SAFE Policy Blog

The Risk of Government Debt on a Central Bank’s Balance Sheet Should not be Exaggerated

Target debate: Jan Krahnen responds to Clemens Fuest and Hans-Werner Sinn: We should stop extrapolating solvable problems into inevitable disasters.

The blog contribution by Clemens Fuest and Hans-Werner Sinn (henceforth Fuest/Sinn) is a helpful step towards a broader discussion of the pros and cons of today’s financial architecture in Europe – of which the common payment system and their (Target-) balances are an embedded element.

Fuest and Sinn point at a number of important risks in the current financial architecture that are indeed worth being considered, like the size of the capital buffer at the bank level, the methods of market risk measurement stipulated by supervisors today, or the incentives for risk-taking embedded in the current supervisory and resolution regime. Indeed, most of my own work in policy advisory circles focuses exactly on these issues: How to improve the existing regulation even further in order to render market discipline in banking sufficiently strong and credible.

The vast set of institutional rules and procedures forming the essence of the Banking Union were unfortunately completely absent from Fuest and Sinn’s original paper. Even if we may diverge as to the trust we put in the current state of these institutions, we nevertheless now all agree that we need to take the banking union´s regulations into consideration when discussing the potential default of a national central bank. That was my point, and I am grateful to Fuest and Sinn for conceding it.

I mention in passing that banking regulation and supervision has greatly developed, and indeed, deeply changed over the past five years (the “banking union”). The reference to the 2007/2008 crisis or the Greek crisis in 2012, as repeatedly done in Fuest and Sinn’s paper, is therefore not really appropriate – as it is out of time, neglecting the current status altogether.

Fuest and Sinn make an important second argument in their paper, when pointing at marketable securities as the other big balance sheet position at central banks, besides lending to financial institutions. Securities, notably government bonds, have become a major asset class on central bank balance sheets since unconventional monetary policy has begun. Of course, government bonds are not risk-free, in particular, if issued by a state with rather precarious fiscal status (the case of Italy comes to mind).

It is a widely shared view among economists, and Fuest, Sinn and myself are no exceptions, that government debt should be treated by the regulator as any other type of debt. Thus, it should carry risk weights that properly reflect the underlying default risk, requiring some Tier 1 equity backing if held on a bank’s balance sheet. However, OECD sovereign debt is treated as undistinguishably risk-free in the current Basel regulatory regime, and there is an apparent need for reform, an issue we all agree on – and regulators in Basel, Brussels, and Berlin are aware of this issue as well.

However, although Fuest and Sinn’s request for non-zero risk weights for government debt is widely shared among economists and policymakers, the market is unimpressed by the political debate – and distinguishes very well between low and high-risk sovereign issuers. Just look at the yield spread between Germany and Italy. That spread measures the difference between Italian and German government bond yields. It has been rising over the past few years from a low of 105 basis points (1.05 percent) in October 2015 to a high of 303 basis points in October 2018. It now stands at around 260.

A longing for doom

To come back to the Fuest and Sinn argument, the risk of government debt on a central bank’s balance sheet is real – but it does not easily grow without any limit. Again, as in the case of the banking union, the market tends to discipline issuers, i.e. governments, by raising the required coupons accordingly, if investors sense irresponsible fiscal policy. The example of Italy in April 2018, when Italian Deputy Prime Minister Matteo Salvini made his infamous announcements on the intended fiscal policies, may serve as an example. The dramatic surge in government debt yields – from 122 basis points to 290 basis points within days – probably taught the Italian government a lesson. In any case, the Prime Minister and his ministers have distanced themselves from a number of extreme policy options they had announced earlier, including the ostentatious break of the agreed budget ceiling.

This brings me to my final point: Fuest and Sinn’s paper, and also their response to my reply, carries a grain of truth – but a grain, even many grains, do not easily add up to a stumbling block. To be more precise, I am more than happy to discuss the criticisms of Fuest and Sinn one by one and to work jointly in identifying means and ways how to improve the current regulatory and monetary regime in Europe accordingly. Everybody should do this. However, I cannot follow in their pursuit of extrapolating isolated, solvable problems into inevitable disasters. There is a longing for doom in almost every argument in their text, which renders a sober discussion difficult.

We should strip off the extreme ballooning of low probability, end-of-the-world scenarios that characterize the argument by Fuest and Sinn, and replace it by more reasonable, high probability, manageable deficiencies of today’s institutional arrangements. Then we would have found common ground for a fertile debate on how to improve the current financial architecture in Europe, and how to make it more sustainable.

That said, there is indeed a good reason to prepare also for more extreme scenarios, and to devise policies for these events. However, in a fatal systemic crisis, as alluded to by Fuest and Sinn, or in a terminal insolvency or illiquidity event of a national central bank, other rules will dictate action than those underlying the debate on Target2 or the banking union more generally.

Jan Pieter Krahnen is Professor for Finance at Goethe University, Frankfurt, and SAFE Director.

See also the other articles in the target debate:



Want to get regular updates?

Subscribe to the Policy Blog RSS feed

Follow us on Twitter

Enter your email address to get regular blog updates.

* Please read our data protection statement (pdf version)