February 26, 2012

Do we need another stock exchange without computers?

The computers who trade on their own at the financial markets are still here. In an earlier article, we found out that the volatility in the markets are not necessarily being caused by High Frequency Trading: HFT Is NOT Responsible for Market Volatility – You Are!

But the machines can still break because of human errors. This happened in 2010. The company Infinium Capital Management is involved in HFT. One of their systems broke down and the computer entered 6767 orders to buy oil futures. An investigation began and it came to the conclusion that Infinium Capital Management had to pay a fine of $350,000. But one interesting thing the investigation found was that the algorithm used was finished the day before the company began trading with it. The algorithm had only been tested for a couple of hours in a simulated trading environment. Each trading algorithm has an emergency break that terminates the algorithm if the order size is too big. But in this case this emergency break did not work and that's why the computer entered 6767 orders.

Other reports of computers who changes the behavior of the markets are:
  • The Flash Crash was not caused by HFT, but reports says that HFT helped to increase the size of the crash
  • Other miniature crashes happens all the time - HFT-computers are interacting with each other and affect each other in an unintended way

What's next? Traders have gone from trading on the street, to trading in large open pits, to trading in front of a computer screen, to teach computers how to trade. The next step may be computers who learn how to trade on their own. 

Is this a good or a bad thing? Eric Ries - the author of The Lean Startup - suggests that we should create a new stock exchange: 
"What is needed is a new kind of stock exchange, designed to trade in the stocks of companies that are organized to sustain long-term thinking."
Eric Ries talks about that we are too short-term. We trade and we focus on the next quarterly earnings - when it sometimes may be better to invest in and focus on long-term growth.

Source: The Economist