High Frequency Trading Risks
UncategorizedOn what seemed to be a fairly normal trading day on New York Stock Exchange in 2007, things suddenly started to get crazy. Multiple buy and sell orders started to flood in almost simultaneously, immediately followed by cancellations. In a matter of minutes, these orders had run into six figures.
These orders caused a major bottleneck for other traders and investors who were unable to get their own orders executed within acceptable timeframes.
The whole thing came to light just last month when the NYSE fined Credit Suisse for allowing a rogue algorithm to go haywire and send all these messages, unfiltered, into the market.
But what was actually behind Credit Suisse’s activities?
Like many Wall Street firms such as Goldman Sachs, JP Morgan and BNY, Credit Suisse has a proprietary trading desk that runs a number of algorithmic trading and high frequency trading strategies. Using complex computer programs, high frequency trading systems generate multiple electronic orders and send them into the market to try to capture best prices and liquidity ahead of other market participants. Over the last few years this practice has grown to such an extent that it now accounts for up to 80% of all volume on US Equity markets, according to some industry analysts.
The problem with these algorithms however is when they go wrong. Fortunately, the NYSE survived the Credit Suisse episode, which was actually a fairly small-scale error. But if it had been multiplied, it could have brought trading on the exchange to a complete halt. Hence the fine.
Ever since the advent of electronic markets, there have always been errors just waiting to happen. Many of these are down to “fat finger” mistakes, where a trader might buy the wrong stock or the wrong amount of stock. But high frequency trading takes the potential damage from such errors to a whole new scale.
This is one of the reasons why regulators are now starting to tighten up on the whole high frequency trading and algorithmic trading world.