European Safe Bonds (ESBies) - (3 of 5)
6. HOW WOULD THE ESBies BE ISSUED?
Beyond the overall structure of the ESBies, there are several implementation features that are worthy of some discussion. We will discuss three topics: the set-up of the EDA and size of the program, the maturities and different vintages of the ESBies, and the gradual introduction of the ESBies into the Euro-zone. Many readers may find what follows, on the nuts and bolts of how the EDA would operate, somewhat technical and may wish to jump to the next section.
6.1 Forming the EDA and Limiting the Amount of ESBies in Circulation
The EDA would play an important role in the European financial system. The amounts at stake would be large, and there are many savvy speculators willing to exploit any mistakes the EDA makes. Therefore, the agency would need a technically competent staff and an independent and knowledgeable board of directors. We do not anticipate that it would be hard to fulfill these requirements. For starters, almost every developed country has an agency in charge of issuing and managing its public debt, so there is some closely related experience on how to manage these institutions. Moreover, there already is a sizable European bureaucracy with decades of experience on how to keep its independence and a close cousin in the ECB. The EDA would require a much smaller staff than the ECB, making it easier to find the right people.
The governance of the EDA would have to be carefully set up to ensure it is immune from political pressure, be it from European or national institutions. One source of pressure would be to increase the share of a country “in need” in the portfolio of bonds held by the EDA. Because our rule for setting weights based on average GDP over the past 5 years is very transparent and rigid, it would be hard to do this under-handedly. Moreover, as we have already discussed, this would be written into the covenants of the ESBies bond contract, so that market participants would have the incentive to supervise the actions of the EDA and proceed to the courts if they were not consistent with keeping the ESBies safe.
A second source of temptation would be to increase the amount of ESBies issued, both as a trend over time and especially during recessions. One important feature that must be part of the debt covenants is a ceiling for the amount of ESBies that the EDA can issue. This maximum prevents the agency from falling into the temptation of issuing too many ESBies and becoming lax about ensuring their safety. Moreover, a hard upper bound would keep in check politicians that are tempted to pressure the EDA to finance persistent public-sector deficits. A hard limit of 60% of GDP, following the Maastricht criteria, may be a good starting point, although a more conservative approach may call for a smaller amount.
Another reason for an upper bound that is not too high, is that if countries place all of their bonds with the EDA, this would lead to complications on both sides of the market. On the side of the sovereigns, they would not be receiving market signals about the sustainability of their fiscal positions, for the EDA, not private investors, would be the marginal buyer of the securities. On the side of the EDA, the portfolio weight of that country in the assets of the EDA would have to be lower, requiring an increase in the weight of the countries that are similar to it in credit risk to preserve the relative safety of the ESBies. Put more succinctly, if there is not enough public debt, there is not enough raw materials with which to build ESBies. This seems to be a remote problem for European public finances in the near future at least.
Beyond the strict rule, there is an automatic mechanism that would put a brake on having too many ESBies. The larger the amount of ESBies issued, the more funds must be committed to the capital guarantee. Keeping with the pessimistic simulation of the last section, if the EDA bought sovereign bonds in the amount of 60% of euro-zone GDP, the capital requirement would require setting aside €800 billions. If instead, it was 20% of Euro-area GDP, it would take only €265 billions. As some countries will always be reticent at some time to contribute more funds to the EDA, requiring that there is a unanimous agreement to raise the upper limit of the EDA will help to prevent such increases.
On the other side, there is a lower bound on the amount of ESBies to issue to ensure that European banks are able to hold them and satisfy their capital requirements, and that the market for them is sufficiently liquid. We do not believe that this lower bound would be that high, since as long as the ESBies become a safe haven for investors during crisis, their liquidity is guaranteed.viii A third concern is that national governments or European institutions may pressure the EDA to use the funds in the capital guarantee to bail out the holders of the junior tranche whenever there is a sovereign default. Recall that the holders of this tranche are taking a risky position and being compensated for it, so it important that they do bear losses when defaults occur. There are a few ways to prevent these bail-outs. First, if the statutes of the EDA explicitly forbid the capital guarantee being used for any purpose beyond absorbing losses in the collateral pool of sovereign pool beyond the tranching threshold, there will be a significant legal hurdle for bailouts. Second, while banks and insurance companies would naturally hold ESBies, they should shy away from the junior tranche and its high risk. Therefore, they and their considerable lobbying power will be on the side of preventing bailouts of the junior tranche, which by eating into the capital guarantee of the ESBies would lower their safety and value. Third, if the sovereigns try to get around this constraint by posting additional capital to the EDA for the bailout, there should be a rule requiring unanimous agreement by all members and their respective parliaments. While the inability of European nations to agree on major changes on short notice has been a weakness throughout this crisis, it can be turned into a mechanism to uphold a commitment against bail-outs that prevents a future crisis.
A final consideration regarding liquidity and the size of the ESBies program is that the EDA would be a very large purchaser in the sovereign bond markets that can have large price impacts. Hence, when buying sovereign bonds it will drive up the price of these bonds to at least the ask price, and symmetrically, it will drive down the price of the junior tranche to the bid. There are a few ways in which the EDA could deal with this problem of buying high and selling low. It could occasionally buy sovereign bonds in the primary market, or it could refuse to temporarily include a bond in the ESBies if the bid-ask spread is not sufficiently low. While this may require some expertise and good sense on the part of the EDA, it should not pose any insurmountable obstacles.
6.2 Vintages and Maturity of the ESBies
Every so often, a new vintage of vintage of ESBies and junior tranches would be issued with new collateral in the form of recently issued sovereign bonds. This could happen as often as every two weeks, as in the case of US Treasuries, or as far apart as every 6 months. As long as the secondary market for ESBies is active and liquid, and as long as the EDA is able to sell the junior tranches, how often there are issues should not be too important. One feature that will arise with time is that each vintage of ESBies would be backed by slightly different collateral. In order to keep them equally safe, the tranching threshold may have to vary slightly across issues. There is nothing wrong with this, and if the ESBies of different vintages are correctly designed, then they would be very close to perfect substitutes even if not exactly identical. Issues may even be “re-opened”, e.g., more of the same bond could be issued one year after the bond’s issuance, to avoid off-the-run type phenomena. Reopening is already common for European sovereign bonds, and is done occasionally for US Treasuries.
As for the maturity of the ESBies, there are different clienteles for safe assets. Pension funds would like long-term ESBies, the ECB would prefer 3-month ESBies, and banks would like a variety in between. Our proposal is that at each issue, the EDA buys sovereign bonds of different maturities, and pools them, by maturity, to create ESBies for each desired maturity. This way, the EDA will not engage in any maturity transformation and so avoid funding liquidity (or rollover) risk. ESBies of all maturities will circulate, generating a risk-free yield curve that can be a useful indicator for policymakers and investors. One slight complication is that for some maturities, there may not be enough bonds available of a particular sovereign to satisfy the GDP-based weights in the collateral pool. In that case, the weights would have to be re-shuffled towards the available bonds of countries that have similar credit risk to the missing bonds. Yet, we suspect this will not be a common problem. Knowing that there is demand form the EDA for a specific amount of bonds with certain maturities, each country has a strong incentive to satisfy this demand, and so adjust the maturity of its primary issuances to ensure it. Moreover, some amount of coordination between the national debt issuers could be encouraged. Another worry for the EDA is that there will be a lag between the time it buys the sovereign bonds, and the time it is able to issue and place both the ESBies and the junior tranches. If there are wild swings in the bonds’ prices and credit risk or if it takes significantly longer than expected to find buyers for the ESBies and junior tranches, then the EDA will bear this cost of keeping the bonds in its warehouse. Insofar as the ESBies are widely held by European banks and investors, similarly to US Treasuries there should be a stable demand for them. If the EDA, moreover, secures the sell-off of the junior tranches before acquiring the sovereign bonds, then this warehousing risk will be limited. Moreover, national authorities should be encouraged to coordinate to a minimal extent on the dates of their issuances. Finally, as the EDA gains experience in the tranching step, it should be able to implement it quite quickly minimizing the lag between the purchase of the bonds and the sale of the ESBies and junior tranches.
Finally, to be clear, if the ESBies are accompanied by a capital guarantee, the latter will apply to all vintages and all maturities of ESBies outstanding. The capital guarantee can, inclusively, be used to make slightly different vintages of ESBies the same in terms of their safety and value. What if the extremely unlikely event occurs where a series of defaults in one European country after another uses up all of this capital? In this event, the member countries should have an option to recapitalize it. However, this should be approved by each national parliament for two reasons. First, so that it prevents illegitimate recapitalizations to bailout the junior tranche, as we discussed before. Second, because it provides a natural reset, whereby countries unhappy with the system can opt out.
6.3 The Transition to a World with ESBies
In our view, introducing ESBies and adapting bank regulation and ECB procedures to accommodate them could have an immediate positive effect on the Eurozone. Breaking the diabolic loop between bank fragility and sovereign debt would go a long way to prevent the risk of contagion that has made the European crisis so difficult tot deal with.
Yet, it may be more appropriate to introduce ESBies gradually. Even with a target of issuing 60% of euro-zone GDP, it could be reached via monthly issues over 5 years building up to that amount. This would give the market and investors time to get used to the new security and learn about its legal details. This may be particularly important for the junior tranche. Moreover, if the ESBies become a safe haven for investors, they can become very popular and effective even with a limited amount of them in circulation. Financial markets are expert in building derivate products on securities like the ESBies that can greatly extend their usefulness and reach.
Another transitional step is the joint choice of the tranching threshold and the size of the capital guarantee. At first, it may be hard to raise capital of even a few tens of billion euros for the ESBies given the economic crisis in Europe. In that case, the tranching threshold could be higher to keep the ESBies very safe, so that for a given amount of sovereign bonds purchased, the EDA would be able to issue fewer ESBies. Then, as capital is raised, the threshold would fall, allowing for more ESBies to be issued.
One issue that cannot be sidestepped is the recapitalization of European banks. The ESBies do not change the need for it, but rather make it even clearer and, hopefully, would therefore help to make it happen sooner rather than later. As regulators raise the risk weights on sovereign bonds, banks will have to trade these bonds for ESBies at market prices, and will inevitably lead to losses for banks the need to recapitalize. Insofar as the introduction of ESBies eliminates part of the uncertainty from the crisis, the prices of sovereign bonds may increase, limiting this loss. But, it is inevitable that some recapitalization must take place.
To be clear, ESBies by themselves will not solve the European sovereign crisis. They are part of a solution that includes at least two other pillars, dealing with the risk of banks and sovereigns, and which we describe in a separate note. Combined, they could put the financial markets of the euro-zone in a sustainable path for the long run. Moreover, the ESBies do not solve the problems of growth, lack of competitiveness and public finances that affect the general well being of countries like Greece, Italy, Portugal, or Spain. What the ESBies do is to prevent these structural problems from feeding into a massive euro-wide financial crisis that puts in danger the whole European Union.
viiiAs an example, Swiss bonds are often used as liquid safe assets because their return tends to rise during crisis, even though the amount in circulation is very small.