Sunday, December 5, 2010

Are Wall Street Computers Making Their Own Decisions?

Are computers controlling our economic future?  IS the entire world economy based on algorithms?
"Goldman spends tens of millions of dollars on this stuff. They have more people working in their technology area than people on the trading desk...The nature of the markets has changed dramatically."AP 2007
Black Box Trading?  Flash Trading?  Algorithmic Trading?  Most Americans and people around the world have not heard of this, much less understand it.  In fact, most executives in Wall Street do not understand it.  How has technology changed Wall Street?  How have the speed of computers changed the very structure and makeup of the world's financial markets?  They have changed them dramatically.  This high speed trading drove the volume of trading on the New York Stock Exchange by 164% since 2005, according to the New York Times.

Blackbox Trading/Algorithmic Trading/Flash Trading
Blackbox trading accounts for at least 73% of all trades on Wall Street.  It is trading done by computers in milliseconds time, with preprogrammed algorithms, which decide how mortgages, interest rates, and value it to be decided.  As most experts in this field will tell you, the floor of the Stock Exchange in New York is practically a show stage, a museum of the way things use to be done.  Most stock exchanges do not even have a floor any longer, only computer screens, being watched by "experts."  It is referred to as "blackbox" because only the programmers of the software understand what is going on.  It is called algorithmic trading because it is based on mathematical formulas coded into the software.  It is called flash trading, because the decisions are made in milliseconds.  The longer it takes to make the decision, the less profit that is garnered from it.

Who are the Quants?
Paul Wilmott via:The Guardian
Robert C. Merton
A Quant is a short term for a Quantitative Analyst. You can look here for a list of the most famous Quants.  These Quants handle every aspect of mathematics as it applies to finance.  This field started in the United States in the 1970s with the Quant pioneer, Robert C. Merton.  Merton's 1952 Ph.D. Thesis, Portfolio Selection was one of the first papers to apply mathematical concepts to finance.  In 1969, he introduced stochastic calculus into the study of finance.  There are several types of Quants now, such as, Front Office Quant, Quantitative Investment Management, Library Quant, Algorithmic Trading Quant, Risk Management Quant, Model Validation Quant and Quantitative Developer.  Out of all of them, the Algorithmic Trading Quant is the most highly paid.  These Quants, use game theory, the Kelly Criterion, market micro structure, econometrics and time series analyses in the writing of their algorithms.  This is a highly technical field, so we will not spend a lot ot time here. We have provided links to the curious.
"Following the formulas was like relying on your seatbelt to drive crazily: it's not going to save your life. People in risk management don't know a fraction of what they should; they're not sceptical, they haven't tested the data or used their imagination to find solutions." Paul Wilmott, 2009

Paul Wilmott, a Quant, considered one of the greatest geniuses of finance in Wall Street, predicted the financial meltdown of 2008. He claims to have warned many of the bankers, that they trusted these formulas too much, and that they were making too much money. Together with Emanuel Derman, he co-authored the Financial Modelers' Manifesto. It is too long to quote here, but it is a fascinating document. You can read the full text of it here.

Emanuel Derman, another Quant, who co-authored one of the first interest-rate finance models, the Black-Derman-Toy model, as well as the Derman-Kani local volatility or implied tree model, also agrees with Wilmott.  In the beginning, they were not viewed with respect in the finance world.  They were viewed as having no understanding of the business of Wall Street, now according to Wilmott, "they are the business."

Mike Osinski, another Quant, but now retired from Wall Street, admitted in 2009, that he built the model that broke the banks.  He wrote the software that turned the mortgages of the banks into bonds.  He had an interesting quote in a New York Magazine interview of March 29, 2009, entitled, My Manhattan Project.  He stated that,
The packaging of heterogeneous home mortgages into uniform securities that can be accurately priced and exchanged has been singled out by many critics as one of the root causes of the mess we're in....The software proved to be more sophisticated than the people who used it, and that has caused the whole world a lot of problems.
This interview becomes more revealing in that even the man who designed the software did not fully realize the effect it could have.  The computers, acting out of the pure logic in their software instructions, carried out the destruction:
I never would have thought, in my most extreme paranoid fantasies, that my software, and the others like it, would have enabled Wall Street to decimate the investments of everyone in my family.  Not even the most jaded observer saw that coming.  I can't deny that it allowed a privileged few to exploit the unsuspecting many.  But catastrophe, depression, busted banks, forced auctions of entire tracts of houses?  The fact that my software, over which I would labor for over a decade, facilitated these events is numbing.  Is capitalism inherently corrupt?  I don't think the free flow of goods in and out of itself is the culprit.  No, it's the complexity masked by thousands of unseen whirring widgets that beguiles people into a sense of power, a feeling of dominion over the future.




What is a Flash Crash?
A Flash Crash is when the stock market can lose many points of value in seconds, due to the activity of computers making decisions in milliseconds.  
Flash Crash May 2010
This was the biggest one-day point decline (998.5 points) in Dow Jones Industrial Average history.  In 15 minutes, $1 trillion dollars in market value disappeared.  The value of Procter & Gamble stock dropped 37% in those 15 minutes and then bounded back to its original levels within minutes of that time.


Things Are Getting Out of Control
Click to read it
The speed at which these transactions move is essential to their effectiveness.  Milliseconds can make all the difference in the world.  Computer systems are geared for speed, and proximity to shave off more minute amounts of time.  According to Derman, these models do not "predict" the future.  They need some basic information that needs to be fed into them.  Based on this view of the future, they predict the most profitable approach.  The Quants are the only ones who understand the models.  The managers trust them, and even the Quants cannot always spot mistakes in the model.

But even worse than this, the Quants cannot predict the outcomes of their programs, or of the actions of the computer systems they designed, and they are the intelligentsia of the group.  This tendency to rely on computer systems, will only increase.  There will not be a retreat into more human control.  On the contrary, more software will be devised to control and resolve the weaknesses of the previous software.  This will only increase our reliance on computers for the very health of our economies.  These automated systems will act without hesitation (unless programmed).
Eventually it is not hard to see where this will all lead.  Eventually, software will be too complicated even for the Quants.  It will require the focus and precision of another machine to program a machine.  This all bring us back to the transhumanist issue.  We must, to compete in this ever more complicated world, surpass our human intelligence and form a more intimate connection with our machines.  Herein lies our only chance to control them.  This flash crash of 2009, is only a shadow of what lies ahead if we ignore the lesson.

Here is a documentary named Quants: The Alchemists of Wall Street:









2 comments:

Stefan said...

are we making a global brain?
Metaman is on my Christmas reading list:
http://en.wikipedia.org/wiki/Metaman

PlusUltraTech said...

I think it is inevitable barring some global disaster that sets us back technologically.