QE - "European style": be bolder, but parsimonious!
The ECB will purchase a monthly €60bn of private and public debt instruments between March 2015 and September 2016 – a total worth over €1 trillion. While the timing and size of purchases are known, there is more leeway than it seems in the way purchases are allocated to each category of assets.
By Urszula Szczerbowicz, Natacha Valla
Post
We argue that instead of focusing on sovereign debt, the Eurosystem should privilege assets that are the closest to job-creating, growth-enhancing and innovation-promoting activities. In particular, instruments issued by “Agencies and European institutions” should be given a prominent role. But they should also be selected very parsimoniously so as to make sure the money created is used to finance long term growth and jobs and not unsustainable government expenditure – as might unfortunately end up being the case under the current list of eligible “Agencies”, which might not please European public opinions once they find out about it. When selecting its eligible agencies, the ECB might be well advised to select them not only on credit worthiness grounds but also according to their economic purpose.
QE as it stands: centered around sovereign purchases
Sovereign bonds are a natural target for the ECB’s Public Sector Purchase Programme (PSPP), the latest – and by far the biggest – of the Quantitative Easing (QE) policy conducted in the euro area. The pool of outstanding sovereign debt instruments is indeed the deepest and the most liquid, especially for highly rated bonds. But shortcomings of sovereign purchases by the Eurosystem have been subject to criticism even before their implementation. First, some have said “too little, too late”. Second, the geographic allocation of purchases, set to follow the share of each EMU member state in the ECB’s capital (the“capital key”), mechanically allocates almost half of all purchases to German and French bonds. Those markets already benefit from exceptionally low interest rates, and net secondary market supply is expected to fall short of planned purchases (both because current bond holders might stick to their portfolios, and because primary market issuance, in particular in Germany, will be small over the foreseen horizon of PSPP). Third, their market impact (flattening of yield curves) and macroeconomic effects (hiking inflation, spurring credit growth) are expected to be limited. Fourth, PSPP might exacerbate tensions in the interbank market by extending the pool of outstanding sovereign debt trading at negative rates. That could in turn accessorily drain the already scarce collateral.
Twisting the programme to diversify the funding structure of the euro area economy
What if the Eurosystem flipped its QE strategy around and selected assets based on their final economic use rather than on their issuer or on the asset class they belong to? In fact, a reasonable operational target for the ECB’s QE could be to facilitate the financing of growth-enhancing private sector activities; to stimulate investment; and to underpin European investment projects or joint initiatives, such as – but not only – the Juncker Plan.
The first way to implement a non-sovereign QE is to focus on private assets. These can be either issued by the financial sector, or by non-financial corporations themselves. In fact, purchases of financial sector securities have been the ECB’s first asset purchase programmes. Way back in 2009, they started going quantitative by conducting outright purchases of covered bonds. They then followed suit in November 2014 with an ABS programme.
But this exclusive focus on securities issued by banks turned problematic. By doing so, the ECB did put all its eggs in the same basket, fully conditioning the transmission of its monetary policy on bank balance sheets. Looking forward, at a time when banks are engaged in a decade long process of balance sheet consolidation, the scope to kick start bank credit supply will remain constrained for a while. But also, evidence suggests that there are many benefits in diversifying the sources of financing for firms so that they can find funding in all cyclical circumstances, in particular in early stages of upturns. Grjebine et al. (2014) show that economies with high share of bonds in corporate debt and high degree of substitutability of bonds for bank loans tend to perform better after economic downturns, i.e. they recover faster and more steeply (see Figure 1). This seems like a reasonable objective for the ECB.
Financial instruments issued directly by the corporate sector would therefore be a premium candidate for QE. By purchasing them, the ECB would foster the development of those markets and most surely improve and speed-up the nascent recovery of the euro area economy. En passant, that would also be a plus to move the euro area towards a Capital Markets Union (CMU), as foreseen by the European Commission.
What size of corporate bond purchases could the ECB aim for? Corporate bonds are already accepted as collateral in the ECB open market operations (OMOs). In fact, according to the most recent published numbers, the ECB identifies an outstanding amount of €1 393 billion of such assets eligible to its OMOs, of which only a mere €74 billion have actually been pledged! This leaves a substantial outstanding stock that could be bought outright by the ECB without running the risk of creating a collateral squeeze. In the recent past, corporate bonds were bought by the Bank of England and the Bank of Japan to facilitate corporate refinancing and investment. The Fed also purchased unsecured corporate commercial paper and asset-backed commercial paper. So that would not be such an exotic endeavor after all.
But we should keep in mind that by buying corporate bonds, the ECB would directly interfere in the development of bond markets in Europe – again a matter to be looked at, and perhaps levered, within the CMU initiative. This can be constructive if done carefully. First, the development of European bond markets must go along with appropriate regulation. Second, the impact of ECB massive interventions on market liquidity should also be anticipated and fine tuned. The BIS and others have recently pointed to bond market liquidity as a cause of concern. This issue should be considered seriously, bearing in mind that the ECB purchases might also stimulate bond issuance by firms that would otherwise not be active in those markets.
Purchases of agencies and international European institutions: be massive, but parsimonious!
A more direct financing of the real economy could take yet another form. In fact, the ECB’s QE programme already explicitly includes securities of European institutions that finance European investment projects, such as the European Investment Bank (EIB). Not only would these purchases support the euro area deficient demand in a way that is geared towards long-term growth objectives, but they would also deepen European integration. At the current juncture, this is no luxury. The total available euro-denominated pool of bonds issued by the EIB is of around €200bn. As the ECB programme currently stands, the purchases of such securities are subject to loss sharing and limited to 12% of the program. It seems that substantially increasing this limit beyond 12% would make QE more efficient (see Figure 2).
In its PSPP, the ECB distinguishes between “Agencies” and “European institutions”. This has gone unnoticed, but might in fact be key. While the ECB purchases of European institutions securities are constrained by 12% ceiling, the purchases from “Agencies” are not subject to the same limit. When publishing the nitty-gritties of its programme, the ECB even indicated that National Central Banks could choose between sovereign or agencies, and that if they were not in a position to execute the planned, monthly purchases of sovereign or agencies debt instruments (for example because there would be none to sell in the market), they would have the possibility to buy “European institutions” instead. This means that the ECB would be in a position to massively support public investment simply by buying bonds from National Promotional Banks such as the German Kreditanstalt für Wiederaufbau (KFW) or Spain’s Instituto de Credito Oficial (ICO), or from the EIB, or even instruments issued under the heading of the Juncker Plan, instead of their sovereign counterparts.
Yet, while we fully support ECB’s purchases of agencies (and in fact we believe that they should be the backbone of QE), parsimony in the choice of agencies will be critical. Indeed, the list of eligible agencies as published by the ECB covers a range of very different animals. For instance, the list includes CADES (a body created to bear the debt of the French social security system) and UNEDIC (body in charge of France’s unemployment insurance, which is largely in deficit): these bodies have little to do with investment and long term growth, and it is not clear that European public opinions will be that pleased when they find that out. When selecting its eligible agencies, the ECB might be well advised to select them not only on the credit worthiness grounds but also according to their economic purpose.
Reference:
Grjebine, T., Szczerbowicz, U., and Tripier, F. (2014). Corporate Debt Structure and Economic Recoveries, CEPII Working Paper, Nr 2014-19 November.
QE as it stands: centered around sovereign purchases
Sovereign bonds are a natural target for the ECB’s Public Sector Purchase Programme (PSPP), the latest – and by far the biggest – of the Quantitative Easing (QE) policy conducted in the euro area. The pool of outstanding sovereign debt instruments is indeed the deepest and the most liquid, especially for highly rated bonds. But shortcomings of sovereign purchases by the Eurosystem have been subject to criticism even before their implementation. First, some have said “too little, too late”. Second, the geographic allocation of purchases, set to follow the share of each EMU member state in the ECB’s capital (the“capital key”), mechanically allocates almost half of all purchases to German and French bonds. Those markets already benefit from exceptionally low interest rates, and net secondary market supply is expected to fall short of planned purchases (both because current bond holders might stick to their portfolios, and because primary market issuance, in particular in Germany, will be small over the foreseen horizon of PSPP). Third, their market impact (flattening of yield curves) and macroeconomic effects (hiking inflation, spurring credit growth) are expected to be limited. Fourth, PSPP might exacerbate tensions in the interbank market by extending the pool of outstanding sovereign debt trading at negative rates. That could in turn accessorily drain the already scarce collateral.
Twisting the programme to diversify the funding structure of the euro area economy
What if the Eurosystem flipped its QE strategy around and selected assets based on their final economic use rather than on their issuer or on the asset class they belong to? In fact, a reasonable operational target for the ECB’s QE could be to facilitate the financing of growth-enhancing private sector activities; to stimulate investment; and to underpin European investment projects or joint initiatives, such as – but not only – the Juncker Plan.
The first way to implement a non-sovereign QE is to focus on private assets. These can be either issued by the financial sector, or by non-financial corporations themselves. In fact, purchases of financial sector securities have been the ECB’s first asset purchase programmes. Way back in 2009, they started going quantitative by conducting outright purchases of covered bonds. They then followed suit in November 2014 with an ABS programme.
But this exclusive focus on securities issued by banks turned problematic. By doing so, the ECB did put all its eggs in the same basket, fully conditioning the transmission of its monetary policy on bank balance sheets. Looking forward, at a time when banks are engaged in a decade long process of balance sheet consolidation, the scope to kick start bank credit supply will remain constrained for a while. But also, evidence suggests that there are many benefits in diversifying the sources of financing for firms so that they can find funding in all cyclical circumstances, in particular in early stages of upturns. Grjebine et al. (2014) show that economies with high share of bonds in corporate debt and high degree of substitutability of bonds for bank loans tend to perform better after economic downturns, i.e. they recover faster and more steeply (see Figure 1). This seems like a reasonable objective for the ECB.
|
Figure 1: Economies with high share of bonds in corporate debt perform better in recoveries |
 |
|
Source: CEPII, Grjebine et al. (2014).
|
Financial instruments issued directly by the corporate sector would therefore be a premium candidate for QE. By purchasing them, the ECB would foster the development of those markets and most surely improve and speed-up the nascent recovery of the euro area economy. En passant, that would also be a plus to move the euro area towards a Capital Markets Union (CMU), as foreseen by the European Commission.
What size of corporate bond purchases could the ECB aim for? Corporate bonds are already accepted as collateral in the ECB open market operations (OMOs). In fact, according to the most recent published numbers, the ECB identifies an outstanding amount of €1 393 billion of such assets eligible to its OMOs, of which only a mere €74 billion have actually been pledged! This leaves a substantial outstanding stock that could be bought outright by the ECB without running the risk of creating a collateral squeeze. In the recent past, corporate bonds were bought by the Bank of England and the Bank of Japan to facilitate corporate refinancing and investment. The Fed also purchased unsecured corporate commercial paper and asset-backed commercial paper. So that would not be such an exotic endeavor after all.
But we should keep in mind that by buying corporate bonds, the ECB would directly interfere in the development of bond markets in Europe – again a matter to be looked at, and perhaps levered, within the CMU initiative. This can be constructive if done carefully. First, the development of European bond markets must go along with appropriate regulation. Second, the impact of ECB massive interventions on market liquidity should also be anticipated and fine tuned. The BIS and others have recently pointed to bond market liquidity as a cause of concern. This issue should be considered seriously, bearing in mind that the ECB purchases might also stimulate bond issuance by firms that would otherwise not be active in those markets.
Purchases of agencies and international European institutions: be massive, but parsimonious!
A more direct financing of the real economy could take yet another form. In fact, the ECB’s QE programme already explicitly includes securities of European institutions that finance European investment projects, such as the European Investment Bank (EIB). Not only would these purchases support the euro area deficient demand in a way that is geared towards long-term growth objectives, but they would also deepen European integration. At the current juncture, this is no luxury. The total available euro-denominated pool of bonds issued by the EIB is of around €200bn. As the ECB programme currently stands, the purchases of such securities are subject to loss sharing and limited to 12% of the program. It seems that substantially increasing this limit beyond 12% would make QE more efficient (see Figure 2).
|
Figure 2: The universe of purchasable Agencies and European Institutions should not be underestimated |
 |
|
Source: RBS, ECB.
|
In its PSPP, the ECB distinguishes between “Agencies” and “European institutions”. This has gone unnoticed, but might in fact be key. While the ECB purchases of European institutions securities are constrained by 12% ceiling, the purchases from “Agencies” are not subject to the same limit. When publishing the nitty-gritties of its programme, the ECB even indicated that National Central Banks could choose between sovereign or agencies, and that if they were not in a position to execute the planned, monthly purchases of sovereign or agencies debt instruments (for example because there would be none to sell in the market), they would have the possibility to buy “European institutions” instead. This means that the ECB would be in a position to massively support public investment simply by buying bonds from National Promotional Banks such as the German Kreditanstalt für Wiederaufbau (KFW) or Spain’s Instituto de Credito Oficial (ICO), or from the EIB, or even instruments issued under the heading of the Juncker Plan, instead of their sovereign counterparts.
Yet, while we fully support ECB’s purchases of agencies (and in fact we believe that they should be the backbone of QE), parsimony in the choice of agencies will be critical. Indeed, the list of eligible agencies as published by the ECB covers a range of very different animals. For instance, the list includes CADES (a body created to bear the debt of the French social security system) and UNEDIC (body in charge of France’s unemployment insurance, which is largely in deficit): these bodies have little to do with investment and long term growth, and it is not clear that European public opinions will be that pleased when they find that out. When selecting its eligible agencies, the ECB might be well advised to select them not only on the credit worthiness grounds but also according to their economic purpose.
Reference:
Grjebine, T., Szczerbowicz, U., and Tripier, F. (2014). Corporate Debt Structure and Economic Recoveries, CEPII Working Paper, Nr 2014-19 November.
< Back
