New Flash Crash Research Back to Table Of Contents Executive Summary We have uncovered new evidence that the primary cause of the Flash Crash originated with an aggressive, complex, cross-asset strategy that overloaded quotation systems (public and private) leading to the loss of liquidity and other ill-effects identified in the SEC report. We have also uncovered evidence that NYSE's direct feed, OpenBook Ultra, would consistently delay 10 - 200ms when quote traffic exceeded a specific threshold amount -- a relationship that was so easy to spot, it had to be widely known by sophisticated trading shops. Curiously, the SEC, which analyzed the same data, made no mention of this crucial information. We also find that CQS was completely saturated (full), to the point where one of the 12 distribution lines began shutting off for 1/2 second every second during the height of the flash crash. The quotation system overload occurs before the beginning of the so-called hot-potato effect mentioned in the SEC report. After publishing our Flash Crash Report(s) last September, we have been developing better real-time monitoring tools to quickly identify data feed delays and determine the underlying cause. Along the way, we discovered unusual market behaviors, some of which we have reported when time permits. Our main focus, and primary concern however, is the runaway explosion and degradation of equity and option quotes. Next month, July 2011, CQS capacity will increase to a whopping 1 million quotes/sec which is 4 times the capacity on the day of the Flash Crash. Each time CQS capacity has upgraded, saturation events could be detected weeks, if not days later. This is just quotes on NYSE, AMEX and NYSE/ARCA stocks. Before the end of the year, unless someone steps in and puts an end to this nonsense, you will need a gigabit connection for non-Nasdaq equity quotes. This growth in quote traffic is not helping investors. It's not helping traders. It's not even helping HFT (in the sense that the information carried by the quote provides pricing information). Cisco or Juniper may sell more gear because of it, but it's destroying the notion of a fair and honest market. You see, when the market is quiet, it appears to function very well with tight penny-wide spreads and quotes that have little or no latency. When there is significant news, however, those penny-wide quotes disappear faster than your mind can read the last word of this sentence and latency skyrockets. Study after study misses the wide disparity between a quiet market and an active one by using 1 minute samples, or 1 minute snap shots (which were used in the SEC study). In fact, both CQS and OpenBook (that's NYSE's direct exchange feed) were thought to have ample capacity during the flash crash, because the network operators measure such things using 1 second samples. It may seem petty, or downright silly to be concerned with sampling at higher rates than one second. But we've past the point of silly a long time ago, and are working through the realm of absurd now. Did you know for example, when the market is busy (hundreds of times on an active day) that 20% or more of the quotes have already been canceled before they've even been sent by CQS? And depending on where you live, up to 100% will have been canceled while traveling at the speed of light towards your computer screen. Do we really need to know, for example, that the bid size changed between 20 and 25 a hundred times in the blink of an eye? What we really need to know, is if we have any reasonable chance at all of buying at the ask price, or selling at the bid. And that is precisely what is missing in today's quotes. More, faster quotes will not improve things. What we need are reliable quotes, where a bid means someone has real intentions of buying and an offer means someone has a real intention of selling. OpenBook Delays Quote Saturation Source of quote surge just prior to 14:42:44.100 event. CQS Saturation Inquiries: pr@nanex.net Publication Date: UNKNOWN http://www.nanex.net This report and all material shown on this website is published by Nanex, LLC and may not be reproduced, disseminated, or distributed, in part or in whole, by any means, outside of the recipient's organization without express written authorization from Nanex. It is a violation of federal copyright law to reproduce all or part of this publication or its contents by any means. This material does not constitute a solicitation for the purchase or sale of any securities or investments. The opinions expressed herein are based on publicly available information and are considered reliable. However, Nanex makes NO WARRANTIES OR REPRESENTATIONS OF ANY SORT with respect to this report. Any person using this material does so solely at their own risk and Nanex and/or its employees shall be under no liability whatsoever in any respect thereof.