GO
Loading...

Nasdaq takes some responsibility for 'flash freeze'

Thursday, 29 Aug 2013 | 6:47 PM ET
Getty Images

Nasdaq has provided an update to the trading halt of Nasdaq-listed securities that occurred Aug. 22.

The broad outline is similar to the story I constructed from different participants and published in my Trader Talk blog on Tuesday.

(Read more: Nasdaq mea culpa not new at all)

While Nasdaq has leaked out much of their version of what happened over the last few days, the update today does advance the story in several important respects.

1) Nasdaq accepts some of the responsibility for the incident, saying, "A number of these issues were clearly within the control of Nasdaq OMX."

2) They continue to place a large part of the blame on NYSE/Arca, the NYSE's electronic trading platform. According to Nasdaq, the SIP's processing capacity was overwhelmed when a matching engine at NYSE Arca had trouble connecting to the SIP. The SIP (security information processor) is the network that carries the quotes and trades for Nasdaq.

There were over 20 connect and disconnect sequences, each of which consumed "significant resources" that eventually overwhelmed the SIP. Essentially, a huge burst of quotes—many of them stale—shut down the system.

Nasdaq's conclusion: There was a "latent flaw" in the SIP software code.

But was the NYSE a part of the problem? The implication is that it was, since the problem started when Arca was trying to reconnect with the SIP.

(Read more: Finger-pointing on Nasdaq outage)

The only external analysis of this event I have seen was done by Eric Scott Hunsader at Nanex. His conclusion: "[T]his burst of quotes had to originate with the SIP itself. It is highly unlikely that anyone or any process external to the SIP environment could have caused or triggered the resending of old quotes. ... Arca, and anyone external to the SIP could not have caused the quote bursts."

Here is the Nanex study.

3) They promise better communication. Nasdaq says it is "undertaking a comprehensive review of the policies and procedures for communicating with customers." That's good news. It has been widely criticized for having to little to say during the crisis.

They have sent a letter to their listing clients that contains a new Commitment to Transparency statement in which they commit to "providing you with information as we know it."

Nasdaq: Working on potential design changes to SIP data feed
CNBC's Bob Pisani reports the Nasdaq's trading freeze was a failure partly due to unprecedented volume sent to SIP.

4) Don't blame high-frequency traders, who say, "Our review indicates that high frequency trading played no role in the technology events of August 22."

No comment from the NYSE, but I've reported officials are furious that the Nasdaq would try to blame them for this. They have freely admitted that a matching engine briefly went down and tried to reconnect to the SIP. But they are adamant there is no way an attempt to reconnect to the SIP should have caused the SIP to fail, whether there was one attempt to connect or twenty.

If that is the problem, NYSE officials say, than there is indeed a design flaw in the SIP.

Well, at least NYSE and Nasdaq can agree on something.

The SEC tells me it is still in the information-gathering stage and has called for a meeting Sept. 12 to get the parties together.

By CNBC's Bob Pisani

Featured

  • A CNBC reporter since 1990, Bob Pisani covers Wall Street from the floor of the New York Stock Exchange.

Wall Street