http://blog.themistrading.com/?p=2256 The paper itself will be over your heads, but if you read carefully, you'll see that my claims of hft being scum are in fact, proved empirically . Winston is toast.
HFT are doing what daytraders, floor traders have been doing for decades, they are just 'efficient' in duplicating what floor traders been doing for decades. there is no way I can manually trade like those HFT machines at that speed. these HFT machines do like thousands of round trips per month the trading strategy by the HFT machines only machines can do. A human trader trying to trade like the HFT machine would burnout or be too exhausted.
Over whose head? That's what I like about ET. Everyobdy assumes they know more then anyone else possibly could. Or in other words, each poster thinks he is the only elite trader on ET. Hilarious. And the paper? It is NOT research. It doesn't prove anything; they had an opinion and looked for a way to prove it. The moment a person performs an experiment, or analyzes a specific event with any preconceived notion of any type, it automatically influences the outcome. How much more so if the researchers are trying to prove something (as in this example). Didn't you notice how many times they used the words "hypothesis", "assuming", "interpret", etc.
It's actually a theoretical paper that proves if HFT traders do "X" and some other assumptions are met then "Y" will be the result. There's nothing wrong with this kind of theoretical analysis but it doesn't prove this really happens outside their model. They don't show that their assumptions are met in the real world and there is no data examined in the paper! So I don't think OP can claim anything is empirically proven.
For the pro side... http://www.futuresindustry.org/ptg/downloads/HFT_Trading.pdf http://venus.unive.it/luca.tiozzo/Summer_Paper.pdf http://finance.wharton.upenn.edu/department/Seminar/micro/Litzenberger_transient_vol5_2010.pdf http://www2.econ.osaka-u.ac.jp/~tsutsui/english/DP Market Efficiency.pdf http://www.cis.upenn.edu/~mkearns/papers/KearnsKuleszaNevmyvaka.pdf http://www.apeaweb.org/confer/bei08/papers/iwatsubo_kitamura.pdf
good,informative read and sounds really nice..almost make me believe that HFT is good,providing liquidity,reducing volatility,increasing efficiency etc..it's all good, until YOU try to buy decent size at ask or sell at bid. when you do this manually tens,sometimes hundreds times a day-you will see the difference or impact of HFT.. subpenny trading is also good and fair for retail guy,right? once again-you have to trade,on order to see the impact. on other hand-it's is competitive field,where one will always be looking on any advantage over the other players. but pulling out orders,before you can hit them,posting the quotes without intention to trade-it's abuse,plain and simple. and that's exactly what HFT do.it's a part of their business model.
HFT is nothing more than automated market making except they have no obligation to provide any liquidity. So they can shut off their machines and go home whenever they feel like. Having said that I suppose some would argue the old nyse specialist and nasdaq mm's didn't do a good job during market "stress" either so, i don't know. But there is anything inherently wrong with hft. I actually hate that its become some stupid buzz word in the industry. And for those who think they ripping off mom and pop when they buy 500 shares of IBM, thats just dumb. The old specialists were masters at that, and if you're that concerned about it, use limit orders.