Actually I was having NYSE delays last week, same as you with ~30sec lag for order to go green. I called and they didn't know what the deal was (surprise) but I assume it's related to the general disarray that IB has been experiencing in the past 2 weeks or so with hardware overall. This week disconnects have still been horrendous but NYSE delays seem to be gone. TMBR executions on NYSE have been instantaneous in my opinion...much better than having to wait for NYSE execution.
I have had only good experiences with TMBR so far, always instantaneous. HOWEVER, with supersoes a situation where my order seems to be "looked at" happens quite frequently. I realize I have said this before, and I realize there are people on this board who love to instruct me regarding SOES rules and regulations, therefore I will not say my orders are being looked at, only that they seem to be looked at, and I will just describe what I mean: Sometimes IB is just a little slow (particularly at times of heavy volume). This situation can be recognized by the fact that my order for a nasdaq stock stays blue and on "BEST". Then, after a few seconds, I either get a fill or the order turns green and "ISLAND". At other times, however, the order turns "SUPERSOES" immediately and after that it takes a few seconds before I get a fill or get rerouted to "ISLAND". I would think that the observed delay in these cases is not due to IB, and I also don't think a supersoes computer would take, like, 10 seconds to match the order and send confirmation. Of course I am aware of the fact (thanks to some very well informed people who keep reminding me of this beautiful truth) that supersoes is absolutely automatic (at least for the 100-share orders I am talking about), so there is absolutely no way a market maker can make money by offering or bidding on supersoes instead of, let's say, on the Island ECN. I guess it will just remain a mystery... Seriously, I don't think MMs bend the rules, but I also don't think that supersoes (or soes) rules prevent your order from being "looked at" under all circumstances, even if its size is just the amount quoted at the time you transmit the order. In case you find that unlikely, just think about how many traders see a certain quote at a given second, how many of them try to take it at that very second, how long it takes the MM to change a quote, how long it takes for this change to reach each of those traders etc. The bottom like is, a perfect broker would offer customizable sets of exchanges that can be configured to be considered for BEST execution. In other words, if I want to define "MYBEST1" as the best among {ISLD, ARCA, INSTINET}, then I should be able to do so and use "MYBEST1" for critical orders say in fast markets. Of course, if you asked me if I would prefer IB's commissions and execution abilities to $.012 per share commissions and customizable sets of exchanges, I would not hesitate a second to choose IB as it is right now.
as for tmbr routing.... I was on a call the other day where the chairman explained the TMBR routing. the algorithm is designed to give an immediate or better fill for the markets that TMBR may be equal or better than the NBBO. Basically, TMBR just becomes another exchagne in the list of exchanges. If an order i sent to a makret and TMBR's internal price is equal or better than the NBBO, the order will be instantly filled and sent back to the client. It doesn't play any BS games like holding an order to see if it can scalp out. It doesn't know or care about client positions. I'm not aware of any of the internalization issues from this AM (not a cop out, i just am not involved in US trading) but I did see a software entry stating "improvements for internalization". Thus, I'll assume things will be better tomorrow.
Can you describe how the fills were, in relation to the current quote and market? I was having a speculative conversation with someone about how they would approach this for listed stocks, and still give you the benefit of the price improvement that occurs on most NYSE fills. I speculated that TMBR would have an internal bid/ask that it showed to the BEST routing algorithm. What I don't know is how they would arrive at that bid/ask on a continuous basis, without intimate, human-based knowledge of the market for a stock. For example, if the market for NOC shows 10x115.00 - 10x115.25 during the afternoon lull, and the last print was 5x115.15 without changing the quote, I might expect a buy order for 100@115.25 sent to NYSE to get a fill at 115.15. Are you getting price improvement from TMBR is such circumstances, or are you just getting filled at the bid or offer? Note that I'm not making any judgements one way or the other - just wondering how it's working. We all know how frustrating it is to have a fast NYSE market run away from you when you would have been happy with a fill at the bid/offer...
as for how TMBR determines a price.... At any given time TMBR has theorectical values for the hundreds of thousands of products in its system around the world. Obviously there are many inputs that determine these values of which I won't go into other than to say that even if the underlying is not moving, it is possible that the internal theoretical value will change. Intimate knowledge of the underlying stock is not required to make a two sided market and excellent hedging systems. From there is it quite simple. Certain parmeters and/or factors determine a spread and a market is made. The TMBR quote basically becomes another internal factor in determining the BEST market price. If TMBR is slightly better than the NBBO bid/offer you'll get price improvement. The real benefit from the clients should come from immediate fills and hopefully some price improvement.
def, from IB management: "There was a technology bug and we are fixing it immediately. The way things are suppose to work is that when TH flags that they are willing to match or beat the NBBO, the order is done, with no further matching or delays" It should be fixed by the end of this week.
Makes sense. My comment was made assuming that you want a neutral position in the stock. I guess there are times, though, when that might not be the case. If IB operates like other option MMs, I'm guessing they are sometimes looking to put on a stock hedge immediately after getting hit for an option execution, usually without trying to make the spread on the stock (i.e. willing to buy at or above the offer). If the TMBR "quote" reflects this view of the stock, in addition to trying to buy the position via the normal routes (DOT/Direct+/floor), customers could theoretically benefit greatly by being able to sell at or close to the offer - a win-win.