if your order gets hit with a partial fill, like order entered is Buy 1000 ABC @ 33.50 and your hit with 500.00 Is your order still in priority to receive the next 500? (from the same exchange) can the tape tick off more shares at that same price with out them going to you? And in that case would it be wiser to just cancel and re-summit the order? thanks in advance R.D.
I am not sure I understand you: Why would you want to cancel and resubmit? Does not make any sense if you want that order to be filled quickly and at that price. If you cancel your order, you logically lose the priority you had when it was working on the exch. and once you send a new one you can expect to be in a cue once again.
Yes, you remain top priority wherever your order is posted. Yes, other trades may go by at or even below your price without you getting filled because they are executing at other exchanges/ECNS, could be late prints, etc. Depending on the instrument, they're not supposed to cross by more than a penny or three, but it happens (can someone provide the current state of the trade-through rule?). If you see a bunch of prints going off at your bid price and can identify where they are executing, and you see the bid drop on that venue, it might be worthwhile to re-send your order there, but probably not. Generally, everyone can see every quote (especially on NASDAQ), so they'll come to you if the guy they're hitting went away and you're the only one left.
NYSE does trhis all the time due to autoex trades going in front of you... I cancel and resubmit all the time to autoex shares as they pop up... (rather than someone else getting them and then having the stock move through the price level....) mnx
If your order has not been wisely routed, or it has not been wisely re-routed in response to events occuring after it was initially routed, then you can expect to sell at substantially lower prices or to buy at substantially higher prices, and you can expect to suffer drastically worse delays, and drastically increased risk that the market will run away from your order without it ever having been executed. This includes the unexecuted balance remaining after you receive a partial fill. It is also not true that everybody can see everybody else's displayed orders. Interactive Brokers (IB) customers, for example, cannot see quotes on regional exchanges; and even if they pay for some datafeed, like E-signal, which does include regional quotes, IB customers cannot route to regional quotes (I am consdering whether to start a thread discussing regional quotes). Most retail customers see even fewer market center quotes than those customers at IB. Most retail orders are routed very poorly and stupidly (IB probably does the best job, despite omitting regionals and other flaws in its routing; and I am not aware of any retail order router as good as IB's). Many, if not most retail orders, are routed on the basis of kickbacks (called "payment for order flow", a form of fraud which the SEC legalized for the benefit of the securities industry), rather than best execution price. If you rely on orders to come to you, instead of routing your order intelligently, then you will be throwing away a great deal of money, and you will be delaying your orders unreasonably and you will suffer far more missed fills. The following advice is highly misinformed: You are almost surely better off re-routing your order in such a situation. You need to route to where the action occurs. I base this on my very careful, blow-by-blow analysis of specific orders, how they executed, and how they could have executed if they had been better routed. Also, block trades, on the same exchange, can trade thru your order, under NYSE/AMEX exceptions to the trade-thru rule. If the amount remaining on your order is for less than 200 shares, then orders on other market centers can trade through your order (another exception). If your order is for SPY or DIA, then it can always be traded thru by up to 3 cents on any other market center, because of a special exception for those two securities. No exception to the trade-thru rule allows for trade-thrus by up to a penny; the only exceptions are 3 cents for SPY and DIA, or no exception at all. Note that for NASDAQ-listed securities, the trade-thru rule simply does not apply at all. QQQQ lost the protection of the trade-thru rule when it moved its listing from AMEX to NASDAQ. NASDAQ has no trade-thru rule. BUT NYSE-listed and AMEX-listed securities, which are traded on NASDAQ (SuperIntermarket), are subject to the trade-thru rule, including any of their executions occurring on NASDAQ. If your order has been routed to a NYSE/AMEX specialist, instead of a NYSE/AMEX auto-ex system, then it will receive manual execution, rather than immediate automatic NYSE/AMEX execution. Manual execution will, on average, be far slower and more costly and far less reliable than exchange auto-ex. If you cancel and re-submit, then under certain conditions (study the NYSE and AMEX rules), you will be able to receive immediate automatic execution, at a better price than you would receive if you simply waited for pending manual execution. Sometimes, cancel and resubmission can get you an execution which you would otherwise not get at all. It might be an even better strategy to avoid execution by specialists entirely, unless you have a strategy that allows you to profit from your insight into how specialist's behave. And yes, routing of equity orders is a pain in the butt. Correct routing is absolutely critical to minimizing trading costs, delays, and missed fills. Stupid order routing will greatly increase your trading costs and delays. Missed fills have an even worse consequence. The missed fills resulting from stupid routing will cause you to tend to receive fills at those times when your trading strategy is wrong (market moving \against you), and to miss fills when your trading strategy is right (market moving in favor of your strategy); so that you will suffer reduced profits and increased losses even on a gross basis, that is, while ignoring commmissions, spreads, slippage, and other trading costs. Stupidly missed fills will damage your underlying trading strategy, by reducing your gross profits, even before we get to the point of considering your net gain after trading costs. Also note that cross trades, on NYSE and AMEX, receive higher priority than orders submitted prior to the orders resulting in the cross trade. Also note that on AMEX and NYSE, time priority is lost, and your order is put on parity with more recently submitted orders, and often (see NYSE/AMEX rules) your order is put BEHIND more recently submitted larger size orders, whenever a trade simultaneously exhausts all shares on both the bid and the ask. This is very different from ECNs, where strict price-time priority is persistent, rather than fleeting, and therefore plays a far more important role, and larger size orders cannot jump in front of newer, smaller size orders, as they often can on NYSE or AMEX (see the rules for details). Many traders do not realize the very limited role played by time priority on NYSE and AMEX, and that order size plays an important role in determining priority on NYSE and AMEX.
Quote from jimrockford: The following advice is highly misinformed: [I said] If you see a bunch of prints going off at your bid price and can identify where they are executing, and you see the bid drop on that venue, it might be worthwhile to re-send your order there, but probably not. You are almost surely better off re-routing your order in such a situation. You need to route to where the action occurs. I base this on my very careful, blow-by-blow analysis of specific orders, how they executed, and how they could have executed if they had been better routed. I dispute your contention that it is "highly" mis-informed. You're bidding for WXYZ on BRUT. INET is also on the bid and getting drilled by someone. He gets exhausted and the INET bid drops, leaving you alone at the inside. Why would you want to re-route, and possibly fall behind hidden orders that may have just been routed there above your price to try to catch some flow? If there's any more stock for sale, they're going to hit you where you are except in the rare instance where they can't see BRUT for some reason. Especially since BRUT is cheaper to hit in this particular case. As you go on to say, and describe some of the nuances, routing in a fragmented market is quite difficult, and there are no simple answers like "always route to INET", as a lot of newbies seem to want. Having a really good smart router that has been developed by equity traders is probably the best solution. Perhaps someone can comment on how well-designed the solutions at some of the prop firms are, since most of the retail routers seem to have been panned here at one point.
Alanm, if you keep your bid on BRUT, even though we assume you are alone at the inside, and that the bulk of WXYZ volume trades on INET, you will only suffer. Your bid will probably be hit sooner if you move it to INET (assuming the bulk of WXYZ trades on INET). If you keep it on BRUT, then trades can go by at your price on INET, either because of hidden INET bids at your price, or because of INET bids submitted and displayed at your price after you decided to remain on BRUT. If you re-route to INET, where we assume the bulk of volume executes, you will have priority over any hidden INET bids at your price, because displayed orders have priority over non-displayed orders (on INET, BRUT, ARCA, etc.) You will also have priority over any subsequently submitted INET orders at your price. If you stay on BRUT, then you will NOT have priority over any INET bids at your price. Note that by re-routing to INET, in our scenario, you only lose priority over subsequent or hidden BRUT bids at your price, but you gain priority over subsequent or hidden INET bids at your price, which are more numerous, since we have assumed that INET receives greater order flow. So the INET priority is more advantageous than the BRUT priority. You raise the risk that if your bid is moved to INET, it will fall behind hidden INET bids at higher prices. But you are ignoring the problem of hidden BRUT bids at higher prices, if you stay on BRUT. And you are ignoring the more serious problem of liquidity. If INET receives greater order flow, then it is more likely to have hidden bids above your bid, but it will also tend to have more marketable sell orders removing those hidden bids. The end result will be that your bid will be hit sooner on INET than on BRUT, where we rely on the assumption that INET receives greater order flow. The risk that your bid will never be hit at all, on BRUT, is much larger than it would be if you re-routed to INET (assuming INET has greater liquidity). You point out that BRUT is cheaper. BRUT may also have other advantages as compared to INET. But these are irrelevant, where we have already included, in our scenario, a postulate that the bulk of order flow goes to INET instead of BRUT. It is a mistake to assume that market participants route their orders rationally in the best interest of the customer. If that were true, nobody would ever trade QQQQ on the AMEX, for example. If that were true, then all retail brokers would offer connectivity to regional exchanges, but they don't. If that were true, then nobody would pay for order flow, but they do, because bribery succeeds in altering order flow to the detriment of the customer. And if most liquidity flows to a market center with higher fees, then this creates a good reason to pay those higher fees, so that one can interact with greater liquidity. If, on the other hand, we assumed that the bulk of order flow goes to BRUT, instead of INET, for WXYZ, then it would make sense to stay on BRUT. Market participants tend to trade where they expect to encounter opposing liquidity. So even if you think BRUT has better fees or other advantages, you only hurt yourself if you ignore our scenario's assumption that most business is done on INET. If you can execute more cheaply by routing to a less liquid market center, then you need to make sure that the cheaper fees compensate for the lack of liquidity, poorer fill prices, and increased delays and missed fills.
Quote from jimrockford: Market participants tend to trade where they expect to encounter opposing liquidity. My scenario was that you are now alone at the inside bid, on BRUT. If someone wants to hit the bid, they are CERTAIN to encounter the other side of the trade on BRUT (i.e. your bid) - anything else has a probability of less than 100%. Not that I disagree with the rest of what you're saying - it's just that it's all situational. Perhaps the original poster will gain from seeing some of the issues involved.
I don't know about NYSE but on NASDAQ I have better success in getting lmt orders filled and stp honored with ARCA than when using SMART-ROUTING where my pathetically small orders get through ISLAND for example. Now I avoid SMART especially on sensitive times like the open and go for ARCA, even if I have to settle for a worse price.