Remember me

Register  |   Lost password?

The Trading Mesh

Knight Capitals $440M loss: algorithms gone wild?

Thu, 02 Aug 2012 11:35:30 GMT           

knight capital $440M loss due to algo glitch Shares of Knight Capital Group Inc. fell nearly 50% earlier today, Thursday August 2nd, after the market-maker announced a $440 million loss due to a software glitch. This technological issue apparently caused Knight to send out numerous incorrect securities orders, causing shares in these companies to rise and fall drastically.

Read the story from Market Watch here.

Knight has traded out of its entire erroneous trade position, which has resulted in a realized pre-tax loss of approximately $440 million, the firm said in a morning statement. Although the companys capital base has been severely impacted, the companys broker/dealer subsidiaries are in full compliance with their net capital requirements.

That doesnt seem to be sitting well with everyone.

Alexis Madrigal, senior editor at The Atlantic, writes: Add another entry to the Encyclopedia of Weird Robot Trading Events This is a field where bugs have major consequences. And thats just it these algorithms and automated trading platforms can have a few quirks, right? Enter Dodd-Frank.

Following the flash crash of May 2010, regulatory and exchange reforms were introduced to put limitations or circuit breakers on these algos. And, according to the FT, the fact that Knight Capitals glitch didnt affect the wider market may be an indication of these reforms working.

Read the story from the Financial Times here.

But whilst Knight Capitals software error didnt cause a major market crash, it did highlight yet again some of the concerns surrounding high frequency trading or HFT.

What do you think? Do we need even stricter regulatory guidelines? Or is this just one of those freak things? Let me know what you think feel free to comment below!

Well be talking about HFT, risk, algos and everything in between at HFT World New York, taking place this December.

, , , , , , , , , , , , , , , , , ,