The Supervisor’s Fallacy: Stability Without Growth Is Not Success
Poland’s financial supervisor is right that markets need rules. But the harder question is whether supervision protects the economy — or merely protects the supervisor.
A senior official of the Polish Financial Supervision Authority recently published a defence of financial supervision. The argument is elegant and, at first sight, hard to reject.
Financial supervision, we are told, is like a seat belt. It does not make the car faster, but it reduces the consequences of an accident. Regulation may increase costs, slow down implementation and reduce risk appetite, but in the long run it is supposed to protect households, firms and the financial system from crises.
There is much truth in this argument. That is exactly why it deserves a serious answer.
The real debate is not between regulation and no regulation. That is a false opposition. No serious economist believes that modern financial markets can function without rules, disclosure, prudential constraints, anti-fraud enforcement and investor protection.
The real question is different:
What kind of supervision, imposed on whom, at what cost, in what time frame — and with what effect on competition, innovation and access to capital?
That question is missing from most institutional defences of supervision. And it is the question that matters most for economic policy.
The supervisor’s welfare function is not the economy’s welfare function
Supervisors are not neutral machines. They are institutions with incentives.
A financial supervisor naturally dislikes failure. It dislikes scandals, political criticism, legal liability and situations in which it can be accused of having allowed too much risk. The safest institutional response is always to require more: more documents, more capital, more reporting, more discretion, more enforcement powers and longer procedures.
From the supervisor’s perspective, this is rational. From the economy’s perspective, it may not be. A market can be stable because it is resilient. It can also be stable because it is stagnant. These are not the same thing.
This is the core of what I would call the supervisor’s fallacy: the belief that supervisory safety is equivalent to economic welfare.
It is not.
The welfare function of the economy is broader. It includes stability, but also competition, access to finance, lower transaction costs, innovation, productivity, investment, employment, tax revenues and the ability of domestic firms to scale. A policy that minimises institutional risk for the supervisor may simultaneously increase costs for entrepreneurs, reduce choice for households, protect incumbents and push innovative firms abroad.
This is why the institutional location of financial-market policy matters. If financial-market policy is effectively led by the supervisory authority, it will tend to reflect the incentives of supervision: control, risk avoidance and institutional defensiveness. If it is led by a ministry of finance or an economic ministry, the calculus should be broader: capital formation, competitiveness, innovation, tax base and the international position of the domestic market.
That is not an argument against supervision. It is an argument against confusing supervision with economic strategy.
Crises are costly. That does not prove every restriction is efficient
The strongest part of the supervisory argument is the reminder that financial crises are expensive. This is true. The IMF’s work by Laeven and Valencia on systemic banking crises documents how costly such crises can be in fiscal and output terms.
But this evidence proves less than supervisors often suggest.
From the proposition “financial crises are costly” it does not follow that every additional supervisory constraint improves social welfare. It does not follow that every reporting obligation, every delayed licence, every conservative interpretation, every additional administrative power and every higher compliance cost is justified.
That would be like saying that because road accidents are costly, every speed limit, roadblock and movement permit is efficient.
The relevant question is marginal: What is the additional benefit of a given supervisory measure relative to its additional cost?
That is the question missing from the institutional defence of supervision.
The evidence does not say: more supervision is always better
The literature is more nuanced than the regulatory narrative.
In one of the classic cross-country studies, Barth, Caprio and Levine do not provide a simple case for “more official supervision”. Their findings point instead to the importance of accurate information disclosure, private-sector monitoring and incentives for private agents to exert corporate control. In other words, the design of supervision matters more than the size of the supervisory apparatus.
The same applies to competition. Claessens and Laeven found that banking systems with greater foreign bank entry and fewer entry and activity restrictions tend to be more competitive. This is not a case for deregulating everything. It is a warning against assuming that official restrictions automatically produce better markets.
Regulation also has distributional effects. A large bank can absorb a new compliance requirement by hiring another team of lawyers, risk officers and reporting specialists. A small financial institution, fintech or start-up faces a different reality: the same rule may become a fixed cost that prevents entry, delays scale or forces relocation.
Even the European Banking Authority’s own work on supervisory reporting costs treats compliance costs as a problem that must be measured, simplified and made more proportionate, especially for small and non-complex institutions.
The problem is not theoretical. The Federal Reserve Bank of St. Louis found that compliance costs represented 8.7% of non-interest expenses for banks with less than $100 million in assets, compared with 2.9% for banks with assets between $1 billion and $10 billion.
That is how regulation becomes a concentration mechanism.
It can protect consumers in one dimension while reducing choice, competition and local access to finance in another.
Good supervision can reduce uncertainty instead of increasing it
There is also another model.
A good supervisor does not merely police the market. It can reduce uncertainty, lower information asymmetries and help serious firms operate legally.
The evidence on regulatory sandboxes is relevant here. A BIS working paper by Cornelli, Doerr, Gambacorta and Merrouche found that firms entering the UK regulatory sandbox saw a 15% increase in capital raised after entry, while their probability of raising capital increased by 50%. The mechanism was not magic. The sandbox reduced regulatory uncertainty and information asymmetry.
This is the difference between a supervisory state that says “prove everything before you start” and one that says “test under controlled conditions, disclose risks, protect clients and learn fast”.
For innovative sectors, this difference is decisive.
Crypto is not the whole story. It is the stress test
Crypto-assets are a useful example not because crypto deserves special treatment, but because it reveals institutional incentives faster than traditional finance.
The European Union adopted MiCA — Regulation (EU) 2023/1114 — to create a harmonised framework for crypto-assets. MiCA is not a libertarian framework. It imposes rules on crypto-asset service providers, issuers, market conduct, governance and disclosure. It is regulation.
But it is also a single-market framework.
That distinction matters. Under MiCA, a firm authorised in one Member State can provide services across the EU through passporting. This means that national implementation is no longer just a domestic legal issue. It becomes industrial policy.
In a passporting market, time-to-licence is industrial policy.
If one Member State implements MiCA in a fast, predictable and proportionate way, it attracts compliance teams, legal work, technology jobs, tax base and institutional learning. If another Member State implements MiCA slowly, adds national gold-plating, creates broad administrative powers or leaves firms without a domestic licensing path, serious firms do not disappear.
They move. They license elsewhere and passport back.
This is exactly the risk Poland now faces.
In my new report for the Warsaw Enterprise Institute, (Over)regulation of Crypto-Assets in Poland, I analyse Polish crypto-asset legislation against EU law, constitutional standards and comparative MiCA implementation across Member States. The full report deserves a separate article. Here, one point is enough: as of April 2026, Poland remained without an operational MiCA implementation, while roughly 1,200 VASP-registered entities faced a hard deadline before 1 July 2026.
The Polish supervisor’s own statement makes the practical problem clear. According to UKNF’s position on MiCA, if no competent authority is designated in Poland after 1 July 2026, domestic entities will lose the ability to provide crypto-asset services until they obtain authorisation. At the same time, crypto-asset service providers authorised in other EU Member States may continue to provide services in Poland under MiCA passporting rules.
This is not consumer protection. This is jurisdictional self-harm.
The result would be paradoxical: Polish firms could be forced out of the Polish licensing path, while foreign-licensed entities could continue serving Polish clients. The market would not become safer because activity disappears from Poland’s regulatory perimeter. It would become less domestic.
Poland would lose firms, jobs, tax revenue, compliance know-how and supervisory learning. Consumers would still use crypto services — but increasingly through entities supervised elsewhere.
The cost is paid outside the supervisor’s balance sheet
This is the part usually missing from institutional arguments.
The cost of excessive supervision is not paid mainly by the supervisor. It is paid elsewhere.
Households pay through higher costs, fewer products and slower access to financial innovation. Firms pay through weaker access to capital and fewer domestic funding channels. Investors pay through limited choice and lower market depth. Small financial institutions pay through fixed compliance costs that favour incumbents. Innovative sectors pay through relocation, legal uncertainty and delayed scaling.
And the state pays through lost tax base, lost expertise and weaker strategic positioning.
These costs are harder to see than a scandal. They do not produce one dramatic headline. They accumulate quietly.
A market that never develops is rarely described as a policy failure. But it should be.
What better supervision would look like
The alternative is not weak supervision. The alternative is better supervision.
A developmental model of financial supervision would still punish fraud. It would still protect client assets. It would still require governance, disclosure and capital where appropriate. But it would also recognise that the financial system is part of the productive economy.
Such a model would require five things:
Proportionality — requirements calibrated to the scale, complexity and risk of the business model.
Predictable licensing — clear criteria, measurable deadlines and no indefinite administrative limbo.
Controlled experimentation — sandboxes, pilots and structured dialogue for innovative models.
Judicial control over intrusive powers — especially blocking websites, accounts or business activity.
Enforcement against fraud, not overregulation of legal firms — because pushing compliant firms abroad does not eliminate risk.
This is not an anti-supervisory agenda. It is a pro-market and pro-state agenda.
A state that understands markets does not abdicate regulation. It regulates in a way that keeps activity, knowledge and responsibility within its jurisdiction.
Stability is not enough
The Polish debate about financial supervision often starts with the wrong question: does supervision slow down the market? The better answer is: sometimes it should.
A supervisor should slow down fraud, excessive leverage, mis-selling, hidden conflicts of interest and systemic risk. But it should not slow down lawful entry, responsible innovation, competition or the formation of domestic financial expertise.
That is the distinction.
Safety without access is not success.
Stability without competition is not success.
Compliance without growth is not success.
The purpose of financial supervision should not be to protect the supervisor from criticism. The purpose should be to protect the market while allowing it to develop.
Poland’s crypto case is only one example. But it reveals a broader institutional problem: when financial-market policy is absorbed by the logic of supervision, the country may end up with fewer scandals, fewer firms, fewer innovations and fewer opportunities.
That may look safe from inside the supervisory office.
It does not look like a development strategy.
Supervision should protect the market from failure. It should not protect the supervisor from making choices.