How Slow is the NBBO?

I have a very very busy day, so I’ll be brief.

I was reading Felix Salmon’s latest article for his Reuters blog, and something was weird. A the beginning of the paper, Salmon talks about another article I missed, by Rob Curran for Fortune. Curran writes : “Terrence Hendershott, a professor at the Haas business school at the University of California at Berkeley, wanted to find out. He was recently given access to high-speed trading technology by tech firm Redline Trading Solutions.” I have read a lot of Hendershott’s papers, so I knew that the study reported by Curran was published on July 8, 2013, in a paper called “How Slow is the NBBO? A Comparison with Direct Exchange Feeds”. I was surprised because the title of the paper was not mentioned by Curran, nor the link to the Berkeley website where I got the file as soon as it was published online (Salmon seems surprised too: “This study is never named, or quoted, or linked to, and I can’t find it on Hendershott’s web page,”).

A usual, as soon as an interesting paper is published, I print it (I am an old school guy). So I printed the paper around July 20, but I did not read it immediately. I read it just after Nasdaq turned Nasdark on Friday 23 August (because the SIP and NBBO was involved in the outage, I thought that some issues could be discussed in the paper). On that day the paper was still on the Berkeley website because I shorted the link through in order to send this tweet to some guys (don’t know if they read it):

Capture d’écran 2013-09-03 à 14.38.53

On August 23, 2013, the paper was still online

When I read Curran’s article, I tried to access the PDF file on the Berkeley website… but the file was gone. So, the paper has vanished between August 23 and August 30. My first question was: why? Is it because the paper is quoted in Curran’s article? Is it because of the sensational title “Make $377,000 trading Apple in one day”? Is it because the Nasdaq went dark–so discussions about the NBBO were banned?

Anyway, its seems that these days if you want to get an academic paper, you have to act like a high frequency trader: find it before cancellation! So, here is the HowSlowIsTheNBBO PDF file. I don’t really have the time to discuss about it, the three writers had access (through Redline Trading) to the data feed of BZX & BZY exchanges by co-locating a computer in Weehawken. Sometimes the paper states the obvious, sometimes other facts are interesting. Perhaps Felix Salmon could find answers to the questions he asked, and perhaps other #hft addicts may find arguments.

PS : Terrence Hendershott will talk here today (an academic event about hft far better than the TradeTech things), but “it will not be broadcast”, unlike other talks–in order to spare sensitive questions? Andrei Kirilenko made a talk yesterday, but was not very convincing.

Update 9/5/1013

That is very strange: as soon as I put the file online (just after Nanex did), the paper was back on the Berkeley website…



  1. Un universitaire gagne 377 000 $ US en une journée avec un système de transactions à haute fréquence.
    Un universitaire qui a fait l’essai d’un système de transactions à haute fréquence a généré près de 377 000 $ US de profits virtuels en une journée avec l’action d’Apple.
    News banques. Mardi 3 Septembre 2013.

    Il s’agit de Terrence Hendershott, professeur à l’école de gestion de l’Université de Californie à Berkeley, a rapporté CNN Money.

    M. Hendershott a eu accès au système de la société technologique Redline Trading Solutions. Il a pu constater l’écart de prix entre les prix des actions sur le marché et ceux affichés sur les écrans des courtiers.

    Le professeur pouvait ainsi faire des achats ou des ventes théoriques sur Apple en profitant des variations durant des fractions de seconde avant même qu’elles soient connues.

    Cet accès « direct » bénéficie à des sociétés spécialisées comme Getco et Knight Capital Group, ainsi qu’à des banques d’affaires comme Goldman Sachs.

    Il avait toutefois été montré du doigt pour un effondrement boursier soudain et temporaire le 6 mai 2010. Cette journée-là, le Dow Jones avait reculé de 9 % en quelques minutes, pour tout récupérer immédiatement.

    • Reprise de l’article de Rob Curran, mais en moins bon.
      “Un universitaire gagne 377 000 $ US en une journée” n’est pas exactement ce que décrit l’article d’Hendershott – il faut lire le papier attentivement…

  2. Mea maxima culpa ! Je n’ai pas pris le temps de vérifier l’original.
    Que le maître des lieux fasse subir à cette approximation une disparition à fréquence nulle ! :-)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s