I agree. I think IB's smart router does not capture much of dark pools liquidity. Which CSFB's algos do you prefer? Have you tried Jeff or Fox River? I have had some success with Fox River's Dark Attack algo.
I only use Guerrilla (Patient) for now. I've used a few others (Sniper, Pathfinder) before. Overall it's not bad but since I use the API, the commissions are significantly higher. I prefer nowadays to use IEX where I can. I understand that since you're so much of the ADV, it's different as you really need to take advantage of dark pools. I'm mostly able to "blend in" and I'm only looking to minimize slippage. I've used Jefferies before briefly, seemed to be roughly on par with CSFB, very slightly worse fills but barely noticeable. Have not tried Fox River yet.
I an assuming that OP is seeing it on Time and Sales. I further assume that TnS is a "special" feed that goes from IB server to TWS. It would be crazy for them show made up trades on TnS, but it's IB afterall, so please correct me if you don't agree. Which exchange reports it and with what sale conditions? Most likely some internalized or dark pool trades. Normal/expected scenerio is you send a 1000 order and get 10x100 executions.
Ran a little (fairly limited) side by side for Guerrilla and Adaptive, similar profiles. As expected the former was better by 3.4 basis points. The only way IB is competitive here is through (over)charging for competitor's orders when sent through API.
Jeeeeez-oh-Pete. ((This is where libel gets the "knowingly-misrepresented" defense -- you can't get sued for ignorant opinion.))
Once again you offer colorful commentary that is irrelevant to trading. What I said was based on actual trades, I'm not giving mine or asking for your opinion which have no value here. To make it wooden and red (blatantly clear) - my statement is based on the fact that API direct routed orders to CSFB are about 80% more expensive, yet manually entered orders cost the same. And, as it appears that's the only way IBALGO can compete. I will keep the comparison running and will advise if the data proves me wrong. I'd also like others who experiment with order routes to talk about their results.
YOU are responsible for the code submitted to IB via the API, not IB. IB has institutions, family funds, and retail traders all trading the API, for years. What do you suppose would happen to that customer base, were a costing fault as you describe, to be found out? (Let's say, via the *annual* trade audits that are performed to demonstrate IB's Smart Routing price improvement performance over that year. Annual. Trade. Audits.) You are saying that you are smarter than alllllllllll these other traders, than federal regulators, than professional DMA auditors. For years. I am skeptical. If you have such an issue, it will be found in your code. Your. Code.
Reading comprehension. Never did I say it's a "costing fault", it's their pricing policy. What do annual audits have to do with anything? IB has no obligation to beat alternative execution venues and they compare improvement to laymen sending visible orders to main venues such as ISLAND, NYSE. What has my code to do with anything? I'm talking about orders submitted measured when they are already executing on IB.