At least since securities regulation became the business of (federal) supervisory agencies, their professional staff has been predominantly composed of lawyers. Attracting the best ones has always been a problem, because governments can’t pay staff as generously as the private sector. But a stint at the securities regulator can be a valuable move in a legal career, whether at the beginning (where the pay gap is smaller) or as a temporary side move to serve the public interest. While the opportunity costs can be high, the lawyer will get valuable connections, learn to think the way regulators think, thus making it easier to anticipate their moves, and, in short, earn higher fees in the next stages of his or her career. On the other hand, banking supervisors have traditionally relied less on lawyers and more on economists. Recruiting top talent among economists has always been easier for banking supervisors, especially when supervisory functions are combined with central banking ones, given central banks’ ability to attract and retain excellent economists more generally and perhaps, at least in the past, due to their lower opportunity costs compared to other professions.
While for decades the cat-and-mouse game has been between the supervisors’ lawyers and the supervisees’ lawyers, or their economists and banks’ management teams, things are changing. With information technology ever more to the core of financial markets, regulators are aware that tomorrow’s (if not already today’s) game is increasingly between their IT engineers and the FinTech engineers. The problem is that it will be far more difficult for regulators to attract top engineers than it has been for them to attract top lawyers and top economists.
First of all, the opportunity costs for top engineers appear to be much higher than for top lawyers’ and economists. According to Laszlo Bock, formerly Google’s head of human resources, ‘a top-notch engineer “is worth 300 times more than an average engineer.”’ Assuming that the average hourly rate of a lawyer in the United States is around $50, if that was true in the legal profession, it would imply a top-notch lawyer hourly rate in the region of $15,000, that is, one order of magnitude more than top New York lawyers are reported to be paid (see eg here and here).
It is also the case that, as hinted before, lawyers spending some of their professional life as regulators gain professionally from connections within the supervisory agency and from soft knowledge of how its internal governance, processes, culture, and power relationships work. Would the value of such connections and soft knowledge be as high for a FinTech engineer? At least at present, that seems unlikely. Developing good RegTech software would appear to be little dependent on the kind of soft knowledge about the internal workings of the supervisory agency that lawyers can exploit, for example, when advising their client as a potential target in a regulatory action.
Regulators may increasingly purchase externally developed software to exercise their core supervisory functions. Yet, an effective management of the relationship with the outside software supplier still requires good engineers. Putting an average engineer in charge may well lead to the choice of lower-quality, less than cutting-edge products. Unfortunately, there is usually no one else within the supervisory agency that can effectively question IT choices. As lawyers in many jurisdictions who use regulators’ websites for professional reasons can confirm, even user-facing products such as regulators’ webpages are often user-unfriendly. Yet, it should not be difficult for laypersons in the higher echelons of the organization to evaluate the IT department’s work by simply looking at the outcome.
Some RegTech firms may be willing to build a reputation as reliable suppliers of software to regulators so as to leverage on it on the market for similar services among supervised entities. But the conflicts of interest in a similar model would be significant. Most likely, regulators would require such firms not only to have firewalls, but also not to sell ancillary software to the market participants the software would help supervise. And if RegTech suppliers were not allowed to directly gain from the product developed for the supervisor, they may well find more profitable ways to signal their quality than becoming an IT supplier for the regulator.
To conclude, FinTech poses a serious challenge to regulators, which goes beyond the already difficult task of adapting their supervisory policies and activities to an environment dominated by artificial intelligence. No less fundamentally, it requires them to develop the ability to attract, retain, and deploy good engineers in addition to good lawyers and economists. Given the higher variance in quality among engineers and the lower attractiveness of a stint at the regulator for the best among them, the hopes that RegTech will make the job of supervisors easier seems to be misplaced for the time being. To the contrary, the risk exists that RegTech firms’ engineers will outsmart supervisory agencies more often than is currently the case with the lawyers and economists in charge. In a financial environment dominated by technology, the comparatively lower quality of engineers may even drag down the effectiveness of lawyers’ and economists’ work within the supervisory agency, as the latter’s models, policies, and enforcement actions will depend more and more on the engineers’ support.
Of course, things may change in unpredictable ways. Perhaps market structures and supervisory dynamics will develop in such a way as to make it valuable for some of the best FinTech engineers to gain the soft skills that top lawyers are currently keen to apprehend during their years as supervisors. But, in the meantime, there is reason to fear that the techies’ cat-and-mouse teams at regulators will be no match for FinTech firms.
Luca Enriques is the Allen & Overy Professor of Corporate Law at the University of Oxford, Faculty of Law.