The problem with the sell off is no one can tell us what happened
whydoeseverythingsuck.com
whydoeseverythingsuck.com
This is Just Plain Wrong. The post offers no evidence for this. For my evidence I'll use the fact that trades were occurring at almost all price levels, which is why some had to be canceled. For the stocks that "went to 0", that in fact means there were no willing buyers for anything other than $.01 (I doubt any trades actually occurred at 0)
"There has to be some explanation, and I can't imagine why it would take days, or even hours to figure out how a $50 stock could get to a price of $0 with no one able to purchase as that price is dropping"
Yes, the answer is that when a deal looks too good to be true, you check it out for at least another second (or perhaps 10 minutes, the duration of the "flash crash") before you put millions to billions of dollars of money into something. Again, I reject the idea that no one was able to purchase at any point.
http://www.washingtonpost.com/wp-dyn/content/article/2010/05...
I've spoken to a recently retired HF trader about it to verify that it seems wrong and he says that the way that quote is written it doesn't make sense.
I've read some articles suggesting that the drop was more dramatic because automated trading systems were taken offline. I've read that the drop was more dramatic because some stocks are sold on multiple markets with different emergency stop conditions.
First, the amount of data is extremely large. See the BATS connectivity manual, page 8. They require a MINIMUM connectivity speed of 1Gbps PER FEED. Page 9 shows that the peak rates are higher than this. Note that this is only one exchange of many. http://batstrading.com/resources/membership/BATS_Connectivit...
Second, even if you could make sense of all the data, there's people on the outside creating the data. A full model of what happened would require knowing what is going through their minds.