Editors' note: This is the second column in a two-part series. Read the first part here.
In a previous column (D’Amico et al. 2021), we discussed a recent proposal regarding changes to the fiscal rules of the Stability and Growth Pact. In this column, we discuss the second element of our proposal: the plan to move a portion of national debts under the umbrella of a European Debt Management Agency (EDA).
In the last decade, there have been many proposals aimed at addressing a basic issue for the EU, namely, the lack of a large central fiscal entity issuing euro-denominated government debt.1 There are various dimensions to this problem, on both the finance and the macroeconomic side, but a common theme is that there is a large potential demand for government-issued, euro-denominated safe assets. The demand for supranational debt is strong even for securities that lack joint and several intergovernmental guarantees and/or a credit enhancement mechanism. Compare, for example, the bonds issued by the European Stability Mechanism (ESM) with those recently issued by the EU in the context of NextGenerationEU (NGEU). The ESM sits on a large amount of paid-in capital, as it could repurchase 70% of its bonds at nominal value. For this reason, it is not surprising that ESM bonds enjoy a AAA rating. The NGEU bonds issued by the EU also attract a AAA rating and trade at prices identical to ESM bonds, even though they are not protected by mutual guarantees or by a cash buffer. What secures the EU bonds in the investors' eyes is solely the EU capacity to extract payments from participating member states.2
This situation hints at an untapped potential for the EU to intermediate debt: moving a share of national tax streams and a share of national debts under the umbrella of a European entity, protected by the enforcement capacity of the EU, can increase the financing capacity of the whole area.
The two main benefits of the plan in our view are: (1) reducing debt costs for the whole Union, and thus increasing the safety of the existing stock of European debt; and (2) helping the operations of the ECB in debt markets.
In past years, the ECB has taken the main responsibility to ensure the stable functioning of European government debt markets. The creation of a European Debt Agency would complement the ECB's work on the fiscal side. The scheme would favour a gradual shift over time in the composition of the ECB’s assets away from a preponderant exposure to country risk towards a more standard configuration - characteristic of the ECB’s peer central banks – that by and large contemplates holdings of riskless bonds. These benefits could materialise in the near future, if monetary policy required a scaling down of the ECB bond purchase programmes, as it would allow the ECB to proceed in that direction without having to worry about destabilising national debt markets.
Let us briefly summarise the mechanism of the plan and then turn to some questions and criticisms.
Under our proposal, the EDA would gradually purchase a certain amount of national debt, at market prices and in proportion to each country's GDP, and finance the purchase with issuances of EDA debt. At the moment of purchase, the EDA would cancel the country’s bond and replace it with a commitment by the country to pay a flow of contributions to the EDA budget. The contributions would be calibrated to cover the net needs of the EDA associated with managing the debt of each given country, keeping the ratio of debt to country GDP constant, after an initial transition period. The formula to calculate the contributions is (r – g)d, where r is the interest rate on the European debt issued by the Agency, g is the growth rate of the country's GDP, and d is the EDA debt issued in proportion to the country’s GDP. The rate r is the same for all countries and would be chosen conservatively to allow the Agency to accumulate liquid reserves. The rate g would be equal to each country's potential growth (to avoid procyclicality of the contributions). The amount of debt acquired by the EDA for each country would be capped at a level corresponding to the debt increases during the pandemic shock. The contributions would be revised at regular intervals of five years by the EDA’s governing body. The same body would decide how to employ the EDA surplus – whether to save it in reserves, to rebate it to participating countries, or to direct it to joint EU projects.
The gains under the plan come from the expectation that debt issued under the EDA would trade at conditions close to those faced by the safest national debt in the area. This implies that the contributions and the use of the EDA surplus can be designed to yield Pareto gains for all countries involved.3 Where do these gains come from? There are basically three reasons for favourable credit conditions for EU debt. One is that EU debt will earn liquidity and safety premia relative to national debts, as it becomes the reference form of euro-denominated government debt. The second is that the EU will employ its enforcement capacity towards member states to assure investors of the reliability of the flow of future contributions. However, we are aware that there is a third, less desirable, channel, as favourable conditions may also come from an implicit perception of mutualisation, even though the scheme does not imply joint and several guarantees. We believe the scheme should be designed to rely on the two first channels only, while minimising the risk of ‘backdoor mutualization’. This can be done in three ways: by frontloading contributions, choosing a conservative r, and accumulating a liquidity buffer; by considering the use of dedicated sources of fiscal revenue instead of generic contributions; and, last but not least, by embedding the scheme in a solid fiscal governance framework. The last argument clearly points to an important element of complementarity between the two parts of our proposal.4
Would the plan create more fiscal space?
The plan would not cancel national debt, but it would replace it with an obligation to the EDA. The implication for the aggregate of EU countries is a reduction in the cost of debt, which does increase fiscal space for the Union. From the point of view of the fiscal rules, the national bonds eliminated by EDA acquisition would be removed from national debt and the new debt issued by EDA would be recorded as EU debt. This is analogous to what happens to the EU debt issued to finance the grant portion of NGEU. That debt is also eventually backed by countries' tax proceeds, but it is not recorded as national debt by Eurostat.
Would the proposal limit the ECB’s capacity to intervene on national debt markets?
No, we see the proposal as complementing and strengthening the ECB’s role. In past years, the ECB has turned to purchases of national debt to avoid segmentation of debt markets, to prevent the risk of multiple equilibria, and to pursue quantitative easing. All of these have been highly desirable developments, but it is useful to plan ahead for situations in which the needs of monetary policy may require a reduction in purchases. In such a situation, the objective of reduced segmentation in non-crisis times is better pursued by a plan based on the issuance of EU debt. The plan leaves more freedom to the ECB and helps remove country risk from national central banks' balance sheets.
More generally, we see a benefit in terms of clarifying the fiscal and monetary dimensions of EU risk sharing. Reichlin et al. (2021) recently argued in favour of moving beyond a status quo of ‘constructive ambiguity’ in terms of fiscal and monetary responsibilities inside the euro area. We see our proposal as going in that direction.
Would the contributions to the EDA be perceived as senior to servicing of national debt, and thus worsen the standing of national debt on the markets?
Payments to the EDA will most likely be considered de facto senior by financial markets. However, we do not think that on net this would negatively impact national debt markets. The fact that we did not see increases in high-debt countries' spreads on the dates when various details of NGEU debt issuance transpired suggests that, even though a portion of national revenues was now implicitly preferentially directed to servicing new European debt, this was not perceived as worsening these countries' capacity to service their own national debt. Our interpretation of these facts is that when the central fiscal capacity of the EU is strengthened, it tends to have positive repercussions on the member states' repayment capacity, exceeding the potential negative effect due to the seniority of EU claims on member states.
How does the plan compare to other approaches to creating safe assets?
The main difference with proposals based on pooling and tranching of existing national debt (Brunnermeier et al. 2012, 2017, Wendorff and Mahle 2015)5 is that the EDA would work more like a government than like a financial fund: its capacity to repay debt is based on its capacity of raising fiscal revenues and it rests explicitly on the EU enforcement capacity towards member states. The advantage of this approach is that it aims more directly at strengthening the fiscal capacity of the euro area. Incidentally, this design also means that the EDA would, in principle, require no capital contributions (paid-in or callable) as its fundamental asset would be given by the present value of future member transfers.
Isn't the national debt held by the ECB already essentially not a burden, as countries receive the interest they pay on that debt back from national central banks?
While this is true, it relies on the ECB continuing its asset purchases indefinitely. We believe there are substantial gains from the safety of knowing that there is a long-term solution to reduce national debt that does not rely on the ECB rolling over its asset purchases indefinitely. It is a reasonable concern that, in the short run, the contributions to EDA would increase the outlays of national governments. However, note, first, that the resources transferred to the EDA by member states could also be rebated back to participating countries; and second, in the current low-rate environment, the EDA could issue zero coupon debt at medium maturities and effectively have zero cash needs in the near future, so member contributions could be designed to essentially replicate the zero burden of ECB-held debt in the transition phase. Overall, the plan dominates an alternative in which the ECB remains the sole central actor with the capacity to acquire national debts, and its gradual implementation leaves flexibility in how the balance sheet of the ECB is adjusted over time.
Isn't it better to use issuance of EU debt to finance future centralised spending and make a stronger push towards a Fiscal Union?
We view our proposal as an additional step in the direction of a fiscal union, not as an alternative to it. Dealing with the burden of high legacy debt in some member states is a strategic issue that a successful fiscal union should tackle, not unlike other challenges that will require joint investment efforts in the future.
We conclude by going back to the theme we started from: the EU has shown strength during the pandemic, and that strength can be directed to improve future fiscal management. Better rules can improve the enforcement capacity of the Union and, in turn, stronger enforcement can be capitalised on to better manage inherited debt stocks.
Authors’ note: The debt management proposal grew out of conversations with Massimo Rostagno and we thank him deeply for that. The questions addressed in this column have been stimulated by useful comments and criticisms by Olivier Blanchard, Laurence Boone, Massimo Bordignon, Carlo Favero, Chiara Goretti, Silvia Merler, Gilles Moec, Jean Pisani-Ferry, Lucrezia Reichlin, Shahin Vallée, and Charles Wyplosz and we thank them all. The views expressed in this column are the responsibility of the authors and do not necessarily reflect the position of the French and Italian authorities.
Brunnermeier, M K, S Langfield, M Pagano, R Reis, S Van Nieuwerburgh, and D Vayanos (2017), “ESBies: Safety in the Tranches”, Economic Policy 32(90): 175-219.
Brunnermeier, M K, L Garicano, P R Lane, M Pagano, R Reis, T Santos, S Van Nieuwerburgh, and D Vayanos (2012), “European Safe Bonds (ESBies)”, Euronomics, p. 30.
Corsetti, C, L P Feld, P R Lane, L Reichlin, H Rey, D Vayanos and B Weder di Mauro (2015), A New Start for the Eurozone: Dealing with Debt, CEPR Press.
D'Amico, L, F Giavazzi, V Guerrieri, G Lorenzoni and C-H Weymuller (2021), “Revising the European fiscal framework, part 1: Rules”, VoxEU.org, 14 January.
Leandro, A and J Zettelmeyer (2019), “The Search for a Euro Area Safe Asset”, PIIE Working Paper 18-3.
Micossi, S and E Avgouleas (2021), “What to Do About Eurozone Sovereign Debt”, Project Syndicate.
Reichlin, L, K Adam, W J McKibbin, M McMahon, R Reis, G Ricco, and B Weder di Mauro (2021), The ECB Strategy: The 2021 Review and Its Future, CEPR Press.
Ubide, A (2015), “Stability Bonds for the Euro Area”, PIIE Policy Brief.
Wendorff, K and A Mahle (2015), “Staatsanleihen Neu Ausgestalten – Für Eine Stabilitätsorientierte Währungsunion”, Wirtschaftsdienst 95(9): 604–608.
Zettelmeyer, J (2017), “A new fiscal governance for the eurozone”, in T Beck and G Underhill (eds), Quo Vadis? Identity, policy and the future of the European Union, CEPR Press.
1 Leandro and Zettelmeyer (2019) offer an excellent overview of different options. Among past proposals some close to the proposal here include
2 The debt service protection scheme behind NGEU borrowing relies on (1) repayments from member states for the loan component, and (2) EU own resources for the grant component. For the latter, repayments will be covered by their gross national income (GNI)-based contributions and by new EU own resources. Should a member state not fulfill its financial obligations, the Commission could undertake several steps. First, the Commission would strive to meet the financing needs through active cash management and recourse to short-term borrowing. Should these means prove insufficient, the Commission could call more resources from other member states on a pro rata basis, while not exceeding their yearly share of 0.6 of GNI and, more importantly, without increasing their ultimately liabilities. The last resort measure would be to follow the standard process for infringement of EU law. None of these steps envisages an unlimited assumption of liability by the other countries.
3 Namely, it is plausible that the borrowing costs of the EDA would be slightly worse than those of some AAA-rated countries, like Germany, as now EU bonds do trade at a small spread relative to German bonds. The contributions of these countries can easily be adjusted to ensure that no country loses from participating to the scheme.
4 Some incentives could be built in the governance of the EDA, as continued participation to the EDA scheme could be made conditional on compliance with the fiscal rules.
5 Proposals closer to our approach here include Corsetti at al. (2015), Ubide (2015), Zettelmeyer (2017), Micossi and Avgouleas (2021).