The Speed Bump From Hell; What's The Reaction To IEX?

Includes: BATS, ICE, NDAQ, RY
by: Kurt Dew


The IEX exchange designation from the SEC is now a foregone conclusion.

But IEX is the question, not the answer.

IEX is a new competitor among many, new and old.

Among them may be the "speed bump from hell."

Johnny come lately, the new kid in town,

They will never forget you 'til somebody new comes along.

- The Eagles

Having received the endorsement of the SEC staff for IEX's exchange designation, the approval of IEX's exchange designation is a foregone conclusion. Perhaps now market participants might attempt a more mature dialogue concerning the issues that underlie the incredible brouhaha that IEX's application created.

This dialogue can follow one or more of three paths:

  1. Other new exchanges that meet customer needs better than either IEX or the High Frequency Traders' pet exchanges [the New York Stock Exchange, a subsidiary of ICE (NYSE:ICE), Nasdaq (NASDAQ:NDAQ), and the BATS exchanges (BATS:BATS)].
  2. The brokers' pet exchanges modify their trading practices to become more competitive with IEX. This might include, for example, changing rules to make the exchange in question less biased toward HFT traders.
  3. New SEC regulation.

Why was nothing settled by IEX? Because IEX represents something other than an answer to the question, "Is existing stock market trading technology an efficient way to 'run a railroad?'" IEX exchange rules and regulations are based upon a philosophy, not simply a technology.

I believe most discussion of the IEX "speed bump" misses this fact. For that reason, investors will be surprised when the debates about exchange technology don't end with the approval of IEX.

IEX's approach was never a solution to the inherently mathematical question, "What is the most efficient way to execute transactions in the market for common stocks?" Nor is IEX's configuration simply the introduction of a "speed bump." There is at least one other IEX innovation designed to prevent HFT attempts to manipulate IEX prices - the "discretionary peg." IEX's approval is just the start of a dialogue, with a technological dimension and an economic and philosophical dimension.

To see why IEX's innovations have no direct relationship to stock market efficiency, it is helpful to take a backward look at IEX CEO Brad Katsuyama's exodus from the role of a trader at Royal Bank of Canada (NYSE:RY).

The basic issue he was addressing at RBC was his inability to perform for the bank and its customers in filling an order at the market bid or offer. In other words, he wanted to do his job and make money in the process. To make a long story short, he and his colleagues found a very clever way of accomplishing this not-so-simple goal. This process had absolutely nothing to do with starting a new exchange.

But to attract customers to his new trading method, he needed to explain to them what he was doing and why it worked. And his customers were impressed, for four big reasons:

  1. He had devoted resources and time to find a new, better way to fill their orders.
  2. The sell side, the buy side suspected, saw meeting their buy side customers' needs as secondary to exploiting their customers' activities, manipulating customer trades to earn trading profits in the sell side's own accounts.
  3. The old exchanges had become the playthings of the HFT traders. The exchanges were locked in a symbiotic relationship with the HFT traders, selling them advantages in trading and accepting the adverse effects on other customers.
  4. Following the end of exchanges as membership organizations and the public listing of exchange corporate shares, the ownership and management of the exchange firms had been increasingly dominated by the large sell side dealers. Thus, the bias of trading technology toward HFT was locked in.

It was the investment houses on the buy side that first suggested Katsuyama begin an exchange. So a philosophy of mine, changing the efficiency of the market without creating new, more complicated, SEC rules, happened to be the philosophy at the heart of the IEX speed bump.

IEX does not object to the behavior of the other exchanges. The pet exchange bias is seen as a market opportunity for profit for IEX, to be exploited by providing better service and attracting business.

A second sign that IEX is a creature of philosophy, not mathematics, is IEX's opposition to the SEC's (ill-considered) proposal that the markets define one millisecond slow as "current."

But the competitive doctrine behind IEX results in a process that is endless. Indeed, the open-ended competitive model is the secret to Adam Smith's success in the history of philosophy. Like evolution, the competitive model posits the absence of an ideal.

So what's the next big thing? If I knew, I wouldn't be writing about the subject. I'd be applying for a patent. But…

One suggestion, by Eric Budish, Peter Cramton, and John Shim, to be found here, is to hold "frequent batch auctions." This way of processing orders would attempt to convert Einstein's time-relative world to Newton's discrete time world.

In other words, for trading purposes, time would be divided into one millisecond (or other convenient period) intervals. In this world, the IEX 350 microsecond speed bump would not affect trading at all, since all orders arriving after crossing the speed bump would be processed at the "same time" (in the same one millisecond time interval) that they were placed.

In effect, this "frequent batch auction" proposal is the "speed bump from hell" because it eliminates the latency advantage created by the HFT pet exchanges and their co-located computers, obviating the need for an IEX speed bump.

Questions arise:

  1. Would it be possible for a single exchange to implement this system while leaving the other exchanges to compete?
  2. Should this become an SEC regulation?
  3. If this becomes an SEC regulation, what is the purpose of multiple exchanges? That is, if the effect of the frequent batch auction is the expected effect - that the orders on all exchanges for each "trading time" are identical - then what is the basis for competition between exchange management firms?

These questions suggest to me that this proposal is indeed "the speed bump from hell." The level of the HFT howls from a proposal to introduce this trading method would indeed be a wonder to behold.

I would like to see the speed bump from hell advanced as a competitive alternative, but I suspect the 1 millisecond "de minimis" time interval proposed by the SEC, which was roundly rejected all around, suggests that the SEC, charged by Congress with the responsibility of designing "fairness" into the market - fairness is a quaint concept that defies definition - may be thinking of introducing the speed bump from hell as a regulation.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.