubble: The Example of the Human Genome Project".

See also:
Econophysicist Predicts Date of Chinese Stock Market Collapse--Part II
Forecasting Financial Crashes: The Ultimate Experiment Begins
And FT Alphaville's wonderfully titled "Dragon-king of the outlier events".

Also:
"Pricing the Future: Finance, Physics, and the 300-Year Journey to the Black-Scholes Equation "
"WARNING: Physics Envy May Be Hazardous To Your Wealth!"
Financial Physics: "Power laws in finance"

Or any of our hundreds of posts on models and modeling.