I'd recommend getting BYXONLY and NSDQBX, in case there isn't enough liquidity pn EDGA. As far as why people use the more expensive exchanges to take liquidity, there are two types of situations: A) Market Participant A takes liquidity from an expensive ECN such as NSDQ/ARCA/EDGX because the cheap/pay you to take ECNs are not present at the best available price, or it they are present, there aren't enough shares to fill the order. If the level is about to "roll over" (change, favorably), people will often take the last available shares and pay the $.003 for them; for example, the offer is .62 but its size is depleting rapidly, having been at 10,000 and now having 500 shares left. The cheap ECNs are gone, but paying $3/1000 to take the last 500 shares at.62 isn't the worst way to enter a trade ever; it's highly probable that bids will come to .62 and the next best offer .63 will be shown on the inside, vs a .62 bid. B) Market Participant B takes liquidity from an expensive ECN even though there is cheap/free/rebate paying liquidity available at the same price for enough shares to fill all of B's order. The technical term for this is "throwing money in the garbage." Some people throw money in the garbage because the difference in ECN price to them is relatively negligible (or believed to be so). Other people are forced to route to a certain ECN, or have no ability whatsoever to choose how their order is order is routed: such people are forced to throw money in the garbage. People like trader B are great to have in the market, because they behave inefficiently. If you get filled on EDGX when there's still EDGA at the same price, you were just given an arb-like fill, or a fill where you are instantly in the money. Of course, the price can blow right through you just as it can go in your direction, but being positioned to get good fills that exploit poor order execution can't hurt profitability.
Random thought about ECNs and credits... At the risk of sounding noobish, I will ask something that's been rattling around in my mind for the last few days: Anyone know of an execution strategy where a credit earning limit order is hit in the event of a stop out on your main position? Let me clarify... let's say I'm using BATS to limit bid my way into a long SPY position of 500 shares. I'm netting ECN credits in the process and (assuming I'm using a direct access broker or prop firm where there's little to no commission charge) even if the current bid is my entry I'm net positive on the trade. I usually enter in a Sell Stop order through EDGE(A or X) to manage my risk (SPY tends to move faster than my fingers can type in a new order at times... so I like managing my risk with an order already sitting at the ECN..) But upon being triggered, it's treated as a removing liquidity market order... worse yet, if the market is moving fast and they have to route it... oh my... it's like getting slipped another penny in price. So... with my mind moving toward hedges and such as an alternative to stop loss orders, is there any way to structure another limit order (adding liquidity) that makes my net position market neutral (in the SPY example that is?) My first thought was shorting another S&P500 tracking ETF to the same exposure as my SPY position, but price would have to hit my order and if the market is moving quickly I might not get the one tick pullback I need to get filled... well, even if I got it within a penny or two, the added difference wouldn't be worth it for the credits. My mind then moved on to inverse ETFs... but they have the same problem, the price has to move into my order to fill, regardless of what direction that is, so upon quick and sharp moves in the market there's a huge chance I can't get the hedge position on quick enough. Sooooo.... maybe I've already answered my own question here, but I figured it's worth tossing out there and asking the community if you can think of a way to do this. The goal is to net credits on the execution of an entry order and subsequent stop out order using the same symbol or multiple symbols so long as the effective exposure is eliminated. Maybe I'm just a dreamer...
Eh. From my experience, at least trading products that move (spy has an average range of above 100 cents so that counts as something that "moves,") I'd have to say no, you want to get out of your loss and the price is you remove liquidity and pay for it. But that's OKAY, you can still make money on ECN credits if you enter by adding liquidity. I figure it like this: (note that the following is assuming use of expensive ECNs; if EDGA can be used before the expensive ones when you take liquidity, you saved additional money). Winning trade: enter (add liquidity), exit with target (add liquidity) -- total of two credits! Losing trade: enter (add liquidity), exit (take liquidity) - depending on your routes, the credit can almost be as much as the take cost, so you're only at a small loss, credit wise. I.E. Enter, +.0024, Exit -.0028, divided by two, equals loss of -.0002 per trade. (on the winning trade, it's +.0024, +.0024, so total of +.0024 per trade). If your stop losses are halfway well placed, price should often continue to move past them, so trying to save a fractional penny will result in you chasing the market and taking a loss much bigger than the one you would take from simply having stopped out.
Thank you for your input. yeah, right now that's how I'm operating... adding on expensive ECNs, and taking from cheap ones... Like anyone who trades as a career, I'm always looking for the 'little' things that can contribute to my income over the long haul. If anyone else has ideas about this, chime in!
You could use either the NQBX or BYX routes for that. You'd want to make sure that the two routes are configured to stay on the book only and not route out. This way in your example when the stop is triggered, it'll become marketable and you'll get paid for removing liquidity from NQBX or BYX. The only problem is if there's no liquidity there your order will post on the book and then you'd get charged for adding liquidity. You could also use a stop limit order instead of a stop loss order. This way you can control the price your order is entered at instead of making it a market order. Let's say SPY is 130.69 x 130.70. You got long 130.69 with a fat BATS rebate. You now put in a stop limit order to sell @ 130.69 that's triggered only when the bid hits 130.68. So if SPY runs up from 130.69 x 130.70, you lock in profits. If SPY falls to 130.68 x 130.69, your stop limit order is immediately triggered and your order is placed at the offer of 130.69 where you can collect a rebate and still be in the green since you captured a fat rebate on both sides.
Yeah, BOSX has a 'pay for removal' setup as well, the problem there is the book often is empty. If I'm trading something thick, like BAC, and watching order flow at a given price level, and I just want to get filled at the bid/offer, I'll use BOSX since someone who's crossing the spread to execute anyway would rather get paid to take liquidity than be hit for removing it as well as paying the spread itself. As for the stop limit orders, I played with this once, it works 90% of the time, but man will it sting when it doesn't... if SPY moves cleanly through that price point, and continues just as fast, then you're left holding a cruddy position and have to scramble to make a market order to get out. I might be of the "video game generation" but I can't type as fast as SPY can slip a penny. One of my ideas was to do two orders, one stop limit at the price I'd accept for a credit based trade, and another 'emergency' order a little further away... connecting the two with a OCO function. I tried this a few times live in the market, and it sorta worked... but once in a while your stop limit will fill then the price would move faster than your system could confirm the cancel of the 2nd order further away. So once in a while you're in 'reversed' trade unintentionally. Hehe.. all this wouldn't be so much of a problem if SPY didn't have the tendency to move 20-30 cents in a split second at times of high volatility. The alternative would be to use a slower moving index ETF, something priced lower but still tracks it % for %, but then you've basically just increased your spread cost since 1 penny becomes worth more of the total % tracking of the S&P. In any case, thanks for your input. I'm still looking for other options, but I think my best bet is to stick with normal stop loss orders and just pay the fee for the security of knowing my order will be executed quickly.