In the aftermath of the ECB's QE announcement one topic has received far less attention than it should: the unexpected collapse of risk-sharing across the Eurosystem as a precursor to QE. This is what prompted "gold-expert [17]" Willem Buiter of Citigroup to pen an analysis titled "The Euro Area: Monetary Union or System of Currency Boards", in which he answers two simple yet suddenly very critical for the Eurozone questions: which "currency boards", aka national central banks, are suddenly most at risk of going insolvent, and should the worst case scenario take place, and one or more NCBs go insolvent what happens then?
First, this is how Buiter calcualtes the loss absorption capacity of individual National Central Banks. In other words, the country whose central bank has the smallest OBLAC number is the one which go under first once the ECB looses control.
The most generous measure of the total conventional loss absorption capacity, or on-balance-sheet loss absorption capacity (OBLAC) of the Eurosystem NCBs is shown in the fourth column of Figure 7. It is the sum of Capital & Reserves, the Revaluation Accounts and Provisions. Note that the Revaluation Accounts are questionable as unconditionally loss absorbing resources. “Revaluation accounts includes unrealised gains related to price movements, foreign exchange rate movements and market valuation differences related to interest rate risk derivatives. Also includes the unrealised gains of euro area NCBs that have arisen due to the change from national accounting rules to harmonised accounting rules for the Eurosystem.” This suggests that the securities whose unrealized gains are recognized in the Revaluation Accounts entry could well be worth less than their Revaluation-inclusive value if they had to be sold in a hurry. A more prudent measure of conventional loss absorption capacity would therefore consist only of Capital and Reserves, plus Provisions, excluding the Revaluation Accounts.
[18]
Note that the ECB, and the NCBs of the Eurosystem live in a world of their own as regards accounting standards, statutory capital requirements etc. There is, as a matter of fact, no statutory capital requirement for the ECB. NCBs tend to have (national) statutory capital requirements. Neither the ECB, nor the NCBs follow the IFRS accounting standards or the IPSAS accounting standards.14 The Governing Council of the ECB has the statutory power to set harmonized accounting standards and principles for all euro area NCBs, making it easier for the ECB to determine whether an NCB is meeting its commitment to do all it can, within the constraints of the NCB’s total resources, to keep the ECB adequately capitalised.
The conventional loss absorption capacity figures are relatively modest for most NCBs, if we include Revaluation Accounts, and very modest if we exclude Revaluation Accounts. For Italy, for instance, the OBLAC was €85.7bn at end-2013. With Italy’s ECB capital key of 17.5%, Italian sovereign debt purchased under the €950bn PSPP would be €166bn. The total amount to be spent under the current QE programme by NCBs on euro area public sector securities for their own risk is €760bn. Italy’s capital-key weighed share of this would be €133bn. The ratio of Broad OBLAC to total NCB assets puts most NCBs into a leverage position that is more prudent than that of the commercial banks now supervised by the ECB. The ratio of Narrow OBLAC (capital and reserves) to total NCB assets puts all NCBs into a leverage position that is no better (and generally worse) than that of the commercial banks now supervised by the ECB. Yet, because individual NCBs have no control over their current and future seigniorage income, they are, as regards insolvency risk much more similar to commercial banks than to a normal central bank – one that has control over its seigniorage revenue.
The numbers for seigniorage and CLAC are rather less discouraging than those for OBLAC. Italy has a CLAC of €625bn compared to an OBLAC in 2013 of €86bn. Greece has an OBLAC of €10bn and a CLAC of €100bn. Spain boasts a CLAC of €407bn but an OBLAC of a mere €20bn.
* * *
We conclude that NCB insolvency, even under the generous ‘no negative CLAC’ definition of solvency, is a risk for a number of Eurozone NCBs. The risk is clearly greater if Revaluation Accounts are excludeed from the definition of unconditional (conventional) loss absorbing capacity. It is a material risk if downward revisions of future potential output growth in the euro area lead to a downward revision of the NPV of future seigniorage revenues (the estimates in Figure 7 are based on the assumption of 1% p.a. real GDP growth for all time in the future).
Here is Buiter on the question of "What happens if an NCB goes bust?"
As we noted above, an ordinary central bank that has only issued non-index-linked, nominal, own-currency-denominated liabilities cannot go bust in the sense that i) it cannot meet a payment due, ii) its comprehensive net worth is negative. However, even such a central bank can suffer losses that exceed its non-inflationary comprehensive loss absorption capacity (NILAC), even though they fall short of its unconstrained (by inflation) CLAC, which could be infinite in nominal terms. In such a situation, we should expect some combination of two related developments, unless the central bank is recapitalized: a domestic and external reduction in the purchasing power of the base money issued by the central bank (i.e. inflation and currency depreciation) and a growing reluctance by the counterparties of the central bank to transact with it, and specifically to accept greater exposure to non-interestbearing liabilities of that central bank or indeed to any fixed nominal interest rate liabilities. The counterparties may be official sector counterparties (e.g. other central banks) or private counterparties (domestic or foreign banks).
The situation is not all that different in the situation where a central bank has issued foreign-currency-denominated, real or index-linked liabilities and has suffered a large loss on its assets such that now even its unconstrained CLAC (and not just its OBLAC or NILAC) is negative: in such a situation, too, would we expect a rise in inflation, a depreciation of the currency and increasing difficulties for the central bank to find willing counterparties. In addition, it is conceivable that such a central bank may indeed default on its obligations or have its debt restructured.
It is clearly possible that, with limited P&L sharing, an NCB ends up with negative OBLAC and even with negative CLAC even if the consolidated Eurosystem shows positive OBLAC and CLAC. This risk increases greatly if the rules under which NCBs purchase sovereign debt for their own risk under the QE programme also require each NCB to purchase only its own sovereign debt (for its own risk). The NCB of a country with a sovereign that is at material risk of defaulting on its debt (Greece, Portugal and Italy, perhaps), could end up with significant holdings of high risk debt (up to 33 percent of the total marketable sovereign debt outstanding, and up to 25% of any individual bond issue).
The rules governing the PSPP operations of the Eurosystem do indeed require that each NCB only buys the debt issued in its own jurisdiction (by the sovereign or other entities) under the ‘own risk’ part of the QE asset purchases.
What happens to an NCB in the Eurosystem that has negative CLAC? There are four possibilities:
- The NCB may get recapitalized by its own sovereign.
- It may get bailed out by other entities.
- The debt issued the NCB gets restructured but it retains access to Target 2 for future funding.
- It may lose the ability to transact with the rest of the Eurosystem.
In a subsequent post we will look in detail at all of these four possibiltiies.
Buiter's summary:
- Individual NCBs do not control their future seigniorage revenues.
- Given limited risk sharing and absence of discretionary seignorage revenue, individual NCBs are more like commercial banks than like a normal central bank.
- Capital adequacy matters for individual NCBs as they could become insolvent even if the Eurosystem as a whole remains solvent.
- Recapitalisation by the NCB’s sovereign may not be feasible, particularly when the trigger for NCB insolvency is sovereign insolvency. Sovereign default remains likely in a handful of euro area member states.
- An insolvent NCB is not a credible counterparty for the private sector or for the rest of the Eurosystem (through Target 2).
- With diminishing risk sharing, the euro area (EA) is effectively becoming a system of currency boards.
- Either the country whose NCB is insolvent is effectively forced out of the Eurosystem, or there is ex-post profit and loss sharing.
- It is time for the euro area to recognise the minimum ex-ante fiscal burden sharing pre-requisites for an effective and long-term viable monetary union.
Speaking of Target 2, here is the just updated chart of the Greek intersystem liabilities.
[19]
And here is Buiter's conclusion:
Profit and loss sharing is essential for a viable monetary union. The Eurosystem’s steady drift away from profit and loss sharing adds to the list of existential risks faced by the euro area during the coming years.
As Winston Churchill once said of America, you can always rely on the EMU, like all of the EU, to do the right thing, once it has exhausted the alternatives. We are hopeful that the Eurosystem will revise what we see as its increasingly dysfunctional arrangements for profit and loss sharing before they cause real harm. If the Governing Council cannot restore something close to full profit and loss sharing for all financial transactions undertaken by the NCBs (if necessary by eliminating the ability of NCBs to engage in transactions and perform roles that are not part of the design and implementation of the common monetary and credit policy or of the Eurosystem’s lender-of-last-resort role), a revision of the Treaty that imposes full profit and loss sharing should be considered.
