HFT Myths

[earlyexit]

Just thought I'd chime in here. HFT strategies are generally capacity constrained. So, that 500% return is on a smaller dollar amount than say a hedge fund with millions or billions under management. Some HFT strategies are extremely constrained on what they can deliver. Throwing more money at it will do absolutely no good on the bottom line. Much easier to get 500% return on $1MM than it is $100MM.

Naturally, the Hedge Funds won't get that sort of return, especially when you combine them with whatever type of strats they are running. So, imo, you need to leave them out of the mix.

Quote from trade2live:

You are saying 500% per year is average or normal ? I guess some firms make that kind of returns but I can't see how this would be the norm. After all many hedge funds run HFT strats now and they aren't anywhere near those numbers.

 

[Pippi436]

Quote from hft:

Last-look is vital to market-making in FX. It's not all free money though. Similar to how you can last-look deals, you can receive reports on how often participants last-look you and turn off your flow to them. So if you last-look too often, you lose order flow.

The last look period is in the tens of millis range. They've changed it a few times (lower) over the years. Not that the actual number really matters, anything in the millis and above changes the game completely, for us at least.

The EBS plan is crazy, and I think/hope they don't go through with it. It'd really throw HFT's in a loop. Market-makers would widen out quotes and market-takers would pay up more crossing bigger spreads. I'd give it a 50 delta of passing though, since EBS has always catered to non-HFT's. I'd also bet that the plan would be rescinded in short order after their volumes drop though. They'd also lose possibly irrecoverable market share to reuters, hotspot, and currenex in the meantime though. Will be interesting to see how it pans out.

I'd heard from a high-level about Lucid but don't know much about it. I don't blame it completely on Lucid though, Reuters needs to manage their systems better. We've taken advantage of some Reuters quirks as well. All exchanges could use some oversight to ensure that their rules are enforced properly.

Those are all interesting points that you bring up. Are you in the industry - have any thoughts on them yourself?

No, i have only ever traded my own money.

My opinion about the matter is that LL may have served a legit purpose, but today even the traditional sell-side MMs should have caught up in technology, and there is really no good reason why it is still there. Every other liquid market works perfectly without it. Obviously, a for-profit venue will cater to those who bring in the bulk of the fees, so the rules are how they are.

Altho it appears that some buyside clients are increasingly not ready to put up with it - hotspot threw out almost half of their LPs last year, apparently people complained too much about too high rejection rates.

Increasing regulatory pressure makes it seem likely that more stuff will move to MTFs/OTFs/SEFs/RIEs or what ever the current favourite acronym of the bureaucracy is (kinda hard to keep up2date with this). And its probably not for the worse. A continous auction market, central limit order book with price/time order matching, only one order type (limit ioc/gtc/gfd). Noone needs anything more than that, one ruleset for everybody. That is, in a nutshell, my opinion.

As for the general attitude towards HFT on this board, i often find it a bit funny. I do remember the fx market from 10 years ago, and i don't want it back (from a primarily price taker POV).


P.S. are you active on some of the newer venues like Fastmatch, LMax, GTX, ParFX, SolidFX and so on? There is a new one every other month - are those interesting to HFT firms?

 

[hft]

Quote from CPTrader:

Hello hft,

Most trading platforms provide a proprietary API and also a FIX connection that supposedly is FIX compliant but as we know every interpretation of FIX is widely different.

1. For a trader without access to DMA i.e. writing to an exchange direct - and writing to a trading platform eg. CQG, TT or T4 would you recommend using the platform's API or using the platform's FIX connection?

2. If writing to an exchange direct, how is an ATS structured such that it can simultaneously connect to CME, ICE, EUREX, LIFFE, HKFE, SGX seamlessly and simultaneously

1. I'd recommend FIX since it's easier and will probably save you some work connecting to multiple exchanges. Since you don't have DMA, you don't care about speed anyway. Better yet, write to your broker's API and let them handle all the custom interactions with each exchange themselves.

2. Multiple threads are one way. Simply firing up connections to each and using them accordingly is pretty straight-forward.

 

[hft]

Quote from QuantWizard:

HFT will holding period for hours or days sounds like an oxymoron to me...

+1

 

[hft]

Quote from earlyexit:

Just thought I'd chime in here. HFT strategies are generally capacity constrained. So, that 500% return is on a smaller dollar amount than say a hedge fund with millions or billions under management. Some HFT strategies are extremely constrained on what they can deliver. Throwing more money at it will do absolutely no good on the bottom line. Much easier to get 500% return on $1MM than it is $100MM.

Naturally, the Hedge Funds won't get that sort of return, especially when you combine them with whatever type of strats they are running. So, imo, you need to leave them out of the mix.

+1

And it's even harder/impossible when it's $500M or larger as many hedge funds employing different types of strategies are.

 

[hft]

Quote from 76132:

So if you wanted to work for a firm like Getco, you would have to bring in your own trading logic and algorithms? Wouldn't that be tough if you previously worked for another firm?

What would compensation be like for a firm like Getco. I'd assume it's base salary plus some sort of bonus based on the profit you make? Or is your entire salary based on the amount you/firm makes?

You just rewrite what you already know.

It varies a lot. For a place like Getco, since they provide you with so much infrastructure to plug into, their payouts are way lower. I don't know exactly what it is but I'd expect in the 10-30% of your PNL range. Base salary even varies between hires let alone employers, but you could expect in the 100K-300K range, and bonus structure varies a lot between employers too. Many do it based on a combination of personal/desk PNL as well as firm PNL, plus firm politics.

 

[hft]

Quote from Pippi436:

P.S. are you active on some of the newer venues like Fastmatch, LMax, GTX, ParFX, SolidFX and so on? There is a new one every other month - are those interesting to HFT firms?

I'm not personally since I feel I have bigger fish to fry on limited resources, but I could see other firms/guys taking a stab at it for some limited PNL if their systems easily support connectivity there.

 

[tonyzhou]

Since there are too many data, what data are your company rely on
for stocks, options and futures?

1. Direct from exchange or from third party vendor?
2. Tick data of filtered data?
3. Full feed or just what you trade?

The question is about too many data, and how your company decide which level of data is enough for hft, the real case.

For example, if strategy is about stock/options combo, since options has too many variants, I think it is impossible to process a full feed.

 

[CPTrader]

Quote from hft:

1. I'd recommend FIX since it's easier and will probably save you some work connecting to multiple exchanges. Since you don't have DMA, you don't care about speed anyway. Better yet, write to your broker's API and let them handle all the custom interactions with each exchange themselves.

2. Multiple threads are one way. Simply firing up connections to each and using them accordingly is pretty straight-forward.

Thank you!

 

[Tradeae]

Quote from Pippi436:

No, i have only ever traded my own money.

My opinion about the matter is that LL may have served a legit purpose, but today even the traditional sell-side MMs should have caught up in technology, and there is really no good reason why it is still there. Every other liquid market works perfectly without it. Obviously, a for-profit venue will cater to those who bring in the bulk of the fees, so the rules are how they are.

Altho it appears that some buyside clients are increasingly not ready to put up with it - hotspot threw out almost half of their LPs last year, apparently people complained too much about too high rejection rates.

Increasing regulatory pressure makes it seem likely that more stuff will move to MTFs/OTFs/SEFs/RIEs or what ever the current favourite acronym of the bureaucracy is (kinda hard to keep up2date with this). And its probably not for the worse. A continous auction market, central limit order book with price/time order matching, only one order type (limit ioc/gtc/gfd). Noone needs anything more than that, one ruleset for everybody. That is, in a nutshell, my opinion.

As for the general attitude towards HFT on this board, i often find it a bit funny. I do remember the fx market from 10 years ago, and i don't want it back (from a primarily price taker POV).


P.S. are you active on some of the newer venues like Fastmatch, LMax, GTX, ParFX, SolidFX and so on? There is a new one every other month - are those interesting to HFT firms?

I agree with you Pippi. There is no justification for last look provision. Yes there should be one "ruleset for everybody" such as: "A continuous auction market, central limit order book with price/time order matching, only one order type (limit ioc/gtc/gfd).

Do you have any new updates/insights on the viability of LMAX, ParFX and SolidFX? I like their structure and wish they gain good liquidity.