The Knight affair evokes the stranglehold of machines on the financial markets:
August 11, 2012
The programming error which caused the American broker Knight Capital to lose 440 million euros again raises the question of the omnipresence on the machine markets accused by their detractors of increasing risks and hurting investors.
On August 1, in order to be able to interact on a new trading platform launched by the stock exchange operator NYSE Euronext, the broker Knight Capital deployed new software which began to pass erratic orders in bursts.
It took about 45 minutes to stop it, time for Knight to lose $ 440 million.
The president of the American market policeman (SEC), Mary Schapiro, denounced the same day an incident "unacceptable" and potential factor "of concern for the investors" classic.
"Will we have to wait for a disaster from which we will no longer be able to recover to ask the right questions?" Asks an analyst, on condition of anonymity.
The ubiquity of computers in the financial markets is not new. Most stock exchanges have been fully electronic for more than 20 years and the few surviving floors (mainly Wall Street) deal with a very marginal volume.
But since the entry into force in 2007 of American and European reforms intended to promote competition, the fragmentation of stock markets has corresponded to the emergence of high frequency trading (HFT).
Every day, millions of orders are driven by predefined algorithms, which seek to take advantage of minute price differences between trading platforms or to anticipate market movements as soon as possible.
Actors like Knight Capital, who practice high frequency trading, argue that it allows investors to get the best price through these algorithms and provides them with better liquidity, i.e. the ability to buy or sell at any time.
The memory of the "Flash Crash"
For Costis Maglaras, director of research at the Graduate School of Business at Columbia University, "markets are much more efficient today than 20 years ago".
At the time, he argues, "you had to go through many intermediaries and transaction costs were high."
As for the risks, Mr. Maglaras believes that, despite Knight's example, the probability of error is less than before the advent of electronics.
He also underlines that there are now, within the firms using these algorithms and at the level of exchange platforms, circuit breakers capable of preventing a large-scale slippage. “Sometimes it takes a few thousandths of a second, sometimes a few minutes,” as with Knight.
However, several recent academic publications have come to relativize the advantages that high frequency trading provides to investors.
According to Pierre-Cyrille Hautcoeur, professor at the Paris School of Economics, improving prices only benefits a handful of stakeholders, but not small investors.
For Hautcoeur, who shares Mary Schapiro's concern, the challenge is to maintain investor confidence in the market, whatever their size.
However, the Knight episode, like that of the "Flash Crash" which saw the Dow Jones index collapse by nearly 1.000 points in a few minutes on May 6, 2010, following a poorly formulated order , ask question. "In some ways, you get the impression that reliability is lower," he observes.
Ms. Schapiro indicated that she had asked her colleagues to quickly make concrete proposals to her on the subject.
http://www.boursorama.com/actualites/l- ... 9e9dab8801
