Def--
Thanks for the breakdown, most firms won't be that blunt. It does show the overhead.
I've been dropping 4-9k/day in commish at IB, lately- and am looking to ramp up velocity and volume--so this topic is certainly of concern.
I've got 2 concerns if going higher velocity at IB--
#1) Blue order hang lag--seeing a couple++++ seconds--as an average, and as much as 15-18 seconds-- a complete mess in sub second arbitrage, as we're waiting on your server checks and acknowledgements before a order hits the markets--quite aparrently HDD based checks.
I've worked recently with one of your general software guys on this--but to no avail. I've worked heavily with your top level developers in the past...with great results--they're excellent at what they do.
#2) Minimum commish --which is quadruple the charge on the same lot charged elseware. It does discourage 100lot API pikers--but for clients who actually do volume, 100lots are part of strategies that are meaningful.
--You need a volume credit against the minimum per share charge to make you competative ....and profitable at the same time.
Until then...I'll be moving the mass of my automated volume elsewhere--but would rather not.
-NBBO
Thanks for the breakdown, most firms won't be that blunt. It does show the overhead.
I've been dropping 4-9k/day in commish at IB, lately- and am looking to ramp up velocity and volume--so this topic is certainly of concern.
I've got 2 concerns if going higher velocity at IB--
#1) Blue order hang lag--seeing a couple++++ seconds--as an average, and as much as 15-18 seconds-- a complete mess in sub second arbitrage, as we're waiting on your server checks and acknowledgements before a order hits the markets--quite aparrently HDD based checks.
I've worked recently with one of your general software guys on this--but to no avail. I've worked heavily with your top level developers in the past...with great results--they're excellent at what they do.
#2) Minimum commish --which is quadruple the charge on the same lot charged elseware. It does discourage 100lot API pikers--but for clients who actually do volume, 100lots are part of strategies that are meaningful.
--You need a volume credit against the minimum per share charge to make you competative ....and profitable at the same time.
Until then...I'll be moving the mass of my automated volume elsewhere--but would rather not.
-NBBO
Quote from def:
Without knowing the market, I'll use HK as an example.
Do you know that for HK stocks, each API is allocated 1 transaction per second. (that is NOT a typo). Each additional throttle costs 100K HKD (13K USD) up front + another 100 US per month + believe it or not, an additional cost of upgrading 2 phone lines. So, lets say a firm orders 10 transactions per second. That's and upfront cost of 130K USD + incremental costs each month.
Now if one client is constantly updating (cancel/replace is two transactions), I think you can see how there is a great impact to the firm and the client is potentially costing the broker money as they slow down other clients. Hope this puts things more into perspective.
On the HKFE, firms are allocated 5 TPS per API. You can order more (and we have) but for example some clients are running code that mimics the futures, it can create speed issues.
Hope you get the picture. Now if you require that many updates I'd suggest you consider paying for your own dedicated API and all the costs associated with it. Some firms (and perhaps even IB) may consider that. Most exchanges have similar restrictions and brokers will need to make a business decision on how to deal with the issue.