The unbearable lightness of regulation

We explain why the competing demands placed on financial regulation necessitate trade-offs and why these should be openly acknowledged rather than ignored or, increasingly, denied. They suggest three steps that regulators can take to help society and policy makers better understand and manage the tension between growth and stability. This article first appeared in Financial World Magazine and went free to read on the 10 September 20226.

Writing to the Financial Times in late 2024, John Glen – Economic Secretary and City Minister from 2018 to 2022, and so effectively responsible for financial regulation – said something that even ex-ministers rarely say out loud. The FCA, he observed, had become the convenient target for two contradictory demands made at the same time: that it should impose lighter and less bureaucratic burdens on growth, and that it should guarantee ever higher standards of consumer protection. His conclusion was blunt: “We cannot have both unless we agree the trade-offs.”

That sentence captures, in nine words, the essential problem – the inherent impossibility – of financial, and much other, regulation. It is also, we think, the most useful thing said on the subject in the past decade. Nobody acted on it.

The first (inherent) impossibility

By the time the Financial Services and Markets Act 2023 reached the statute book, the FCA had accumulated a formidable collection of things it was required to do: protect consumers, maintain market integrity, promote competition, facilitate competitiveness and growth, have regard to innovation, to proportionality, to the UK’s international standing etc, etc. The PRA’s list is shorter but has grown in the same way. Each item is defensible on its own. But, taken together, they form an optimisation problem with no solution.

Consider a stripped-down version, with two objectives and no ‘have regards’: protect consumers and promote competition. These sound complementary. In practice, they collide constantly. Detailed disclosure requirements help sophisticated consumers make informed choices and allow meaningful price comparison – but they confuse everyone else, while imposing costs that are passed on to all customers, including those who never read the disclosure. Whichever way the regulator moves, they are subordinating one defensible objective to another.

Add innovation, stability, proportionality and cost-benefit rigour, and the problem does not become harder in degree. It becomes different in kind. It is a spinning-plates problem: three plates can be kept going if you watch them closely enough but add a fourth and calamity beckons.

This is impossibility in the first sense: no regulator can maximise stability, growth, competition, consumer protection and international coordination simultaneously. No hierarchy, sequencing or set of weights dissolves it. It is permanent, and no reform programme will remove it.

But it is also navigable. Regulators navigate it every working day, and the mechanism they use is regulatory judgement rather than the statutory framework. The virtue of the arrangement that prevailed after 2008 – call it the peacetime constitution – was never that it resolved the tension. It was that it implicitly acknowledged the tension existed, accepted that the regulator was the one making the choice, and left that choice visible enough to be argued over.

Why the argument is never fair

However, visibility is not the same as fairness, and the peacetime settlement was never fair to regulators.

The costs of regulation are specific, immediate and attributable. A firm can say precisely what a rule costs, how many staff it requires and what activity it prevents. By contrast, the benefits are typically diffuse, long-term and unattributable. Regulators cannot point to the crisis that did not occur. A measure that quietly averts disaster looks, both before and after the non-event, like over-caution; a requirement that curbs excessive risk-taking looks like a drag on growth.

The result is an accountability asymmetry that no amount of good communication fixes. Regulators are continually criticised for the visible costs they impose and receive little credit for the invisible benefits they provide. The single moment at which crisis prevention becomes visible is the moment it fails.

The second (dishonest) impossibility

Something has changed in the last three years, and it is more than the pendulum swinging back.

The UK has written competitiveness and growth into statute as a secondary objective for both the PRA and the FCA – no longer merely something they must have regard to. The sleight of hand is easy to miss. Stability and consumer protection genuinely conflicted only in a narrow set of scenarios, typically where a large firm prospered by exploiting its customers. Competitiveness is different: you don’t have to be trying to maximise it for the conflict with the other objectives to kick in. Instead, it pulls against the other objectives more often than not. Meanwhile, the political rhetoric rarely acknowledges its secondary status, and in practice it has become the litmus test against which every regulatory action is measured.

The US has gone further, mounting an assault on the administrative state itself – executive orders, capital recalibrations, and a series of judicial decisions that strip agencies of the deference their technical expertise once attracted. The EU, unable to talk openly about ‘deregulation’ after the tribulations of its sovereign debt crisis, has adopted ‘simplification’ instead, a label promising tidier rules that increasingly delivers thinner ones.

The vocabularies differ. The premise is identical: that regulation – rather than weak demand, productivity or structural economic change – is the binding constraint on growth.

Denial has consequences

We have seen deregulatory phases before. What is new is not the direction of travel but the denial of what it involves. Earlier swings happened within a framework that acknowledged competition between objectives. This one instructs regulators to deliver growth and competitiveness alongside stability and resilience, while denying that any choice between them is being made.

That is not a harder version of the peacetime task. It is a different task. The peacetime regulator navigated trade-offs the system acknowledged. The wartime regulator is required to pretend they don’t exist.

And the accountability trap tightens accordingly. When risk accumulates, leverage rises and asset prices overshoot. Eventually, a shock will trigger an unwind, possibly a fire sale. At this point, the blame will not fall on the politicians who demanded lighter regulation, the industry that lobbied for each recalibration, or the courts that narrowed agency authority. Instead, the regulator’s defence – you told us to prioritise growth – will sound like an excuse.

This is not speculation. In 2019 the Federal Reserve tailored its enhanced prudential standards by size and risk, placing banks holding between US$100 billion and US$250 billion in assets into a new, more lightly regulated category. Silicon Valley Bank, previously treated as systemic, now sat in this new category, subject to considerably less scrutiny. Under the previous regime, it would have faced full standardised liquidity requirements, enhanced capital requirements, annual stress testing, and tailored resolution planning. Under the new one, it just faced stress-testing every two years. The change also shifted the burden of proof towards supervisors. It made it harder for them to act decisively, and their role shifted from requiring firms to fix issues, to recording that they had been raised.

When SVB failed in March 2023, the authorities invoked the systemic risk exception and protected every depositor – conceding, in the act of rescue, that the new category had been wrong all along. The post-mortem was nonetheless conducted almost entirely as an inquiry into supervisory failure. The policy decision that created the gap supervisors were then expected to fill had largely dropped out of the story.

This is not a new pattern. The pre-crisis FSA operated under an explicitly endorsed light-touch philosophy. After Northern Rock, light touch became a term of abuse, and the fact that ministers had demanded it, and that industry had celebrated it, likewise disappeared quietly from the story. None of this excuses poor supervision, but the usual telling of these events misses some important elements. These omissions make similar failures in the future more likely.

What can regulators do?

There is no solution, no avoiding a trade-off. A democracy is entitled to decide that it will accept a higher probability of crisis in exchange for faster growth. That is a legitimate political choice.

What is not legitimate is pretending the choice has not been made, then blaming the people instructed to prioritise growth for failing to deliver stability.

Three things follow – limited but not trivial – steps that regulators can take to improve the existing situation. Record trade-offs at the time they are made, rather than reconstructing them after the fact. Preserve institutional memory, so that when the pendulum swings back towards caution, policymakers are not starting from nothing. And work to reduce the amplitude of the swing: less extreme deregulation, less extreme crises, less extreme re-regulation. We aren’t aware of a regulator that currently does any of these in a systematic way.

Above all, and as John Glen called for in his letter, politicians and regulators should communicate honestly about the risks being created as well as the potential benefits. This will not prevent losses when the next crisis arrives. It will make it considerably harder for anyone to say afterwards that nobody could have seen this coming.