Eric Posner and E. Glen Weyl have attracted a lot of attention for their proposal to create an agency that would approve trading of new financial products, in the same way the FDA must approve new drug entities before they can be sold to the public. Their initiative is fundamentally flawed, both in its diagnosis and its prescription.
There are a variety of problems. The most basic is their false dichotomy between contracts that are used for speculative purposes, and those that are used for hedging. In fact, speculation and hedging are highly complementary functions. The most important function of derivatives markets is to facilitate risk transfer. This risk transfer often involves a hedger on one side of a transaction, and a speculator on the other. The speculator may be, as Posner and Weyl claim, trading on the basis of incorrect beliefs. So be it. His existence allows the hedger to reduce risk, and makes both parties subjectively better off, although Posner and Weyl seem to suggest that it would be possible to determine objectively that the speculator is acting irrationally.
In other words, much speculation in financial markets facilitates hedging. To suggest that these activities are antithetical, or unrelated, as Posner-Weyl do, is highly misleading.
Posner and Weyl use mutual funds as an example of a salutary financial innovation. No disagreement here. But these can be used for activities that are commonly and properly considered “speculative.” Moreover, equity and bond futures are widely used to speculate on broad movements in stock and interest rate changes. But they are also used by pension funds and institutional investors to achieve the same type of objectives as mutual fund investments. So what, really, is the difference between mutual funds and derivatives?
There are indeed well-known problems with speculation. Much of it (including excessive acquisition of costly information, the development of faster ways to trade) is properly understood as rent seeking. But these problems are not uniquely associated with particular kinds of instruments, as Posner-Weyl insinuate. There can be inefficient acquisition of information for speculative purposes in derivatives that are widely used for hedging–including the agricultural futures that Posner-Weyl consider to be salutary. (Although it does not speak highly of their understanding of these markets that they use farmer hedging as their justification of ag futures, though that is hardly the primary hedging use of these instruments.)
Other possible problems highlighted by Posner-Weyl include the use of derivatives for tax avoidance or regulatory arbitrage. I say “possible problems” because the real problem can be with inefficient regulations or taxes, not the means that are used to mitigate these inefficiencies. But even abstracting from this difficulty, there is also a fundamental problem distinguishing “bad” derivatives that are used for tax or regulatory avoidance, and “good” derivatives that are used to hedge.
Back in the 1970s and 1980s, simple plain vanilla futures contracts were widely used in a tax reduction strategy called “tax straddles” involving a spread trade (buying one future, selling another with a different delivery day). The trader would liquidate the losing leg of the trade to generate losses to offset against other gains, or to convert ordinary income into capital gains. This strategy was eventually disallowed in 1981: the instruments themselves were not banned because they could be used to avoid taxes.
In sum, hedging and speculation are complementary; speculation is not always inefficient; and inefficient forms of speculation are not confined to specific instruments. Therefore, regulations designed to impede speculation by requiring the approval of specific products are misguided because impeding speculation interferes with efficient transfer of risk, and because the inefficient uses of derivatives cannot be reduced reliably by preventing the marketing of particular instruments.
This means that charging some financial FDA with the responsibility of choosing which innovations to permit and which to proscribe, based on the goal of eliminating speculation, is basically hopeless. Such an agency’s choices will be plagued by Type I and Type II errors.
Not to mention that whereas the effects of a drug can be determined with some (though incomplete) precision in clinical trials, there is no comparable way of determining how financial innovations will be used, the volume of their use, or their economic effects (in part because these effects depend on the scale of adoption, and the interaction of these products with other products in the market). All of these things depend on the choices of myriad individuals, and these choices can change over time: who would have envisioned, in the 19th century when wheat and corn futures were first traded in Chicago, that they would one day be used by index funds to offer diversification opportunities to pension funds?
Drugs are taken by individuals and their effects can be understood by studying individuals who take them: the systemic effects are just the sum of individual effects. Things are very different in financial markets. The introduction of financial products can lead to very complex responses in the entire system that cannot be understood or quantified in any clinical trial.
When I originally heard of the Posner-Weyl paper, I thought they would focus on how particular instruments can be systemically risky. They in fact pay very little attention to this issue. In this regard, it should be noted that the FDA/clinical trial model is completely inapposite. You cannot study the systemic effect of a financial innovation in an isolated trial: you have to see how it interacts with, and affects, the entire system.
And to try to understand these effects a priori is certainly futile. The system is too complex, and there is no way to understand how a particular innovation will fare in the market, and how it will affect the entire system. As a concrete example, no one truly understood the consequences of CDOs on subprime ex ante, just based on the characteristics of the instrument themselves. The risks embedded in the instruments were not well understood when they were first created, and these risks depended on the scale of their use, which was impossible to predict at the time they were first created. In the end, there were risks that were not widely understood at the time of introduction, and these risks had systemic import only because these instruments proved very popular and grew to massive proportions completely unanticipated when they were first introduced.
We actually have historical experience with ex ante entry regulations on financial innovation. Until the 1990s, no exchange could introduce a futures contract unless the CFTC approved it. The approval was contingent on a CFTC finding that the proposed contract passed an “economic purpose” test, and was designed in a way that would permit it to achieve this purpose. In some ways, the economic purpose test was analogous to what Posner and Weyl propose today.
This system was costly and reduced competition. Like many entry barriers, it was exploited by incumbents to hamstring competitors. I am not familiar with any dangerous product that was kept off the market through this procedure, and the introduction of good products was delayed. All pain, no gain. Why would we want to repeat that experience.
Maybe Posner and Weyl would disagree with my appraisal of the CFTC precedent. They don’t seem to be aware of it, in fact. Before making such sweeping proposals perhaps they should have performed an analysis of a similar system that operated for decades.
Being a Sam Peltzman student, I was never all that impressed with the FDA model in the first place, even in pharmaceuticals. I am even less impressed with the FDA as a model for the regulation of products that are fundamentally different from drugs. Especially when those advocating this model make basic errors about the instruments that they believe should be regulated, and the uses to which they are put.