High Frequency Trading (HFT) seems to be the new rave on hedge fund street/wall Street. Is there really an edge in this strategy or is it just lot's of the typical Wall Street hype? All comments, experiences, insights, will be much appreciated. Many Thanks.
HFT is not an inherently better way to trade the market. Indeed, given transaction costs, it is a relatively expensive way to trade. Furthermore, it emphasizes the importance of nano-second order transmission and smart order-execution routing. Subject to these constraints, HFT tries to exploit current inefficiencies in trading opportunities available to the broad market, whether based on improvements to data latency, unique algorithms, or both.
Well, it just sounds cooler. The term "daytrading" is out of fashion: people think of the stock bubble and how they lost lots of money at best, at worst they think of guys who go into offices and kill a bunch of people. "Scalping" is an even less savoury word: an even more painful death than getting shot by the berserk daytrader. "High frequency trading" also implies some kind of automation support. Saying "I do automated scalping" just doesn't have that ring.
My concern has always beenwith the issue of transaction costs. How can you make money given the high transaction costs and the small margins you attempt to capture? In addition, what is the 'edge' in this so called HFT? I can't discern it. Furthermore, I think HFT and many so-called "arb" strategies have great embedded AND hidden risks. Just my thoughts... I'm open to debate and insights from those more experienced and enlightened.
There is some stuff -- including a few interesting articles -- stashed here . The recent competition and the winning buying the book strategy was interesting...
I think the heart of the idea of high frequency trading is to look at the market as a repeative sets of daily or hourly trading instead of a continuous string of trades going back months and years. Markets will often behave the same way everyday at a specific time. Where the criticism of "high frequency trading" is that a physics Ph.D. can write a program to trade the market based on "universal truths of the physical world" and use it a brand to market under. I does not mean that it is of no use to traders.
Even if your thesis that the market is just repetitions of minute scale patterns (and this may be true), is HFT the way to exploit this? Is HFT still a viable proposition given the high risks involved and the very high transaction costs. HFT seems intriguing because of its novelty, but it reminds me of an option selling strategy - it seems very good until you gat a volatility explosion and in one INfrequent event you get decimated.
Here is another way to look at it. High frequency trading is trading based on decisions made from looking at high frequency data( that is a reapetive series of hourly or daily periods) as opposed to traditional trading based on low frequency data(a continuous series of data over week or months). High frequency data is available because of increases in computer power. This goes hand in hand with more computer power enabling low commission electronic trading. Add low volatility and markets that seem to be in congestion all the time and you can see why the new way of looking at the market is popular. A lot of traders come on here and complain that nothing runs anymore. These traders are trading with a mind set left over from the days when the dow curbed in twice a week and round trips cost $60. I guess the bottom line is that traders who have been successful lately are traders that excel in congested markets and high frequency data is a very useful tool for this style of trading.