Are SBBS really the safe asset the euro area is looking for?
The European Commission is pushing to create a synthetic euro-area-wide safe asset in the form of sovereign bond-backed securities (SBBS). However, SBBS do not fully fulfil their original promises. If introduced on a massive scale, they might increase the supply of safe assets in good times and loosen the link between sovereigns and banks. But they will not give governments a means to maintain market access during crises, they might change incentives for governments to default, and they could pose a problem to individual bonds not included in SBBS if, in the end, they are put at a regulatory advantage vis-à-vis individual bonds.
A safe asset is a liquid asset that credibly stores value at all times, in particular during adverse systemic crises (Caballero et al., 2017). There is high demand for this type of asset: from savers in need of a vehicle to store their wealth for the future; from domestic financial institutions to satisfy capital requirements, liquidity ratios and more generally to post collateral in many financial operations; and from abroad, from emerging market economies looking for a means to invest their foreign exchange reserves.
Several assets can play such a role: cash, central bank reserves, bank deposits, etc. Sovereign debt securities also play this role, in advanced countries in particular, thanks to their high liquidity and simplicity – as long as public finances are considered sound by the markets.
From the establishment of the monetary union through to the crisis, most of the bonds from euro-area countries enjoyed this status – but several of them lost it during the euro crisis, and spreads with German bonds increased rapidly. The loss of the safe-asset status resulted in two main problems: 1) these countries lost the possibility to put in place countercyclical fiscal policies at low cost during the recession, and 2) the fall in sovereign bond prices put an additional strain on European banks holding a significant amount of these bonds, having been drained already by the global financial crisis and the burst of housing bubbles in several countries.
Given the inability of some euro-area sovereign bonds to play the role of safe assets during the crisis, ideas on how to create synthetic euro-area-wide safe assets have emerged (see Leandro and Zettelmeyer, 2018, for a good overview of the main proposals). The SBBS proposal – originating from the European Safe Bonds (“ESBies”) proposed by Brunnermeier et al. (2011, 2017) – has recently gained some traction in the policy sphere. As a result, it has been developed in more detail by the European Systemic Risk Board High-Level Task Force on Safe Assets (ESRB, 2018a and 2018b) and is supported by the European Commission (2017a, 2017b and 2018), which sees it as a way to strengthen the euro architecture.
What are SBBS and why would we need them?
The main idea behind ESBies/SBBS is to pool sovereign bonds from all euro-area countries (in proportion to the ECB’s capital keys, i.e. more or less the size of each economy in the euro area) to use them as collateral and to provide cash flows for securities issued in several tranches, with different seniorities: senior ‘safe’ assets, and mezzanine/junior ‘risky’ assets. This has the additional (political) advantage over other proposals – such as Eurobonds or blue bonds/red bonds (Delpla and von Weizsäcker, 2010) – that it would create safe assets without resorting to any mutualisation of debt between countries.
Originally, the ESBies proposal had three main objectives:
1) To ensure a large supply of euro-area safe assets. As some countries of the eurozone lost their safe-asset status during the crisis, this reduced significantly the quantity of European safe assets in circulation. SBBS, thanks to pooling and tranching, could increase the supply again by offering a euro-area wide safe asset in the form of senior SBBS.
2) To help break the link between euro-area sovereigns and their domestic banks, which resulted in doom loops during the crisis. Introducing SBBS could help reduce the domestic bias in sovereign bond holdings of banks by providing them with an easy way to diversify their portfolios, thus making them more immune to the default risk of their sovereign. An additional advantage is that it would also make bank deposits (another type of safe asset available to ordinary savers) safer by making the banks, issuers of these deposits, safer.
3) To avoid intra-euro-area flights to safety during crises, which reduce the fiscal space of some countries at the worst moment. SBBS could help countries from the monetary union maintain market access in times of crisis in order to be able to put in place countercyclical fiscal policies, which are needed as euro-area countries cannot rely on their own independent monetary policy, exchange rate or on a federal budget to respond to asymmetric shocks.
Would SBSS really help achieve these three objectives?
In practice, it is possible that SBBS would help increase the stock of supply for safe assets and break the link between banks and sovereigns, but it is hard to imagine how they would provide a way for euro-area governments to maintain market access in times of crisis.
However, even for the first two objectives to be met, SBBS would need to be introduced on a massive scale. This would require significant regulatory changes so that SBBS could be treated as a risk-free asset, while the holding of individual sovereign bonds would be penalised (just aligning the regulatory treatment of SBBS with that of sovereign bonds as proposed by the Commission might not be sufficient to push investors to switch from simple sovereign bonds to more complex SBBS). A change in the ECB’s refinancing operations and haircut valuation to give a preferential treatment to SBBS would also be crucial to encourage financial institutions to use them as collateral. Such a drastic regulatory advantage for SBBS over individual bonds would give an incentive to the euro area’s banks to switch, at least partly, from the latter to the former. This would result in some diversification of their portfolios and thus reduce the probability of a doom loop. However, the extent to which this probability would be reduced would depend on the development of SBBS. This would need to be very high to have a significant impact on banks’ balance sheets.
The issuance of SBBS would be constrained by lower debt levels prevailing in some countries. Putting aside countries like Luxembourg or Estonia, which have very low debt-to-GDP ratios (and which would force SBBS to deviate from the official ECB capita keys very quickly), the maximum issuance of SBBS would be constrained at some point by German and Dutch levels of debt. This means that there would be a significant share of Italian, Spanish, Belgian, French, Portuguese, Austrian and other bonds from the EMU that would not be included in SBBS. As the regulatory treatment would have changed, putting them at a disadvantage against SBBS, funding for these bonds could actually become more problematic.
Turning to the third objective of SBBS – to offer a flow of “safe liabilities” (as put by Coeuré, 2017) for governments to rely on in bad times – this is clearly a false promise, for two reasons. First, the ESRB insists that sovereign bonds of countries losing market access will not be included in the pooling. Second, for other countries, finding buyers for junior SBBS in bad times would become crucial as this would limit the possibility of issuing any SBBS at all during stress periods.
Proponents of SBBS argue that there would be investors interested in buying both senior and junior tranches, given the low level of complexity and transparency of these assets. It is true that SBBS would be less complex and risky than some of the collateralised debt obligations (CDOs) produced before the global financial crisis. But this is not an assurance that the market in SBBS will function correctly in stress periods. These episodes are often characterised by flights to safety but also to flights to simplicity. As the financial panic of 2007 showed, runs on asset-backed securities (ABS) were indiscriminate, and depended neither on their complexity nor on their intrinsic performance (Covitz et al., 2009). This general panic about ABS made new issuances impossible even for simple products, while secondary markets froze completely. The same could happen to SBBS. In addition, the idea to exclude some countries from the pool (whether because they are losing market access or, on the contrary, because their debt is too low) would make matters worse by creating a variety of SBBS with different compositions, resulting in the lack of a homogenous market, lower liquidity, and less transparency. This is an issue because, in a financial crisis, even a bit of complexity can make these assets vulnerable. To play the role of safe assets, securities need to be information-insensitive – i.e. they need to be “simple in complex times”, as Caballero et al. (2017) put it, which would not be the case with SBBS. That’s why the ESRB concludes that “SBBS do not entail any built-in promise to offer a stable source of finance for governments during a crisis”.
In addition to not fulfilling one of its main objectives, there could be other unintended consequences of introducing SBBS. As highlighted by De Grauwe and Ji (2018) – but also by the recent increase in spreads in Spain and Portugal, following the announcement of the coalition agreement in Italy between The League and the Five Star Movement – correlations between sovereign bond prices of the euro area have had a tendency to change quickly in stress times. Coupled with the low diversification of SBBS –almost 80% of SBBS portfolios being invested in the four biggest countries of the euro area (Germany, France, Italy and Spain), thanks to the use of ECB capital keys weights – and with modified incentives related to the introduction of SBBS, the risk of default on European debts could end up increasing. That is why the use of historical data in SBBS simulations is generally hard to justify. Indeed, the incentives of governments to default could be completely altered by the introduction of SBBS and regulatory changes. For instance, if domestic banks only hold senior SBBS and, if the junior SBBS were only held abroad by non-systemic institutions, a default would not impact the domestic banking sector (which is one of the goals intended) and countries might have an incentive to default more often. Although the ESRB insists that SBBS are not “build to fail” or a “precursor to debt restructuring”, incentives to default might be affected by their introduction.
Overall, SBBS could lead to an increase in the stock of European safe assets, available in particular for banks during good times – which could appear to be a good thing, as it would help them diversify away from their domestic sovereigns. However, SBBS do not seem to provide a greater possibility for governments to finance themselves in crisis times, nor do they increase the flow of new safe assets to which investors could turn to in bad times (as SBBS could be vulnerable to panics due to their (even limited) complexity, markets could freeze and SBBS could quickly become illiquid assets).
Of course, even if other objectives are not fulfilled, the diversification of euro-area banks’ portfolios is desirable per se in order to share risk between countries, to avoid doom loop situations, and to be able complete the banking union from a political perspective (because some countries will be reluctant to accept a European Deposit Insurance Scheme as long as banks are biased in their exposures towards their own sovereigns). However, SBBS are not indispensable for this task; concentration charges, as proposed by Véron (2017), should be able to play that role by themselves. It is true that SBBS, in addition to diversification, also result in some de-risking for banks if they only hold senior tranches of SBBS, but the same would be true if they would hedge themselves against sovereign default by buying sovereign credit default swaps (CDS) for instance (as long as the counterparty risk of the CDS is low).
The proposal from the Commission to pursue SBBS could be seen as a way to convince countries reluctant to introduce concentration charges on their banks’ portfolios that they could rely on this new instrument to be able to sell their bonds to investors at all times. However, as discussed above, SBBS (at least in the version proposed by the ESRB) might not have this property.
The key to be considered a safe asset is, of course, sound fiscal and structural policy. A pool of euro-area bonds can contribute to diversify risks but faces numerous drawbacks. The SBBS idea remains untested and politically controversial. Policymakers and public debt management agencies in many euro-area countries are afraid of the unintended consequences that such a drastic change could entail for the sovereign bond markets of the eurozone.
Moreover, as argued in another blog post, the most compelling argument against SBBS is that they represent a distraction in the current debate. Instead, making all euro-area sovereign bonds safe again – and for good – remains the most desirable way to increase the supply of safe assets and avoid intra-euro-area flights to safety during bad times. This can be achieved through sound policies at the national level, combined with an ambitious reform of the eurozone architecture. This would render unnecessary the provision of safe assets in the euro area via potentially hazardous SBBSs.
Republishing and referencing
Bruegel considers itself a public good and takes no institutional standpoint. Anyone is free to republish and/or quote this post without prior consent. Please provide a full reference, clearly stating Bruegel and the relevant author as the source, and include a prominent hyperlink to the original post.