To: Eric P who wrote (7042 ) 4/26/1999 12:37:00 PM From: Dan Duchardt Read Replies (2) | Respond to of 12617
Obviously, I agree that SOES and SNET are basically useless now. However, I think that only one simple change is needed to make the market 'fair' again. => Eliminate the 17 second wait that market makers are currently allowed before updating their quote. I'm not sure this is all that needs to be fixed, but I agree with you that this is the big one. In fact, in my experience the uncertainty of whether or not SOES quotes were live, or dormant, was the only reason for choosing SNET for 1000 shares or less, and the primary reason for choosing an ECN to get executed. What I found was that as soon as a preference was directed to an MM at the inside market, instead of executing my order the MM would most often pull his quote down and decline the order. This is a clear violation of NASD's own interpretation of the "Firm Price Rule", but how can you prove your order hit his screen before he pulled down his quote?? You can't and the MM knows it. However, I believe it was you Eric who noted that in your automated trading program you have seen sequences of numerous declines as you go through the list of MMs preferencing at their quoted prices. That's pretty compelling evidence!! It seems to me, the 17 second siesta has created a burden on the MMs to deal (fairly or unfairly ) with many more SNET orders. And, as you have stated: Needless to say, the stock will quickly ignore the market makers quote and trade at a much higher price against the available ECNs. However, this disrupts the market in the stock and should be eliminated I believe it is the primary reason for emergence of the popularity of ECNs as preferred execution systems. The big complaint of NASD, which they use as a foundation for their proposal, is that it is unfair to the MMs to be required to monitor two, asynchronous systems, both of which expose them to firm liability to execute orders, a sort of double jeopardy if you will. As you and I (in an earlier post) have both noted, the MMs can and do freeze the inside market at virtually no cost compared to their working capital, resulting in big, sudden jumps in the quotes, and fueling even bigger jumps in the prints as the ECNs are executing. To give the devil his due, I should point out that NASD acknowledges the 17 sec delay as unnecessary and are proposing that it be reduced to 5 sec. I agree with you that it should be done away with entirely, but I believe the NASD would view that as being too difficult a change. It's easy to reprogram the 17 to 5. It's more difficult to program a new response that will automatically change the MMs quotes after every execution, or require him to create a new one. I expect you know better than I the options the MMs now have for managing their quotes. Perhaps you have an idea how to implement the change you are suggesting in a way the NASD could not dismiss as being too complicated.