Predicting the Future of Regulation: Deregulation, AI, and Déjà Vu

Last week, David Collins posed the question is it too much regulation?

My own thoughts in the space are, maybe are we not regulating enough of the right things?

What I mean by that is – we should be regulating the individuals more, and therefore placing a greater burden on people and behaviours. Moving the Banking industry to a certified model on a far greater scale, than registered persons and SMR today. People should see working in any role in the industry as a privilege. If they don’t respect that, neglect it, or make short term decisions for personal gain, it should have consequences to them.

Instead, we have domestic and international sets of frameworks, that look like a patchwork rug, added to over 30 years. This was often driven by the fact that banks’ growth over the period was fuelled by acquisitions upon acquisitions, leading to tech stack on top of tech stack, held together by some brilliant hard-working souls who often took a “sticky plaster” approach to solving issues. Regulation has followed suit. It has targeted different problems, generally reactionary, and often without fully thinking through the objectives or at least, applying them broad brush.

What has this period of Regulation achieved?

With my background being Transaction Reporting, I often find myself in the debate of what does the regulator do with all that data. And whilst I can see the question, we can increasingly show evidence of the regulators trying to demonstrate this via reports, publications and references in market statements etc.

But the question is misplaced. Why do we have to send all this data? Because the industry itself allowed greed, complacency and short-termism behaviours to drive itself to the edge of the abyss.

The failure to stop and challenge the creation of insanely complex products that almost no one really understood pushed the industry to the point. You didn’t just poke the bear, you full on whacked it, danced in its face, and then asked it to help you.

Now – does sending 200+ data fields on every trade mean we are safer? I doubt it.

Does it mean people have a better understanding of what is being traded? Unlikely.

Has it changed the behaviours that lead to the regulation being put in place? I am not sure.

Has the immense burden of cost to implement the regulation in the space stifled innovation, both from a product but also tech space, impacting the end consumer? Yes. I have no doubt in my mind that in attempts to centralise, standardise, harmonise, the end consumer has suffered.

I think that Regulators can point to successes and stability over the period, navigating challenges. But, if they were to write the regulations for a post 2008 financial crisis work now, with everything they know from the last 15 years, I am not sure too many of them would copy and paste what we have today. And I am confident in saying, banks would have taken a different approach than the hugely costly investments they have had to make in the same period.

Is AI the answer or the problem?

Now, David in his post last week suggested we don’t need AI Regulation. I am not sure there, and think we do. Maybe that’s because I grew up in a Terminator generation and worry what happens when the machines rise up. – joking! (sort of).

But, I worry about the super powers AI can give bad actors. It is in areas like this, that I would advocate it is the behaviour, the “what can you use AI for”, that should really be regulated.

(Interesting side note – western culture tends to portray AI as “Tech vs Human”, whereas eastern is often more “Tech empowering Humans” and therefore tech or AI is embraced further).

I think used in the right way and by the right people, AI in regulation will have a significant shift in what can be done in this space and therefore, in firms’ compliance with the current regulator.

If I was to look into a crystal ball and predict what next?

Deregulation – maybe, but not as big or dramatic as suggested. I would suggest you will see more delays of Regulations such as the remaining parts of Basel 4 and MiFID III, with some softening of rules aimed at allowing growth in areas that do not pose a systemic risk in the near future.

Fines and enforcement – I think we will see a more pragmatic period, where the need to be competitive internationally results in Regulators inclined to address breaches in a manner that isn’t front page news.

A large institution having a significant loss due to an AI based tool in the next few years – almost certainly.

The last one, unfortunately, will then trigger a series of reactionary regulations and controls imposed on banks… Déjà vu.