Streetwise Professor

Bart Chilton, apparently all knowing, has decided that the priority rules that exchanges have chosen are all wrong, and need to be changed.  In his continuing war on high frequency trading (not to be confused with his war on commodity speculation), he has condemned size priority rules.  He claims that these give “cheetah” (stop it Bart!  You’re killing me with your clever repartee) traders that can move faster than others an unfair advantage that they use to game the markets.

Any centralized trading mechanism has to have priority rules—rules that determine the sequence in which orders are filled.  Price priority—the highest bid and higher offer get executed first—is universal, but secondary priority rules are needed to break ties between bids and offers at the same price.  Different secondary priority rules are possible.  A common one is first-come-first-served.  This is called time priority, giving the earlier orders at a given price priority in execution over ones entered later.  Another secondary rule may give some priority to designated market makers.  Still another is size priority, which gives bigger orders preference over smaller ones.  There can also be hybrid rules which use time, market maker status, and size to allocate trades between quotes entered at a given price.

Each priority rule provides different incentives.  There is no obvious “best” rule: every alternative involves trade-offs.  Chilton apparently believes that time priority is best, and that size priority is unfair.  But he provides zero, zip, nada in terms of an analysis of the trade-offs involved.

Size priority, no surprise, gives an incentive to quote in size.  This can be very beneficial, as it means that big market orders can be accommodated more readily without large price movements.  This can be important in markets in which liquditiy demanders desire to trade in size—which is the case, and increasingly so.  For somebody worried about flash crashes, you’d think this would be a virtue.

Exchanges internalize many—perhaps not necessarily all, but many—of the costs and benefits of alternative secondary priority rules.  Whether they are competing, or monopolists, exchanges that choose priority rules that result in the creation of the amount and type of liquidity that is preferred by most traders in the marketplace generate more volume and can charge higher fees and can make more money.

A coherent objection to exchange choices would require some sort of demonstration that there is an externality resulting from these choices.  Chilton, to belabor the obvious, does no such thing.

Chilton says that high frequency traders program their algorithms to optimize results given the secondary priority rule.  Really?  Wow.  Who knew?  But that would be true regardless of the rule.  If exchanges went with strict time priority, the HFTs would adjust their algos accordingly—and those with the fastest connections and best programmers would win.  The algorithms would be different, but odds are that the winners and losers would be the same.

It is particularly strange that Chilton simultaneously objects to size priority rules and HFT in which the fastest algos win.  Under time priority, speed (and hence things like collocation, computing power, etc.) would be even more important.  If he has a problem with the race going to the swiftest, protesting against size priority rules is a strange way to go about it.

Unless and until there is a convincing demonstration of that exchanges face systematically distorted incentives in their choice of priority rules, regulators should butt out.  Objections to priority rule choices should be based on a clear and concise identification of some sort of externality.  For absent such an externality, exchanges have every incentive to make the trade-off that maximizes value.  Not everybody will be happy with the choice because traders are diverse.  But that’s the nature of the real world in which trade-offs are pervasive, as opposed to the Never Never land where too many regulators and legislators seem to dwell.