You can make plenty of money in HFT without FPGA's. Profile your hotpath and only pursue it when it makes sense given the full context of your other latencies.
You can definitely do it between US exchanges, why couldn't you? It's more competitive, but the rules of arbitrage dictate that someone has to do it. I don't really follow banks for price changes. Except in very specific use cases (think highly illiquid markets like Power), bank feeds are nowhere near as informative as exchange pricefeeds.
HFT, thank you for advice. I guess I'll try start without FPGA and then switch to FPGA solution because it's guaranty of the best data feed handle in my opinion.
Here's my question: At what point is the trading technique classified as algorithmic vs. high frequency ? In other words, what is the average time-in-trade where it is no longer characterized as high frequency. 100 milliseconds ? 1 second ? 5 seconds ? Also, I've read the articles where HFT profitability has been on the decline for many firms. I wonder if they are looking to lengthen their time horizon, use different algos and move away from the millisecond trading ?
Very interesting. I wonder if they use a classic measure like the VIX to detect a shut off condition or do they roll their own volatility measure ?
I focused on less 1 millisecond. Long time horizon system is not guaranty of stable profit. I have system than worked 3 years for FOREX and apr-may dropped down [ http://forexgrowthbot.com/ ]. My IMHO only.
i would imagine most hft's just observe their p&l's when their daily p&l's start losing consistently. ie from just a few loss days within 100 days to big losses every day within a week. i remember citadel and rennaisance went thru that in 2008. i am sure the smart co's who do a good job risk managing will use VIX or some other signal to lessen exposure, but the truth is the vast majority still use archaic risk management, ie 2% down, stop trading. for example, the 2010 flash crash was when hft's started losing millions within a few minutes and then all shut off, no liquidity = no market = crash.
HFT's are just robot versions of the old specialist/human market makers. specialists on the NYSE were high frequency too, they were the market makers who bid/offered or matched thousands to millions of trades a day. in turn, they get to essentially "count cards" and know the market's real time open interest. this is how today's big HFT's still make. in other words, if you're HFT and small like a minnow, the whales will eat you up. so it's difficult to lump all HFT's into profitable or not profitable. they have their own food chain to deal with.
That is admittedly not far from the truth. Archaic as they might be, they are still effective as long as they are implemented properly. For example, a 2% drawdown limit would have kept Knight in business had it actually been in place. The good firms, which I like to think include myself, might have simplistic risk measures, but highly robust levels of redundancy to ensure that they are implemented consistently and effectively.