European governments are facing ballooning budget deficits in their quest to mitigate the economic effects of containment measures against the covid-19 pandemic. As some of these countries start this fight from a much weaker fiscal position than others, there’s a revived discussion about fiscal risk sharing among member states of the euro area. In the previous post we have discussed whether that risk sharing should only mean A) liquidity support whereby the community lends to troubled members on favourable rates; or B) also “solvency support” in the form of outright cross-country transfers. That discussion highlighted, that given the common nature of the current shock, positive externalities (spillovers), and the underlying reasons for differing initial fiscal positions, usual moral hazard concerns might be dominated by arguments for more solidarity, which could justify a more ambitious approach to European fiscal risk sharing, that is, cross-country transfers beyond liquidity support.

This post, in turn, tries to organize the various proposals out there into a consistent framework, which could help in highlighting the relative merits and essence of those proposals. There are two rather separate and independent questions:

1. Do European countries want to give liquidity support to each other, or are they also willing to provide “solvency support”, putting up with cross-country transfers?
2. How is that support financed? Via issuing more central bank money (a common liability of members tates), creating a new joint debt instrument (“Eurobond” or “Coronabond”), or via real time direct financial help from other member states (e.g. through a fund)?

Those options can be combined in all potential ways, leading to six main scenarios (not counting the status quo of bond market financing, and ruling out a default scenario inside the euro area):

The main difference between the liquidity and solvency scenarios is that in the former case any financial support originates a debt liability for the receiving member state (e.g. Italy) which remains on their books and will constitute a burden for their future taxpayers, potentially leading to debt-sustainability concerns. In contrast, the transfers involved in solvency support constitute income for the receiving country, i.e. they are “given” (like gifts) by the taxpayers of fellow member states, as opposed to being lent. Of course, the line between liquidity and solvency is not a very sharp one. Liquidity support also improves the solvency of a member state through reducing interest expenditures on outstanding debt, which can add up to a substantial amount in certain cases. In fact, the liquidity support role of central banks as a lender of last resort for governments is essential precisely because it can help ruling out self-fulfilling confidence crises in bond markets, which could otherwise force the country into default through bidding up its debt servicing costs. However, the scenarios listed in the “solvency support” category above potentially involve transfers which are larger by an order of magnitude, and their main distinctive feature is that they don’t result in a liability remaining on the books of the receiving country.

In what follows, we will walk through the above scenarios using an illustrative framework which is a very stylized “hybrid” of balance sheets (involving stocks) and income-expense statements (involving flows, i.e. change in stocks), so it is definitely not airtight from a technical accounting point of view. Nevertheless, it can perhaps better visually facilitate our understanding of what is going on. To ease the pain of readers with a more serious accounting background, I would suggest to view government spending not as an expense, but rather as an investment into an asset which is the well-being of the country – then it seems more palatable that they appear on the asset side of the balance sheet. Similarly, fiscal incomes such as tax revenues and transfers from abroad, contribute towards a higher “net worth of the government” which is why they are illustrated on the right side of the balance sheet, along with actual liabilities like bonds which also provide financing for government spending.

1. Liquidity via the ECB: QE, PEPP, OMT

Under this scenario (shown in Figure 1 by blue arrows), the ECB purchases the government bonds of member states with freshly created central bank money, potentially biasing purchases towards those countries which would face higher yields (e.g. Italy, which is the only country depicted in the figure below). The ECB is in effect lending to Italy to help finance Covid-19 related fiscal spending, but beyond liquidity provision it also helps via keeping borrowing costs low for Italy by alleviating any pressure from sovereign debt markets. Both the expanded Quantitative Easing (QE) program and the recently announced Pandemic Emergency Purchase Programme (PEPP) help in this respect, which are a nice demonstration of the ECB’s willingness to step up as lender of last resort for euro area governments.

The ECB’s capacity to buy sovereign debt is in principle unlimited. However, the current QE and PEPP programs both have a self-imposed upper limit which, although expandable if needed, might prove legally and politically challenging to increase. This issue can be addressed if the government in question applied for an ESM program, which would allow the ECB to activate its Outright Monetary Transactions (OMT) program. Under OMT, which was introduced in 2012 in the spirit of Mario Draghi’s “whatever it takes”, the central bank can buy unlimited amounts of government debt, in order to eliminate redenomination risk from bond yields, and to keep the currency union together. This can substantially increase the firepower of the ECB.

To the extent that Italian taxpayers are still liable for the purchased debt, and that they are expected to honor these bonds at maturity, QE does not involve cross-country transfers, since barring a future default event ECB shareholders (fellow member states) are not expected to make losses on them.

Figure 1: ECB QE and IMF-like ESM scenarios

2. Liquidity directly from member states: IMF-like ESM

In this scenario (shown in Figure 1 by grey arrows), the European Stability Mechanism (ESM) is lending to Italy (which is equivalent to buying Italian government bonds) against its capital which was pledged by member states. In this sense the ESM fulfils the role of the IMF for euro countries: it is a rainy day fund, in which member states stash away some cash, which then can be lent to members should the need arise. Italian taxpayers remain fully liable for paying back these loans. The question arises then, what the added value of ESM-loans is, once ECB bond purchases already ensure favourable financing conditions for Italy.

The capacity of the ESM is currently around 500 billion euros, which is rather limited. However, it can still improve financing conditions for member states as long as it is willing to provide loans below market rates, which are themselves already reduced by the ECB’s bond purchases (this subsidy would constitute a direct transfer, i.e. solvency support from ESM-members). More importantly, engaging with the ESM unlocks the potential for using the OMT program of the ECB, which can enable unlimited bond purchases by the central bank (as opposed to the more constrained QE, and PEPP programs). This would basically solve any remaining liquidity problems, and would also help a great deal via lower interest expenditures. In sum, it is the combination of an ESM program with the ECB’s OMT which can make a huge difference.

3. Liquidity via eurobonds: levered up ESM

Under this scenario (shown in Figure 2 by green arrows), the ESM can expand its lending capacity beyond the equity pledged by its members, by itself borrowing through issuing eurobonds. These eurobonds would constitute a joint liability of member states, proportional to their share in ESM equity, and can be sold to private investors as well as the central bank.

In Figure 2, for simplicity, it is only the ECB which buys these eurobonds, in effect swapping Italian debt for the common eurobonds. The ESM essentially becomes a middle man for the ECB’s asset purchase programs: the ECB buys eurobonds from the ESM, and the ESM, in turn, purchases Italian bonds from the proceeds. The ESM could become a “buyer of last resort” for Italian debt, while the ECB provides a monetary backstop to the ESM, guaranteeing that eurobonds are non-defaultable. Thus, at the end of the day, Italian spending is still financed by printing more money at the ECB, but inserting this extra step via the ESM and eurobonds helps in terms of democratic legitimacy. Deciding on the cross-country allocation of emergency bond purchases, with potential distributional implications in the form of future default risk, might be politically more acceptable in the hands of elected finance ministers (the board of ESM), rather than in those of independent central bankers.

Figure 2: levered up ESM scenario

So far we’re still only talking about liquidity support, which is essentially the same as ECB bond purchases, i.e. Italian taxpayers remaining liable for Italian bonds. Of course, if Italian public finances were to become unsustainable in the future, leading to default on their bonds, then the ESM would make losses on their holdings of Italian debt, while other member states would still have to honour their joint eurobond liability (obviously, in this case Italy could not chip in to pay their share from the eurobond obligations if even the ESM, which supposedly has seniority among creditors, could not have its claims from Italian bonds). This would constitute a transfer to Italy from other countries. However, this risk is there even in the alternative case without joint eurobonds as long as we assume that euro members cannot afford to let Italy default. Should Italy get to this point, a bailout of some sorts would have to occur anyway: like monetization via the Eurosystem’s balance sheet or an outright transfer from member states – rather than taking over Italy’s share of eurobond obligations – but a bailout would happen nonetheless.

Apart from making central bank bond purchases more politically and democratically acceptable, eurobonds would also have the extra advantage of improving the euro area’s financial infrastructure. It would create a euro safe asset for which there is great demand among investors, and would potentially constitute a much deeper and more liquid market than that for the current safety benchmark German Bund. Since eurobonds are backed by a diversified set of governments relative to a national bond, holders of eurobonds would not be exposed to one particular sovereign which could facilitate spreading the risk across the financial system. This would also help in arresting the so-called sovereign-bank doom loop in which distressed governments and their banks holding mainly their national bonds drag each other down in a vicious cycle.

4. Transfers via eurobonds: a Euro Treasury

Under this scenario (illustrated in Figure 3 with red dashed arrows), the ESM is transformed into something like a Euro Treasury which, instead of lending money to member states, engages in giving money to them in the form of transfers (solvency, as opposed to pure liquidity support). If these transfers are not immediately backed by own tax revenues or payments from member states, then the Treasury runs a deficit. This deficit can be financed by issuing eurobonds, which are joint liabilities of member states.

This scenario makes it clear that, in addition to improving liquidity and easing financing conditions (as in the levered-up ESM scenario), eurobonds can also be a vehicle for cross-country transfers. Under the Euro Treasury member states borrow jointly, and spend (not lend!) where needed. Depending on the exact institutional setup this can either mean a “transfer union” or a full-fledged “fiscal union” (or something in between). In a transfer union country leaders decide jointly only on the overall size and the recipients of the transfers, while also committing to make future payments to honour their share of eurobond obligations. The recipient government decides how exactly to spend the funds, while all member states decide on their own how to raise taxes at home to honour obligations to the Treasury. In a fiscal union, instead of giving the transfers to national governments, the common Treasury spends the funds directly where it sees fit, and could potentially also raise tax revenues on its own right – much like a federal government. The fiscal union setup could mitigate moral hazard and free rider concerns attached to joint debt since spending decisions are also taken jointly as opposed to individually. However, at their core both the transfer and fiscal union setups implement a type of fiscal risk sharing which could potentially entail transfers from the taxpayers of one country (e.g. Germany) to those of another (e.g. Italy).

Figure 3: Euro Treasury scenario

In Figure 3, for simplicity eurobonds are bought only by the ECB, but could just as well be bought by private investors in the bond market. The funds raised by issuing eurobonds are spent by the Treasury, or given to member states to spend, for their fight against the Covid-19 crisis. Notice that since this is a transfer to Italy, it does not contribute towards Italian public debt and Italian taxpayers are not directly liable to pay this back. However, they are indirectly liable for some part of this transfer through their payment obligations to the Euro Treasury. By making the transfers, the Treasury gives away assets (the funds raised from eurobonds) without acquiring any other asset in return, while its eurobond liabilities remain unchanged. This deficit creates a negative net worth (indicated by prolonging the right side of the Treasury balance sheet downwards: this is where this illustrative framework starts breaking down…), which constitutes a loss to member states in the sense that they are jointly liable for the Treasury’s eurobond obligations, and will be expected to make future payments to the Treasury such that it can make good on those promises.

Each member state should take part in recapitalizing the Treasury in proportion to their “capital share”, which is independent of their share from the disbursed transfers. The net transfer to Italy is therefore the difference between the gross transfer it received directly from the Treasury (the light green area), and the payments it needs to make for Treasury recapitalization (light grey area with red bars). The part of Italian Covid-19 spending which is financed by this net transfer is indicated with a green X. This area exactly corresponds to the one indicated by black X in the Treasury balance sheet, and by red X in Germany’s balance sheet. This also makes it clear that the net transfer received by Italian taxpayers came from German taxpayers as they need to pay in more capital to the Treasury (dark grey area with red bars) than the gross transfers they received from it (dark green area).

The scenario above describes a general asymmetric case where gross transfers and Treasury recapitalization shares are dissociated from each other, leading to net cross-country transfers, with some ending up as net beneficiaries, while others are net contributors – analogous to what happens within the budget of any welfare state. In the special knife edge case of completely symmetric gross covid-19 transfers, which perfectly correspond to member states’ capital share, we can see that no net transfers would occur. The reason why Italy might need asymmetric help could be either i) that it is hit by a more severe covid-19 epidemic; or ii) that it can afford from its own pockets the same amount of spending (necessitated by a symmetric covid-19 shock) to a lesser extent, due to “pre-existing conditions” such as higher initial public debt. The first case is fiscal risk sharing best described as insurance, while the second case is fiscal risk sharing which needs to involve a larger degree of solidarity. Whether that is subject to moral hazard concerns or can be deemed fair, was the subject of discussion in my previous post. A widely-cited proposal by Odendahl et al is something which partially falls under this scenario.

5. Transfers directly from member states: Corona crisis fund

This scenario (shown in Figure 4 with purple dashed arrows) is the combination of the IMF-like ESM and the Euro Treasury: just like the former, it does not issue joint debt, and similarly to the latter, it provides transfers as opposed to loans. This Corona fund can be thought of as an emergency rescue fund in which member states donate some money to be given to other members in need. These transfers are financed by direct payments from member states as opposed to issuing common eurobonds (similarly to the loans of the IMF-like ESM). Similarly to the Euro Treasury, the pattern of cross-country net transfers depends on the difference between the gross transfer a government receives and their initial contribution to the Corona fund – the only difference is that now they need to pay in the contribution directly and a priori, instead of financing it by creating a joint debt instrument.

Figure 4: Corona crisis fund scenario

While this Corona rescue fund can in theory achieve the same amount of redistribution across the taxpayers of different countries as the Euro Treasury setup with the eurobonds, in practice it is likely to have a much more limited firepower since all the funds need to be presented in advance. Absent eurobonds, the Corona fund cannot channel the extra liquidity which could be raised via joint borrowing, e.g. from the ECB. Of course, better-off members like Germany could borrow via their national bonds if needed to finance a larger equity pay-in to the Corona fund. But even in this case, the disadvantage of this scenario is that it strips the eurozone financial infrastructure of a much needed safe asset, like a eurobond, which can diversify exposures to individual sovereign risk (as discussed in the previous section).

6. Transfers via the ECB: monetization

Finally, cross-country transfers can also take place in the form of a “stealth bailout” through the balance sheet of the Eurosystem. Figure 5 illustrates this scenario with yellow arrows, while also combining all the previously discussed scenarios as well. If the ECB monetizes Italian debt, meaning that the bonds purchased during QE, PEPP or OMT are destroyed such that Italian taxpayers no longer need to keep servicing them, nor to ever pay them back, then this constitutes a transfer to Italy from the shareholders of the ECB. As in the case of the Euro Treasury, the net transfer to Italy depends on the difference between the monetized Italian debt (yellow area with red bars) and the corresponding fall in Italy’s equity at the ECB which Italy is supposed to recapitalize (light orange area with red bars). This net transfer from German taxpayers to Italian ones is exactly the area denoted by yellow X, which is equivalent to the loss of German equity at the ECB (which they also need to recapitalize according to Germany’s capital key) due to forgiving some of the central banks holdings of Italian debt.

Figure 5: All scenarios combined

Of course, the EU Treaty forbids such explicit monetization of public debt, but keeping it forever on the central bank balance sheet and rolling it over indefinitely (as well as rebating interest payments to the Banca d’Italia) would very much have the same effect on Italian taxpayers. In addition, to the extent that this strategy requires keeping the ECB balance sheet larger than might be justified from the aspect of future price stability considerations, then it imposes an “inflation tax” on all the citizens of the euro area just so that Italian taxpayers can keep rolling over their debt. It is in this sense that a “stealth transfer” can occur through the balance sheet of the Eurosystem.

On a related note, debt monetization is intimately connected to the issue of helicopter money. For simplicity, Figure 5 depicts a case where the ECB’s capital is completely wiped out by debt monetization, but does not go negative. However, equity could potentially become negative: in fact, the very essence of helicopter money proposals is a transfer to households against central bank equity. They point out that a central bank should not worry about negative net worth, as it can never go insolvent (since it prints the paper in which its liabilities are denominated), meaning that in our example member states would never need to recapitalize the ECB after it monetized Italian debt, and the whole amount of destroyed bonds would be gained by Italian taxpayers, but more importantly, without any de facto loss to German taxpayers. (Of course they would all lose in an accounting sense due to lost ECB equity shares, but they would never have to sacrifice actual resources to recapitalize it).

The above argument points out that the crucial way through which helicopter money is more stimulating than a simple combination of quantitative easing and fiscal stimulus, is by beating the Ricardian equivalence channel, i.e. by making the increase in the money supply permanent which never has to be “paid back” via increased future taxation (as opposed to financing it by repayable public debt). As I explain in more detail here, the problem with this argument is that it assumes the central bank potentially ceding control over the size of its balance sheet, and consequently over inflation. If at any point in the future, the central bank would like to lean against inflationary pressures by selling assets or increasing interest rates, a sufficiently negative equity (less assets than monetary liabilities) would mean that it might run out assets to sell or that it would need to keep increasing the monetary base to pay higher interest expenditures. In order to avoid this, the fiscal authorities would need to recapitalize the central bank, which would require raising additional tax revenue, invalidating the original argument about helicopter money being able to escape Ricardian equivalence: then we are essentially back at the QE + fiscal deficit combination.

In other words, if governments care about price stability in the future, fiscal backing of the central bank balance sheet is necessary, which also means that at the end of the day there is unlikely to be a free lunch from debt monetization on the aggregate euro area level. However, the ECB balance sheet can still be a vehicle for cross-country transfers between the shareholders of the Eurosystem.

* * *

This post aimed at organizing the various proposals for European fiscal risk sharing in a consistent framework along the dimensions of liquidity support vs transfers (solvency support) on the one hand, and different means of financing on the other hand. Of course, there are other valid organizational frameworks, like this detailed overview of fiscal options by Grégory Claeys.