Make sure the ESM’s pandemic crisis support is fit for purpose!
Maturity, seniority and size: Make sure the ESM’s pandemic crisis support is fit for purpose!
Most of the discussion regarding access to European Stability Mechanism (ESM) financing to fight the Covid-19 crisis has focused on the importance of removing conditionality from ESM loans. Memories of harsh austerity and fears of an electoral backlash made it imperative that a detailed memorandum of understanding (MoU) was not part of ESM loans. Removing conditionality would not only eliminate the stigma that asking for help could create (already an important concern), it would also prevent the potential damage from ill-timed fiscal measures during a reconstruction process that is quite uncertain in all but one feature – it will be long and difficult. On 9 April, EU finance ministers agreed that the ESM can provide loans with the sole requirement to commit to use the money to pay for direct and indirect healthcare, cures, and crisis-related costs. This new dedicated facility will be called the Pandemic Crisis Support Credit Line (PCSCL).
However, while we know the PCSCL will not include thorough conditionality in the form of a MoU, there is much less clarity regarding what type of financing terms it will offer. According to officially disclosed information, the PCSCL will be based on the existing Enhanced Conditions Credit Line (ECCL). This implies that there are additional characteristics of the loans that are known:
Maximum size: The finance ministers have agreed that the ESM can grant PCSCLs amounting to 2% of each member’s GDP.
Interest rate on the loan: The standard pricing of an ECCL, once it is activated, is 35 basis points (plus fees) over funding costs. To this, commitment fees for making the credit line available should be added.1
Type of access: ECCL credit lines have an availability of one year (renewable) and can be drawn via a standard loan or a primary market purchase.2
What is missing?
The following are financial aspects that would shape the impact of the PCSCL, but for which we still have no detail:
Funding strategy: In order to understand what the cost of the PCSCL will be, we need to know not only what margins and fees the ESM will charge, but also how the ESM is going to finance the loans.
What issuance strategy by the ESM should back PCSCL support? While ongoing debates have brought proposals for the issuance of perpetual bonds, as we argue in Erce et al. (2020), the most effective way to provide official support is to obtain financing using instruments that are as short as possible. In this way, the ESM can leverage the currently ultra-low short-term interest rates and perform a strong maturity transformation.
Loan maturities: According to the existing guidelines, there is no written limit to the maturity of the ECCL support. We know, however, that the amended ESM treaty, the discussion of which was frozen by the Covid crisis, wanted to impose a maximum average maturity of five years. In a recent interview, Klaus Regling argued that average maturities for the PCSCL could be as long as 10 years. In Erce et al. (2020), we show quantitatively the extent to which maturities longer than 10 years would help smooth refinancing needs in the medium term.
Seniority: The typical loan from the ESM is senior (its repayment has priority in default) to every creditor but the IMF, although ESM seniority can be waived (as in the 2012 Spanish programme). Will the ESM remain senior under the PCSC? Or will it, as the ECB, lend in a pari-passu fashion (i.e. be treated equally as the rest of creditors)?3
The risk to avoid: “Too senior not to disrupt”
During the euro area crisis, the seniority status of official lenders became an especially contentious issue. The senior role of the ESM replaced the de jure pari-passu approach initially pursued. The rationale for claiming seniority is that official support (both through liquidity and conditionality) increases a county’s ability to repay, to the benefit of all existing creditors. In exchange for ‘enlarging the size of the cake’, the ESM claims seniority (Corsetti et al. 2020).
The seniority of ESM loans can have undesired effects on sovereign (junior) bond markets. According to Ghezzi (2012), the euro area experience proved that private sector subordination can have perverse effects when markets are not certain about the success of the programme. If the sovereign borrower does not recover following official lending, the loss given default increases as the official loan dilutes private creditors. This can undermine market access.
On the ‘positive’ side, the small loan size proposed makes dilution less important. How important dilution is would depend on the repayment structure of the loan. A bullet repayment approach would imply an increase of refinancing needs of 2% of GDP the year the ESM loan would mature.
What are the interest savings from the PCSCL?
As long as the ECB prevents belief-driven speculative behaviour in the euro area bond market (i.e. Europe remains in a good equilibrium), the interest savings are not great.
Italy’s three-year yield currently hovers at around 1%. This is in contrast to the ESM loan, which would add 40 basis points (fees included) to the borrowing costs. If the average maturity of the ESM bonds is three years, the ESM loan would cost 30 basis points. In this scenario of stress, actual savings would then amount to (0.01-0.003)*27 billion= €150 million of direct interest savings a year.4
This benefit seems too small to add senior creditors and dilution risk into the sovereign bond market.
Seniority can have consequences: Evidence from 2011
Despite the relevance of seniority in sovereign debt markets, there is very little systematic evidence on seniority in practice (Schlegl et al. 2015). Here, we provide evidence using the recent experience of the euro area, where the official sector seniority was effectively diluted by its longer maturities (Hatchondo et al. 2016).
In December 2010, Ireland received support for almost €65 billion, of which €40.2 billion came from the EFSF/EFSM, the prequels to the ESM.5 Originally, the loan carried a margin of 250 basis points over funding costs and an average maturity of 7.5 years. In turn, Portugal entered its programme in May 2011. The size of its package was €78 billion, of which €52 billion came from the EFS. Initially, Portugal’s loan featured a 210 basis-point margin and an average maturity of 7.5 years. In July 2011, the Eurogroup granted both countries a reduction of the loan margins (to 0 basis points) and an extension of the average maturities to 15 years.
This experience provides an opportunity to analyse whether and how longer maturities can mitigate the negative effect of seniority (Chatterjee and Eyigungor 2015). We study how the bond market responded to the change in the repayment profiles of official lending, such that senior debt repayments became more backloaded. This change substantially reduced the concern of bondholders that, at maturity, they would have to stand in line with senior creditors able to dilute them.
To obtain a first insight, we compare the change in both the sovereign yield curve and bond market liquidity (as measured by bid-ask spreads) around the time in which official lending, originally with a repayment structure falling in a window five to ten years ahead (prior to the loan amendment), was “extended” to be repaid beyond ten years. We do this in Figures 1 and 2.
Figure 1 Irish and Portuguese sovereign yield curves around the 2011 extension of maturities and reduction of margins
Source: Corsetti et al. (forthcoming), from Bloomberg and author’s calculations. Notes: The right-hand axis contains the difference (in basis points) between average yields (at each maturity) one month after the amendment. The left-hand axis measures yields in percentage points.
Figure 1 describes the dramatic yield compression and flattening of the yield curve that followed the extension of maturities and suppression of margins. In Figure 2, we present the change in the value of the average bid-ask spread of the corresponding sovereign bonds, in the month before and after the amendments to the lending terms. While in Ireland the liquidity improvement was sizeable, this positive effect did not play out in Portugal.
Figure 2 Bond market liquidity
Source: Corsetti et al. (forthcoming).
There is an important caveat to keep in mind when looking at this evidence. Many concurrent factors, first and foremost ECB actions, could contribute to explaining the dynamics in Figures 1 and 2.
However, we also use a regression-based event-analysis framework (Gourinchas and Obstfeld 2012), which allows us to understand how the difference between the 10-year and 3-year bond yields (term spread) deviated from a fair-value model (defined by a rich set of variables, including ECB policy measures) during the weeks around (i) the signing of the ESM loan and (ii) the 2011 modification of the loan. We build the term spread using daily data from 2009 to 2016 for the corresponding benchmark 3-year and 10-year maturity sovereign bonds. Figure 3 presents the results.
Figure 3 Econometric evidence
Panel A in Figure 3 shows the results for the Irish case. In the week ahead of signing its EFSF loan, the excess term premium hovered around 0. In the week following the announcement, as yields went up, the term spread went 5 basis points negative. An inversion of the curve, at such high-yield levels, is an indicator that markets are increasingly concerned about default. In contrast, while ahead of the July 2011 change in lending terms the excess term premium remained deeply negative (-20 basis points), following the amendments yields fell dramatically (Figure 1) and the negative term spread disappeared. This bullish flattening of the curve indicated that default concerns receded as official loan repayments were postponed.
The results for Portugal, presented in Panel B of Figure 3, are consistent with those for Ireland. In the run-up to the signing of the ESM loan in March 2011 there was a very short-lived improvement in the term spread, but this soon turned back to negative, indicating markets were not convinced the programme solved the problem. Instead, when maturities were extended, the drop in yields and fattening of the yield curve (Figure 1) coincided with the disappearance of a negative excess term spread.
Policy implications: Dilute official loan seniority, de jure or by maturity
A small PCSCL, together with standard margins, would deliver only a modest amount of savings. Moreover, if the maturity of the loans is not long, the loans would not deliver significant smoothing of financing needs and would embed senior creditors and dilution risk in sovereign bond markets.
Experience shows seniority can be destabilising. De-jure seniority does not eliminate credit risk for the official sector, but it can tilt the balance of risk against private creditors and disturb bond markets (Schlegl et al. 2015). If waiving seniority is not possible, the use of longer maturities can smooth such effects by placing repayment senior creditors later in time.
Why should the ESM restrict the maturity of its lending through the PCSCL to a window below ten years? In fact, the longer the loans the ESM provides, the easier it will be for borrowers to recover swiftly, and the more room for manoeuvre it will have to design an adequate financing strategy.
These considerations imply that to enhance the effectiveness of the PCSCL, access should be provided both without seniority and by back‐loading loan repayment. These features would to avoid dilution risk, making the PCSCL more attractive for potential borrowers.
Chatterjee, S and B Eyigungor (2015), “A Seniority Arrangement for Sovereign Debt”, The American Economic Review 105(12): 3740–3765.
Corsetti, G, A Erce and T Uy (forthcoming), “Official Sector Lending in the Euro Area”, Review of International Organizations.
2 The detailed guideline for the ECCL is available here.
3 See Gelpern (2016). A complication is that the preferred creditor status of the ESM is anchored in the ESM Treaty. Spain benefitted from a waiver that was explicitly foreseen as a possibility only during the transition from the EFSF.
4 Direct savings may be larger if uncertainty causes yields to rise. Indirect savings may be substantial if PSCSL triggers OMT.
5 The overall programme financing was above €80 billion but, as Daragh Clancy likes to remind us, almost €20 billion of that amount were Irish savings.