2009 has seen much coverage in the media on High Frequency Trading (HFT), but I find that many people are still not sure what it is or how it works, and whether or not it is good or bad. So let’s take a closer look.
High Frequency Trading
The general description of HFT is the buying and selling of securities at lightning fast speeds, which are facilitated by computers running on pre-programmed algorithms. Programs are out there that take inputs and parameters from programmers to look for specific patterns or situations and then immediately execute a trading strategy based on the HFT firm’s proprietary models.
An Example of an HFT Strategy
This is only one type of HFT strategy, there are other which are based on news releases, or technical trading signals.
Let’s pretend that a mutual fund manager wants to buy 100,000 shares of XYZ which is trading at $10.00/share. The manager has determined that they do not want to pay more than $10.20/share and therefore places a limit order of $10.20/share. Some exchanges provide HFT firms with the ability to see new market orders before everyone else for a fee, but only by milliseconds at a time. So our HFT firm will see that there has been a flood of demand for XYZ stock and will start trying to figure out the limit price, after having bought up some XYZ stock. So for example, it will buy 100,000 shares of XYZ at $10.00/share and then start issuing small sell orders at specific prices. For example, it would start selling small positions at $10.50 which might get rejected because this is above the limit price set by the fund manager. It might then send a small order for $10.25/share, which would also get rejected. It would then send a small order for $10.13 which would get filled. It might continue to send small orders until it figures out that the limit price is at $10.20/share and then flood the market with a large order at $10.20/share.
This is a fictitious example designed to illustrate how HFT could work, and all of this could happen in less than a second. Profits per share are measured in pennies but the high volume of trading means that HFT firms can make big bucks. According to the TABB Group, HFT firms represent only 2% of the trading firms in the US, but 70% of the trading volume. I’ve mentioned before that the algorithms behind the Renaissance Technologies Medallion Hedge Fund are rumoured to be responsible for 10% of the NASDAQ trading volume alone. The TABB Group report estimates that over $21 billion in profits were earned by HFT firms in the US in 2008.
We’ll look at it some more in future posts…