Discovery Processes at Work
The first post I wrote on SWP related to the pricing of message traffic on electronic trading systems. This has become a major issue in the debate over HFT, with (as I noted last week), the SEC musing about impose by regulatory fiat pricing schemes that would penalize cancellations.
But as Jeremy Grant notes in a couple of articles in the FT, exchanges are already implementing pricing schemes to deal with “stupid algos”-and I would argue, “evil algos” too. There is a process of market experimentation going on, with some disagreement among exchanges. Deutsche Borse is considering a pricing scheme similar to that the SEC is considering-imposing a tax on all cancellations above some level. ICE has criticized this idea, arguing that such a scheme does not discriminate between “good” cancellations and “bad” ones: that was my objection to the SEC idea. Instead, ICE assess charges for cancellations based on how far the orders are away from the inside market. Orders close to the inside bid-ask are more likely to be liquidity enhancing, and less likely to be opportunistic efforts to clog the exchange systems. This is far more discriminating approach.
But there is a bigger point here: who should be making these decisions, exchanges or regulators? A strong case can be made for the former. Exchanges internalize many, and arguably most, of the costs and benefits. Opportunistic algos reduce the derived demand for exchange services, costing the exchange money. Good algos that contribute to liquidity increase the derived demand for exchange services, making the exchange money. Exchanges have an incentive to get the pricing right to encourage the good and discourage the bad, and moreover are likely to have better information, and to be less susceptible to rent seeking influence activities, than government regulators.
Different exchanges engage in different experiments on pricing. This contributes to a discovery process that is likely to reach an efficient outcome than the process that legislatures and regulators use to set prices and procedures. Exchanges that develop effective pricing policies gain imitators: those that don’t, lose business.
Yes, it is possible that there are externalities that exchanges don’t take into consideration. But only fools believe that regulators can understand, quantify, and correct these externalities. Indeed, the process by which regulators internalize such effects is often highly distorted and inefficient, due to constraints on information and public choice problems.
Exchanges are clearly aware of the relevant issues, are highly motivated to address them in a discriminating and effective way, and are in fact trying to do just that.
Regulators were once defined to me as conflicted by their desire for power, laziness and fear of being made to look stupid. While this is harsh, and unfair to many, your point has been that ruling by fiat when one has no skin in the game is not a good idea, especially the more technical and detailed the decisions become. They do have the power, however, to force things: maybe the model we pursue is one where the markets and exchanges make these decisions, and the regulators enforce the law and keep all the players in the defined markets (e.g. no purposefully non transparent practices as arose in the old days in the third, and fourth markets for NYSE, etc.). This would require that regulators’ (and their political masters’)power grabbing, mission creep, or whatever term you like is kept firmly under control.
Comment by sotos — March 5, 2012 @ 9:32 pm
And that’s where I am skeptical, sotos-I just don’t see how to keep the grabbing and creeping–the grabbing by creeps, if you will–from occurring. Everything pushes in that direction, and very little seems to push the other way.
The most perverse aspect of this is that egregious failures are the pretext for expanding regulators’ powers. They (and their political masters) are experts at laying blame on markets and market participants, even when regulators and legislators are primarily culpable.