Regulatory change

The JC pontificates about technology
An occasional series.


Comments? Questions? Suggestions? Requests? Insults? We’d love to 📧 hear from you.
Sign up for our newsletter.

By way of falsification of Daniel Susskind’s thesis that technological unemployment is nigh,[1] consider regulatory change. It is a good example of unanticipated business change and increased complexity that is not solved by technology, but created by it.

Take trade reporting. Many jurisdictions now have trade and transaction reporting regimes requiring automated reporting in real-time to centralised trade data repositories on pain of significant financial penalty, the the millisecond, of all trades in the market and a range of second order metadata — 59 fields, to be precise — including unique transaction identifiers, the time, date and price of trade, the exact time of the decision to deal, the underlier, the counterparty, legal entity identifiers and so on. For every transaction across the market.

These regulations arrived in the aftermath of the global financial crisis but were spurred in part by advances in trading technology — especially high-frequency trading: Flash Boys and all that jazz. Trading firms harnessed quickly developing technology to create algorithmic high-frequency trading systems to capture the infinitesimal edge. The development continues apace.

The data supposedly improve market transparency, aid liquidity, provide early warnings of market dislocation and may indicate market abuse. None of this was available in the run up to the global financial crisis. Before about 2012, there was no transaction reporting to speak of.[2]. Like high-frequency trading, trade reporting is only possible because of technology. In “the good old days”, it just didn’t happen. This is not saving anyone any work; it is creating a whole new field of work. There are people whose entire career specialism is “regulatory change management”. I mean — shoot me, right? According to Google Trends the expression “regulatory change manager” was first searched for in October 2008, and then not again until December 2012.[3] It now appears 11,400 times on LinkedIn alone.

These new roles are not going to go away. The programme for implementing these trade reporting systems adds technology, personnel in change management, compliance, operations and technology (none of whom were hitherto required; because it is not as if there a programme of manual trade reporting: there was no transaction reporting, it creates equivalent roles within trade repositories and regulators, and for third-party businesses providing connectivity, reporting and application interfaces to facilitate this.

This whole new subsystem of interoperating parts needs maintenance and oversight; interpretation of new rules and changes to existing ones. It requires governance. It requires audit. It presents its own risks of failure, and no doubt adds to the risks of systemic failure elsewhere. This subsystem may be infinitely divisible into discrete algorithmic tasks, but the emergent whole - overseeing, understanding and managing this entire programme cannot be. There is no deep mind, no alpha-go playing machine, that can oversee manage and administer this programme unaided. Humans; lots of humans will always be needed. To be sure, humans may need even more technology to help them oversee that system — but each such application introduces more complexity, via exactly the same positive feedback loop, creates more risk, necessitates more governance and more audit, and will require more narratising specialists to diagnose and manage it.


See also

References