Leading derivative expert, author and regular contributor at the Econotwist’s, Espen Gaarder Haug, says algorithms have many weaknesses and should be monitored on a daily basis by people with extensive market knowledge and experience. The former Norwegian Wall Street trader have a thing or two to say in relation to the ongoing circus at Oslo District Court – also known as the “robot case.”
“Should the major players be allowed to carry on trading algorithm with no monitoring, without people with broad market experience monitoring them?”
Espen Gaarder Haug
The day traders in the so-called “robot case” – now appearing before the Oslo District Court accused of market manipulation – run their scheme undisturbed for nearly six months. First, when the Oslo Stock Exchange contacted the brokerage and investment firm that owned the shares that the robot was trading, the alarm was raised and the possibility to manipulate the trading program was removed.
“A rational investor would have plugged these holes quickly,” Dr. Espen Gaarder Haug comments.
“If you choose to outsource market making to a computer and you lose money.. I mean, you just have to accept it,” he adds.
Espen Gaarder Haug holds a Ph.D. from NTNU and is regarded as one of the worlds leading experts on derivatives and option pricing models.
He has worked almost 17 years as a trader on Wall Street, among others, at JP Morgan and Chase Manhattan Bank.
He has also worked as an active manager of multi-billion dollar hedge funds, like Amaranth and Paloma.
Dr. Haug underlines that he don’t know anything about robot manipulation other than what’s been written in media, but says he recognize similar issues from his experience in the international financial industry.
He believes that the trading algorithm requires good internal monitoring y individuals with broad market experience.
“I know many who run algorithm trading, and they’re always following the trades with several well-qualified employees – and often by the one who created the algorithm,” he says.
There’s Always A Risk
“Robot trading always involves a risk of weakness in the algorithm,” Haug points out.
“It is therefore important to monitor whether such risks pops up, so that these algorithms can be adjusted, or turn off. If you choose to let the robot run without supervision, you take an unnecessary, additional risk,” says Haug.
The ongoing court case in Oslo have revealed that the two charged day traders was able to exploit the same weakness over and over again.
Timber Hill (who is a part of the Interactive Broker Group) was not aware of this before the Oslo Stock Exchange contacted them on March 14th this year.
Testifying before the court on Tuesday, Thomas Borchgrevink, manager of market surveillance at the Oslo Stock Exchange, said: “I felt that they were not aware of this. They were not on the ball.”
Timber Hill closed down, temporarily, the robot in question when the Oslo Stock Exchange made them aware of the error, and has since modified the algorithms.
Chill, Timber Hill
Espen Gaarder Haug assumes that the Timber Hills algorithm was not only used to make trading in stock (so-called market making) in the Hafslund (B shares), Wilh. Wilhelmsen and Odfjell (B shares), which is relevant for this trial.
“In a way, Timber Hill, should be glad that someone intervened in these low liquidity stocks, and that no other big caps suffered even greater losses,” Haug points out.
There may be several reasons why a brokerage firm choose to let a robot trade without supervision, he says.
“They can, for example, have too much faith in their algorithm, so they do not fear failure. It may also be that they trade many shares in as many markets as possible, so that each share has little importance. And they choose to take a risk that in some cases leads to losses”.
“Around-the-clock monitoring is costly, but the algorithms have many weaknesses and should be monitored on a daily basis by people with extensive market knowledge and experience,” he notes.
Haug believes it is reason to ask whether monitoring should be mandatory:
“Should the major players be allowed to carry on trading algorithm with no monitoring, without people with broad market experience monitoring them?” Dr. Haug asks.
He believes naive algorithm trading, could not only harm itself, but at worst, damage the entire economy by reinforce large price movements.
Espen Gaarder Haug predicted in detail the financial crisis by the end of 2006, beginning of 2007, in a series of interviews and articles published by the Norwegian-The-Economist-peer – Økonomisk Rapport in 2007 and 2008.
“Although it is difficult to prove, there are good indications that the US stock market crash in 1987 was reinforced by the naive algorithm trading. The algorithms were based on a number of imaginative assumptions that broke completely with the crash. Some major players followed their naive algorithms slavic and sold more and more the more as the market dropped based on signals from the algorithms they slavishly followed,” Dr. Haug says.
Actually I have never heard of a quant strike, are they overpaid or could it simply be because they still are working on finding the optimal strike price? I wonder if it strikes you fellows, as it strikes me, that all we ever seem to talk about is integration and derivation over strike prices!
A new study of blue whales off the coast of California has given researchers the first ocean recordings of their famous hitchhiking partner -- the remora -- revealing the suckerfish's secret whale-surfing skills as well as their knack for grabbing the most flow-optimal spots while riding aboard the world's largest vertebrate.
Neuroscientists have identified a brain circuit critical for learning to make decisions that require evaluating the cost or reward of an action. They showed this circuit is negatively affected by aging and in Huntington's disease.
In the nearly two centuries since German physician Carl Wunderlich established 98.6°F as the standard 'normal' body temperature, it has been used by parents and doctors alike as the measure by which fevers -- and often the severity of illness -- have been assessed.