There is an obvious problem with HFT: it gives an advantage to lower latency analysis and communication, leading to huge amounts of wasted effort getting faster pipes and software, co-locating automated trading, etc. None of this is delivering real value to the economy, as far as I can see. If you have an argument for why trades that happen at millisecond time scales help liquidity and price discovery over what you'd get with a system that handled trades on the scale of minutes, I'd like to hear it. I have a hard time imagining that the benefits (to society at large) could outweigh the costs of the waste when I see what's being done.
That said, I agree that the high frequency of trades themselves are probably not the problem. The problem is rewarding a latency/processing advantage.
That said, I agree that the high frequency of trades themselves are probably not the problem. The problem is rewarding a latency/processing advantage.