Beyond ESBies: safety without tranching

Jeromin Zettelmeyer, Álvaro Leandro 01 June 2018

a

A

In a recent paper entitled “ESBies: Safety in the tranches”, Brunnermeier et al. (2017) proposed the creation of tranched securities backed by a diversified pool of euro area sovereign bonds. By choosing a sufficiently high subordination level (i.e. ‘thickness’ of the junior tranches), the senior tranche – called a European Senior Bond, or ESBie – could in principle be rendered as low-risk (in terms of expected loss rate) as a German government bond, without requiring either member state guarantees or a euro area budget. It could also be issued in sufficiently large volumes to replace national sovereign bonds in bank balance sheets, hence contributing to financial stability and reducing financial fragmentation in the euro area.

In spite of these qualities and the fact that many of the technical questions surrounding sovereign-bond backed securities (SBBS) have recently been addressed by a task force of the European Systemic Risk Board (ESRB 2018), ESBies continue to be heavily criticised (e.g. De Grauwe and Ji 2018). Critics focus mainly on three arguments.

  • First, in a crisis involving correlated sovereign risks and multiple defaults, the supposedly ‘safe’ senior tranche might end up being much less safe than its proponents claim.
  • Second, SBBS envisage the simultaneous issuance of senior and junior tranches, but would anyone want to buy the junior tranches, particularly in a crisis?
  • Third, ESBies might upset the functioning of national bond markets, raising sovereign borrowing costs.

Are the criticisms justified? Are there better alternatives to SBBS/ESBies? In a new CEPR Policy Insight (Leandro and Zettelmeyer 2018), we answer these questions by comparing SBBs/ESBies to three alternatives. Like ESBies, the alternatives all avoid joint and several guarantees, but unlike ESBies, they do not require financial engineering to achieve safety.

  1. Bonds issued by a capitalised intermediary (‘capitalisation approach’). The safe asset could consist of a single-tranche bond issued by a public intermediary backed by a diversified pool of euro area sovereign bonds. Rather than through tranching, it would be made ‘safe’ through a sufficiently thick capital cushion. The required size would depend on how the capital is invested: if invested in the same portfolio that backs debt issuance, it would be exactly the same as that of the junior tranches in the SBBS proposal – namely, about 30% of total assets. If it is kept in cash, capitalisation could be a little lighter, in the order of 25%. The need to provide this large public capital cushion can be interpreted as the cost of avoiding tranching.
  2. E-bonds. In an idea going back to Monti (2010), a senior public intermediary would issue a single bond backed by a diversified portfolio of euro area sovereign debt bought at face value. The funding costs of the bond would be passed on to the sovereigns in proportion of volumes of debt held in the portfolio of the public intermediary. As in the SBBS proposal, safety would hence be created through a combination of diversification and seniority, except that seniority now refers to the seniority of the issuer of E-bonds in the sovereign bond market, not to a senior tranche debt instrument.  In a variant of this proposal, the safety of E-bonds could be increased further by endowing the senior public intermediary with some capital. This would make it similar to the capitalisation approach, except that much less capital would be required.
  3. Bonds issued by a leveraged euro area sovereign wealth fund. As in the capitalisation approach, a publicly owned, capitalised institution would issue debt backed by a portfolio of assets. However, this portfolio would be internationally invested to maximise long-term returns subject to a desired risk level. This would allow the intermediary to start small, based on some ‘seed capital’, and gradually grow in size by reinvesting its earning and leveraging its capital subject to maintaining a prescribed capital adequacy. Once the desired size has been reached, the fund could start paying a dividend.

Safety

Using a default simulation model, all approaches can be calibrated to target a particular expected loss rate. In that sense, all can be made equally safe. However, they differ in how these losses are distributed, particularly in the risk tail. This can be illustrated using the ‘value at risk’ (VaR), a common measure for unexpected losses. The VaR at probability p measures the maximum loss occurring with probability p or higher. Figure 1 shows VaRs for ESBies, the capitalisation approach (assuming that capital is kept in cash) and E-Bonds (in two versions, including a ‘capitalised’ version).

Figure 1 Value at risk of different approaches

Sources: Alvaro and Zettelmeyer (2018), based on the simulation model of Brunnermeier et al 2017.

The figure shows that ESBies (and bonds issued by an equivalently capitalised intermediary) are more effective than E-bonds or national tranching in protecting their holders against a wide range of tail risks, but are more vulnerable to extreme tail risks. This is because ESBies are fully protected against individual or multiple defaults by euro area member states whose total losses-given-default do not exceed the size of the non-senior SBBS tranches. Once this cushion is depleted, however, holders of ESBies would bear the full cost of further defaults. In contrast, E-bonds and national tranching bear losses any time the losses-given-default of a single country exceed the portion of its debt held by subordinated debt holders. At the same time, they continue to offer partial or even complete protection (depending on loss-given-default assumptions) in the very unlikely event that a country such as France, Germany, or the Netherlands were to default in addition to lower rated countries.

To offer better protection against severe defaults by smaller countries or combinations of smaller countries, the E-bond intermediary could be capitalised. Figure 1 shows that a 2% capitalisation (at least ten times less than required in the pure capitalisation approach) would offer almost the same protection as ESBies for moderate tail risks but much higher protection against extreme risks.

Borrowing costs

To the extent that safe assets replace sovereign bonds on bank balance sheets, they should lower the risks of crises and hence reduce borrowing costs. Offsetting effects are possible – for example, by reducing bond market liquidity – but likely small, with one main exception. Since the E-bond proposal implies that investors would bear higher losses-given-default, the marginal cost of debt issuance for countries that have reached their debt issuance limit to the E-bond intermediary would go up. However, this does not necessarily translate into a rise in overall borrowing costs. The reason is that a share of the debt would be borrowed from the E-bond intermediary at the much lower cost, assuming that E-bonds are designed to match the expected loss rate of a German bund, and the E-bond issuer passes its low funding costs on to its borrowers. It can be shown that the net effect is to slightly raise average borrowing costs for the highest rated borrowers and to slightly lower average borrowing costs for the lowest rated borrowers.

Redistribution

Assuming that they operate as intended, the SBBS and capitalisation approaches rule out redistribution, as bond purchases occur at market prices. A euro area sovereign wealth fund will also avoid redistribution, provided that profits and losses are distributed in proportion with the capital shares.

In contrast, the E-bond proposal would lead to redistribution because the funding costs of the intermediary are distributed to its debtors according to their portfolio weights, regardless of how much risk each debtor contributes. However, because of the E-bond intermediary’s senior creditor status, the redistributive effect would be quantitatively small. Simulations indicate that total redistribution could be in the order of €10 billion over five years, mostly at the expense of Germany (-€3 billion) and France (-€2.2 billion) and to the benefit of Greece (€4 billion), Spain (€1.4 billion) and Portugal (€1 billion). Redistribution could be reduced by excluding exceptionally risky borrowers, such as Greece, from the portfolio, or by capitalising the intermediary in a way that reflects the contribution by each member state to the risks borne by the intermediary.

Unintended consequences

A frequent criticism of SBBS is that the junior tranches might not attract buyers in future debt cries, or perhaps even in normal times. It is easy to show that the latter is unwarranted, because SBBS would generally reduce, rather than increase, the net supply of lower-rated sovereign and sovereign-based securities (they ‘use up’ bonds in the same rating categories that would correspond to the junior tranches).  In debt crises, the junior tranches might indeed lose market access, but only if some of the countries in the SBBS portfolio also lose market access and/or sovereign issuers discriminate against SBBS intermediaries in a default situation. Furthermore, even if SBBS were to lose market access, the consequences would be benign, as countries could continue to issue plain sovereign debt directly to the markets.

Conclusion

Most criticisms directed at SBBS/ESBies appear exaggerated. SBBS do well compared to several alternative proposals to create safe assets that would not require tranching. They would protect their holders against a wide range of risks including correlated defaults, would not lead to redistribution across countries, and would not require public capital.

At the same time, some competing proposals could be superior to ESBies in at least some dimensions. In particular, a lightly capitalised version of Monti’s (2010) ‘E-bond’ approach would have two advantages. First, it would offer the same protection as ESBies against moderate tail risks, but higher protection in extreme cases. Second, it would have a fiscally disciplining effect on sovereigns, by raising marginal borrowing costs in countries with high debt levels, without however raising their average borrowing costs. Unlike SBBS, however, capitalised E-bonds would imply some (if modest) redistribution, have a much larger impact on national bond markets than ESBies, and require some public capital.

References

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.

De Grauwe, P and Y Ji (2018), “How safe is a safe asset? “, CEPS Policy Insight No. 2018-08. February.

ESRB (2018), Sovereign bond-backed securities: A feasibility study, European Systemic Risk Board High-Level Task Force on Safe Assets.

Leandro, A and J Zettelmeyer (2018), “Safety Without Tranches: Creating a ‘real’ safe asset for the euro area”, CEPR Policy Insight No. 93. May 2018

Monti, M (2010), A New Strategy for the Single Market, Report to the President of the European Commission, May.

a

A

Topics:  EU policies Europe's nations and regions Financial markets Financial regulation and banking International finance Monetary policy

Tags:  ESBies, tranching, sovereign bond-backed securities, E-bonds, capitalisation

Senior Fellow, Peterson Institute for International Economics; CEPR Research Fellow

Research analyst at the Peterson Institute for International Economics

Events

CEPR Policy Research