i wasn't involved in these incidents - and i am not so much astonished at all that such things can happen (by accident or not) - but the main problem i see is that some trades get busted - others not. yesterday the es was the starting point - es trades got busted - other trades not. why not? because if they bust trades in other markets - where should they stop? bust trades in every single stock that went down also? i guess not. maybe they should not bust any trades - there are limit down barriers anyway - and as long as es isn't down at least 5% - what's the problem?
5% on the ES is a pretty huge move - and a pretty big "problem" if you're leveraged. They could theoretically introduce tighter limits for price moves over very short time frames, but there's a natural reluctance to interfere with "natural" price discovery. If enough of these events occurred, they might still be forced to introduce tighter rules constraining extreme price moves and large trades - or lose business - , but the next time we get a spike like this, it may actually be a real event driving the market way up or way down. Any of us can imagine what those events might be, and we all know far too well that they really do happen. When this problem came up with YM, I suggested that the exchanges should be urged to draft much more specific rules about busting and timely reporting - call it a "sunshine policy." (The decision by the CME was actually relatively quick, I think, at least compared to what happened with the YM event and certainly compared to the handling of outage fiasco.) Even when the trades get busted, the fat-fingered trader's firm does pay a price, but if there was a higher penalty for making errors of this sort, then firms would also have an incentive to introduce failsafes that would prevent them from making such mis-trades in the first place. Publicizing the name of the individual trader would be a strong disincentive. But I don't think the exchanges would want to get into heavy litigation with the trader's estate after some ruined 1-lotter tracked him down and blew him away.
Did have a long position on and a stop down there What happened though is that I have an alert set to warn me if there is a sudden move of 4 to 5 points as I usually do other things during the day. I cancelled my stop order on the alert as I saw no parallel move in the other indexes - just in time before the flush touches my price And yes, I waited out the whole 15 minute before exiting the trade. It was not as painful as the other times that are incidents driven which usually taken longer to recover. This is the exact reason why I do not completely trade on mechanical model.
SethArb since you are new here 14:08 is CST = 308PM EST I agree... the point i was making was re: the bulletin by IB... indicating the spike at 14:08 est. I thinks they screwed up... Nereus PS. Thnx for being gentle on the new guy...
sounds like a good idea - why not? "natural" price discovery - yesterdays action was also just natural price discovery - right? if it was all about avoiding interference - there wouldn't be limit barriers - and there wouldn't be busted trades - imho.
something just doesnt jive about the delay between when the trade happens and when they tell everyone it's busted. i think globex should introduce more sophisticated circuit breakers, or do a quick shutdown when this happens. i may have missed something in this thread, but mechanically, do we have a fix on exactly what type of trade can cause this to happen? if they go back and bust it afterward, then the system shouldnt accept the trade in the first place. i wasnt trading es, but several other globex futures got jerked to lesser degrees at the same time. personally, it triggered a short for me that probably wouldnt have been there otherwise, and that stopped out in a big way this morning.
how about this.. there's hackers in the system. they can't say "there's hackers in the system" for obvious reasons. so people will keep entering "erroneous orders" until they can figure out how to close the holes in the systems.
The rules are actually very concrete and it is easy to know immediately which trades will be busted. The CME will bust any mini trades that occur more than 6 points away from the min/max in the big contract. Yesterday the big contract traded down to 1002 on the move down and so any trades lower than 996 were busted. That is the standard policy. No need to wait to see "if" your trade will not hold up or not.