http://blogs.wsj.com/deals/2012/11/...r-shows-scars-of-trading-glitch/?mod=yahoo_hs Probably a bad thing for the industry if GETCO wins the bid. One less major market-maker competing for our orders.
It's also bad because virtually every retail order through the big online brokers such as Ameritrade and E-Trade goes directly to a market maker, paying a small fee to the broker for the privilege, rather than to a competitive exchange. GETCO and Knight are two of the big market makers in that space now, so this will further reduce "competition" in this space (although "competition" is tough to pin down in the convoluted world of payment-for-order-flow in off-exchange transactions). The ultimate loser is the small investor, who has no that he or she is effectively paying a lot more than the stated $7.99 comission to the broker to get into those 1000 shares of GE.
Payment for order flow doesn't hurt investors. If I want to buy GE and the market is 21.05 x 21.06, I'll get filled at 25.06. It only hurts the traders offering .06, which is mostly professional traders.
Yeah, but what if it's 21.00 x 21.10, and the order gets internalized at 21.10 instead of probing for midpoints or dark liquidity between the spread? Then the investor gets totally screwed, unless they're using smart routing algorithms that will check all the pools. That's broker dependent. Even for 21.05 vs 21.06, you can evade the midpoint if you go to the wrong route and get short changed a half tick.
How do you figure you get short changed? If you see midpoints going off on one ECN's T&S for half a penny better than you were willing to execute on, and you hit a different ECN up, you only short changed yourself. Or, if you hit that same ECN that was printing midpoints and got filled at the round penny, the book dried up before you got some... If you just hit an ECN without knowing anything about midpoints and expected to get some god-sent crazy good fill at a price better than you listed on your marketable limit order, well, you have your head up your ass. And the 21.00 x 21.10 example doesn't apply to midpoints.. midpoint orders don't stretch out over whole penny units like that.. and why you'd cross a 10 penny wide spread without probing for liquidity part way first on various trading venues is beyond me. You don't even have to flash your whole order, just test the waters to see where people/algos are interested... heck, probe dark liquidity too... unless the stock is flying off the handle and you need to trade it that instant (like flattening a bad position,) there's no reason not to try such things first. gah..
A) "If you see midpoints" -- on what? The lit exchanges? Ok. But what about one of several dark pools? Unless you know something I don't, I'm not sure how an investor can tell which dark pool was responsible for a midpoint fill on the consolidated tape. The other issue is whether your broker even offers connectivity to the dark pool that had the right midpoint. If they (the broker) don't have connectivity to the right pools in this fragmented market and decide to send it to an internalizer, it is possible for people to miss out on a better fill. B) Mid-points do apply between 10-cent spreads. As an example, DirectEdge even has order types to put limits on midpoint pegged orders intended to execute between the NBBO. My guess is that they added these features because people were upset with poor midpoint fills when the NBBO was fluttering about. The investor who trades through E-Trade is probably not going to probe multiple venues on his web interface with limit orders. At best, he ends up using a smart routing order option, at which point he's dependent on the underlying technology of the broker or a routing strategy provided by the exchange. (Moreover, none of the exchanges reveal openly who their dark liquidity partners are, so it's anyone's guess whether you probed wide enough for your dark fill.) Some smart routing options don't probe all the various exchanges, and you're lucky if your broker even discloses which dark venues your order will access. So it's entirely possible you get short-changed for a number of reasons. Internalizers definitely benefit when someone isn't in front of them at the midpoint in their fragmented corner of the market.
An investor on ETrade isn't going to do anything like this, heck, I'd be surprised if their "active" traders even know midpoints exist. In most cases, it's better for retailers to hit up their broker's smart router anyway, some even help them reduce their fees overall and they can't get filled worse than the NBBO. This was really more about handling orders at a DMA prop or firm. As for dark pool midpoints, well, it's dark.. not seeing their book and not having a good idea of the T&S is part of the point. But to be honest, in my time I haven't had much success hitting dark pool midpoints or sitting on a midpoint order in a dark's book and getting filled... there's too many dark pools to list them all, but this includes MS and a few other larger ones. I have had plenty of success, however, with ARCA/NASD/EDGE/BOSX/etc... lite exchange midpoints. The only headache is the different order handling between them all (Like when they are pegged, etc... NASDAQ has burned me a few times filling me a penny worse when the level breaks without my midpoint filling. ARCA is a lot more forgiving, not moving the order..)
The retail investor is the cash cow. If GETCO and Virtu could strip Knight down and just take what they wanted, they'd drop everything and just take over probably some NYSE privileges and then take all the special order flow relationships. They'd effectively build a relatively consistent source of revenue, despite falling retail volumes. Btw, do you trade out of a brokerage firm that has access to MS's dark pool? My broker offers a few separate routes, but MS isn't one of them. Send me a private message; maybe we could discuss routing options.