Search results

Ignore trading advise. Do the back testing.

I get the same thing, on fill events. Two of them. Here's the problem: If I want to do something as-soon-as a fill arrives, which involves resetting an order status, I get screwed by the second fill event. So .. I now have an intermediate state "KIND_OF_FILLED", which indicates that I...

Certain algos will post in lit and dark venues at the same time. If they get done in the dark [where you are not], they will cancel on lit, but on some exchanges, still report on lit as a trade. Blame dark pools -- the land of anti-price discovery.

Nice work. Have you considered some back testing? p.s. Interested to know which charting component that is.

hehe .. it is *because* the spreads are wide that the MM strategies work .. the high frequency bit is about them coming to you a lot -- not the other way around.

Lets say I flip a 1c, 5c & 50c coin each 100 times. Results come out as 1c : 55% heads. 5c: 49% heads. 50c: 52% heads. Is my 1c strategy worth more? ... Personal view on "signals" is if you want to use them, them paper trade them for at least 3 months & do it seriously!

IMHO a lot posted above is over complicated & over engineered. The IB API is OK as a solution, and yes, of course is async. How can communicating with an exchange not be? - I send a new order -- and get an async ack. - I send an amend -- and get an async ack. - I get a *partial*...

Time at exchange ack, minus time sent. For IB this is (time when orderStatus = SUBMITTED) - (time at placeOrder). From Australia, I see 300ms to 900ms against the IB test market for ES. Pretty hard to arb anything at those latencies..