Really depends on the situation. Most traders/quants these days do most of their own coding, with a software developer maybe to make performance enhancements to the code. Then within a team or firm there's a few trade operators monitoring the algos making sure nothing goes wrong. The idea guys get the most money, particularly if they do all the coding too, but high-quality performance software dev's can get paid good cash too. Trade operators don't make too much (relative to the quants/dev's), but have an avenue into moving higher up on the food chain.
interesting... is that a function of the current environment you think (low vol)? i would have thought a lot of these groups would be interested in the etf/futures/index arb side of things as well. am i right to assume then that the majority of the hft volume you're talking about is primarily futures-only rv/delta1 market making?
I think I'm overextending myself trying to estimate what those bare-bones firms really specialize in. But I think a key barrier for them is the lack of top-notch inter-colocation connectivity that is required to successfully do HFT in equities. Microwave lines between all of the protected equities exchanges is expensive, and their operations aren't large enough to justify the cost. On the other hand, simply colocating at the major exchanges and giving you good intra-colocation connectivity is relatively cheap and straight-forward. So yeah, futures only trading, both market-making and liquidity-removing, are common strategies for them. Options trading where you don't need good inter-colocation connectivity is something they can offer relatively easily as well. That's not to say they don't trade equities, it's just that it's on a slightly longer timeframe than ultra-HFT trading (like holding periods of seconds and minutes instead of milliseconds). At least that's my educated guess, I really don't know for sure.
Maybe a silly question, but why would microwave links be more expensive than say, a fiber-optic cable, which has to be dug into the ground? Somehow it feels like this infrastructure thingy is sort of a "re-discovered novelty" and hence overpriced...?
It is definitely faster, I read about it in few places. However, the draw back is bad weather from what I saw.
Can't imagine that's a very critical factor - you obviously need a line of sight, and humidity etc. can influence attenuation, but we're talking about microwaves here and not visible light...
hi hft, does HFT affect the market so much that it is not wise to trade except we can also do HFT? HFT affects the market so much that normal speed intraday or swing trades would have very hard to time to detect any profitable pattern or has any way to have good return?
Thanks for posting here, hft -- this gets my vote for best ET thread of all-time. If you don't mind my asking, how does a firm with tremendous volume handle the increasingly nightmarish US tax code? Does it just get passed off to Green & Co or some similar firm, or can outside firms even handle that? I can't imagine that the (often horribly inaccurate) 1099's being passed around are helping matters -- or are most HFT's B/D's that prepare their 1099's themselves?
No, I think there is still value in intraday/swing trading, just not tick-by-tick trading/scalping that manual traders used to be able to do.
I'm not sure how the exact logistics work, but I've never heard of large volumes themselves being an issue with Accounting. PNL is very straight-forward for liquidly traded products - just stick it in a database and add up buys and sells more or less. There's some minor issues here, but they have to do with things other than high volumes - illiquid instrument pricing, currency/timezone differences between us and the clearing firm, etc. For tax treatments, the larger problems become those similar to other businesses - depreciation, allocation, international tax differences, etc.