I could be wrong about this, but switched from IB's SMART router which mainly routes to the NYSE to routing to ISLAND and ARCA on Genesis. At first I was so excited because I though of all the ECN rebates I would collect that basically my comissions would be free. If you are trading a NYSE stock and route to ARCA or ISLAND, the fact is you really just do not get hit unless the price trades through. I get filled much better when I'm routing to NYSE. I'm starting to think it's really not worth shooting for the ECNs on NYSE stocks... what do you guys think?
Why don't you improve your trading... And trade something with a decent edge... Instead of trying to eke out micro-pennies from rebates. You did not discover the rebate approach, dude... And your competition is people you DEFINITELY cannot compete with. I don't even know what the rebates are... Because my NYSE margins and profits are large enough... That the rebate issue not worth addressing for me. Like is poker... The #1 goal is to find a "soft game"... Avoid the strongest players unless you are one of them.
IMHO, rebates are vital for DMA brokers (especially with SmartRouter) to keep their execution/routing cost down. If you want to fill a buy order with the size more than posted on NBBO, and a couple ECN's match the NBBO price, so you'd just take NBBO size and route the rest to ECN with the lowest take fee instead of sweeping the book. BTW, it's not that hard to implement, your ATS just need to watch multiple feeds. But I was never focusing on rebates routing all trades directly to NYSE...
I sent all of my non-marketable NYSE limit orders to ARCA. While I'm sure that ARCA is less likely to get a fill when at the inside along with NYSE, I like the rebate and know that in fast moving stocks, the NYSE will rarely even be at the inside market. That said, I have never found a good way to test the true 'cost' on different routes on non-marketable orders. Specifically, what is the slightly increased fill rate 'worth' on NYSE, versus the rebate on ARCA. I'm guessing that the speed and rebate on ARCA make it a better route for non-marketable NYSE limit orders, but I could be wrong.
We trade pretty liquid stocks, and we like to "take" on NYSE and "park" on ARCA when closing a trade. We also park on ARCA when pair trading because we have predetermined pricing for the most part. Some of our guys get $2,000+ per month in ARCA rebates. But (always a but, LOL)....if you find that you're focusing on the sub pennies vs. the profitable trades, then just stick with NYSE. I trade GE a lot, and generally use ARCA for exits, FWIW. Don
Don, as a very respected trader I definitely take your advice... however I could be wrong about this.. but isn't the "cost" of entering a function of where in line you get filled at a certain price point. Like if there's 100 orders at a certain price point you'd increase your expected value if you are #1 / 100 vs #99 / 100 to get filled... I have really never have checked this but I have to imagine that @ certain price points NYSE bids / offers are taken out first and then ARCA last so it really becomes a question of if the rebate is worth it? ~maybe it's a wash?
I, like Don, park on ARCA (sometimes INET & BATS as well). If necessary I will take on NYSE also if I find that I'm doing too much chasing on a name. I don't look at rebates as my edge, but as a ROI enhancer. I do enough volume daily that my rebates are significant & easily cover my trading fees. I measure my fills against the spread at the time I place my initial order and use this as my guide as to whether trying to provide liquidity is penalizing me an amount greater than the rebates I'm collecting.
this is the whole point... how do you know that it's an ROI enhancer... if you believe that ARCA, BATS and INET orders are just as likely to be filled as NYSE order then I agree with your statement, however I doubt that's the case... so you have to try to estimate the cost of getting filled later vs the ECN rebates. You have to have a database to really get the answer...
Not sure what would be in that database that would give me the answer. Even if that db had tick data AND trades/fills (which I have) it would be tough to estimate where you would have been filled. Probably the best way to answer would be to running a significant number of shares in a live side by side comparison. What I can tell from my trading (~ 500k shares/day) is that I do beat the NBBO spread on average BEFORE I give myself credit for rebates.