IEX, Diversion Tactics, and the New Market Share HFT Math!


Since the release of Flash Boys, we have become accustomed to side issues blown up to divert attention away from real problems and issues we have in our equity markets.

You saw what happened with the use of the word “rigged” – Twitter, the SEC, Direct Edge’s Bill O’Brien, some brokerage firms, and Congressman Garret all had coronaries. As a result, IEX’s Brad Katsuyama has had to fend off numerous attacks from multiple fronts, and constantly correct a series of lies and innuendo spread by those entrenched in the status quo. For example:

–          BNP Paribas claimed that Michael Lewis owns an IEX stake.

–          Bart Chilton wrote a NYT July 7th 2014 op-ed titled No Need to Demonize High Frequency Trading in which he claimed HFT is 50% of volume on other stock exchanges, AND ALSO 50% of volume on IEX.

–          Raymond James’s Patrick O’Shaughnessy has written three brokerage reports in one month about IEX, questioning market share math, including IEX Market Share: Lies, Damned Lies, and Statistics.


Brad Katsuyama finally felt compelled to answer these allegations publicly, and he did so in this Bloomberg View piece – Flash Boys and the Speed of Lies. He closed that op-ed with this:

With so many billions of dollars at stake, it is not surprising that some people will spin rich fictions about IEX — to deter others from believing in a fairer stock market. This is a battle being fought with words and numbers. So before you accept them as fact, make sure you consider the sources and their motivations.


In this morning’s note, we want to make sure you know those sources and motivations – Chilton’s and Raymond James’s. And we want to concisely address the “market share math”.


Bart Chilton

Bart Chilton worked for the CFTC as a commissioner and public servant for years. He is best known in that role for coining the term “cheetahs” to describe high frequency traders, and has repeatedly advocated for more regulation for these firms. See here, here, and even in this article, where he says HFTs should be fined by the second!

Then a funny thing happened on the way to the forum… Chilton got hired by law firm DLA Piper, where he is now working with “clients who will be subject to increased regulation over the next few years”. We read that to mean HFT firms (ultimately time and lobbying disclosures will tell the tale).

Since he was hired by DLA Piper in the above role, Chilton subsequently wrote his HFT-friendly NYT Op-Ed linked to above. The change in his tone was… notable.


Raymond James

We do not know very much about Raymond James, and its analyst Patrick O’Shaughnessy. We do know this, however:

–          From public disclosures, his firm routes a substantial amount of its order flow specifically to HFT firms.

–          Raymond James writes research reports on competitors of IEX (like NASDAQ).

–          Raymond James expects to receive underwriting fees within the next three months from NASDAQ.

–          Go have a look-see who NASDAQ’s largest customers are by volume and revenue.


Does the above present a material conflict of interest? Maybe. Maybe not. You decide.

“Market Share Math”

 Calculating a specific class of market participant’s market share volume in the market place is no easy task. It is not science. It involves estimates, discussions, as well as hard data. Probably nobody understands this better than Tabb Group’s Larry Tabb.

Tabb has estimated “HFT market share” over the past five years, and his estimates have ranged from 50% to 73%, depending on the year as well as his data inputs. In this CNBC piece from September 2010, Tabb puts the number at 56%, and breaks it down as follows:


–          HFT                       56%

–          Institutions            17%

–          Hedge Funds        15%

–          Retail                     11%

–          Other                     1%

–          Sum                       100%


Please note that Tabb’s figures have the percentages summing to 100%.

In roughly the same time frame, in this 2009 Bloomberg article, Raymond James’s O’Shaughnessy claims that HFT is 70% of the equity marketplace.  He should be asked whether his estimate then, and his estimate now of about 50%, are calculated the same way Larry Tabb calculates his 50-70% estimates. In other words, Mr. O’Shaughnessy, what are YOU using in your denominator? Do you have your percentages summing to 100%?

 Why do we point this out? Mr. O’Shaughnessy claims that Brad Katsuyama needs to do the “Market Share Math” better. He believes that Katsuyama is using single counted numbers to estimate that HFT is 15-17% of the volume in his pool, and that accounts for the difference between Brad’s numbers and his own, since Brad’s numbers should be doubled to 30-34%.

We believe Mr. O’Shaughnessy needs to do the math batter. Here is why:

–          30-34% is still less than 50%.

–          Percent comes from the word combination “per centum”, which is per 100, and not per 200.

–           Also, Katsuyama calculates his figures using the same methodology that the good folks at Tabb do – they believe that market share sums should add up to 100%.  How does Mr. O’Shaughnessy do it? How did he do it when he agreed with Tabb’s numbers back in 2010? What was his methodology then? What is it now? Was he inaccurate then, or is he inaccurate now?


Here is the bottom line. If you want to claim that IEX has HFT accounting for 30-34% of its volume, then you have to be consistent and then say that on the other stock exchanges, HFT accounts for nearly ALL of their volumes. If you want to say that HFT is 50% of volume on the other stock exchanges, then you have to say that HFT is 15-17% of the volume on IEX. Right?


Closing Thoughts

 The whole “what percentage of volume is HFT on IEX” discussion is yet another diversion thrown out by those interested in our current markets remaining as they have been for the last 7 years. It is silly. We could not care less who is on the other side of our clients’ trades in our anonymous (kind-of) marketplace, as long as all of us are competing on fair platforms that treat every participant equally.

We do object to trading platforms creating artificial and needless opportunities for “arbitrage” by design. If a trading center is maintaining and encouraging a plumbing arbitrage, it is subordinating the treatment of some clients to others.

We have been very pleased in the last few months to see the debate and focus on market fairness advance responsibly. Even at our recent Cowen Themis Market Structure Teach-In, we were graced by Hudson River Trading’s Adam Nunes challenging our own viewpoints on market structure. While not in agreement with us on all issues, Adam eloquently presented his viewpoints on what fair market should be like, and he has challenged us to rethink and question some of our own tightly held beliefs. We thank him for that! And that is how debates move forward.

Distractions such as the recent “what percentage of volume is HFT on IEX” discussion, on the other hand, move them backwards. They especially do so when the debaters can’t even be consistent with their own “math”.