unbenannter blogbeitrag 20260123 en
When good money disappears – what this CBDC experiment really shows
A critical look at an SNB working paper Central Bank Digital Currency and Gresham's Law: An experimental analysis
The Swiss National Bank is conducting research. In the working paper "Central Bank Digital Currency and Gresham's Law: An experimental analysis" (SNB Working Papers 3/2026), the authors examine an old monetary theory thesis with new technical packaging in a laboratory experiment: Gresham's Law – bad money drives out good.
The central message of the paper is quickly told: When a risk-free digital central bank currency (CBDC) exists alongside risky bank deposits, people tend to hoard the safe money and spend the riskier money. Limitations such as holding caps or negative interest rates are supposed to correct this behavior – with questionable side effects.
But how reliable are these conclusions? And what do they not say?
Summary of results
In the experiment, two forms of money face each other:
- Account A: risk-free (CBDC analogue)
- Account B: risky (bank deposit with default risk)
The main findings:
- Without restrictions, the risk-free money is held and also used.
- With holding caps, the safe money is almost exclusively hoarded, payments are made via the risky money.
- With negative interest rates, the safe money is hoarded less – but still more as a store of value than as a means of payment.
The authors conclude from this: In a system with predominantly risky bank deposits, it would be "better" to build the payment system on these risky deposits – not on CBDC.
That's a strong thesis. Perhaps too strong.
The critical questions that arise
1. Isn't a political goal being experimentally validated here?
The paper begins with an implicit premise: CBDC is dangerous for banks because it could withdraw deposits. The experiment then tests design options that are supposed to minimize exactly this risk.
Critical question:
Is the experiment open-ended – or is it seeking experimental legitimacy for already established monetary policy guardrails?
2. How realistic is the human model in the laboratory?
The test participants are students in a highly simplified economy:
- Two actors
- Fixed prices
- No alternatives like cash, crypto, stablecoins
- No institutional protection of bank deposits
Critical question:
Can one really infer the behavior of households and companies in a complex economy from such a setting?
Or differently: Do people in real life really behave like laboratory agents with Excel logic?
3. Is risk artificially exaggerated?
The risk of Account B is drastic: 10% probability per period, 50% loss.
Critical question:
Does this risk profile reflect real bank deposits – or does it deliberately create a dramatic contrast image to risk-free CBDC?
In reality, there exist:
- Deposit insurance
- Implicit government guarantees
- Bank bailouts
All of this is missing in the experiment – and systematically shifts the results.
4. Is "means of payment" confused with "sacrifice money"?
When people are forced to pay with riskier money while they hoard the safe money, is that then a functioning payment system?
Or rather:
A system in which citizens rationally try to avoid losses – at the cost of stability and trust?
The authors interpret the spending of risky deposits as confirmation of Gresham's Law. But one could also read it as a signal of distrust.
5. Is an intentionally unattractive CBDC democratically legitimate?
The paper discusses:
- Holding caps
- Negative interest rates
Both are coercive instruments to make an actually attractive public money artificially unattractive.
Critical question:
Why should the state create a means of payment that is intentionally worse than private alternatives?
And further:
Who decides how much "good money" citizens are allowed to own?
6. Is Gresham's Law being overstretched?
Gresham's Law emerged under metallic money regimes with state-fixed exchange rates. Whether it is transferable 1:1 to digital forms of money is disputed.
Critical question:
Is this a timeless economic law – or a convenient metaphor that legitimizes political decisions?
The uncomfortable conclusion
The paper shows one thing very clearly above all:
A well-designed CBDC would be too attractive.
And precisely because of that, according to the implicit logic, it must not be.
This raises a fundamental question that is not asked in the paper:
Does the monetary system primarily serve the stability of banks – or the users of money?
As long as this question is not openly discussed, experiments like this remain technically clean but politically incomplete.
Conclusion
The SNB working paper is methodologically interesting and cleanly executed. But its interpretation is normatively loaded. It shows less what CBDC necessarily causes, but rather what central banks are afraid of.
For a real societal debate about digital central bank money, more than laboratory experiments are needed:
- Real usage scenarios
- Alternatives beyond the banking system
- And above all: political honesty about goal conflicts
Because good money doesn't disappear by itself.
It is displaced – or intentionally kept small.