Covered Bond Watch – No tapering of covered bonds

by: Joost Beaumont

  • ECB likely to announce extension of asset purchases at a slower pace on Thursday
  • But we think that the covered bond purchases will not be affected materially
  • The CBPP3’s share of total asset purchases has already declined significantly…
  • …while net purchases seem on a steady path since beginning of last year
  • Reinvestments will keep the Eurosystem a dominant market player in any case
  • It currently owns some 42% of all CBPP3 eligible euro benchmark covered bonds
  • We think that the country breakdown of CBPP3 portfolio is not a top priority…
  • …but the central bank might need to smooth reinvestments somewhat next year…
  • …as reinvestments are heavily tilted towards the first half of 2018
  • We see EUR 27bn of reinvestments in 2018, EUR 29bn in 2019, and EUR 33bn in 2020
  • So, the market impact from ‘tapering’ announcement should remain limited…
  • …with a significant spread widening being a buying opportunity

DISCLAIMER: This report has not been prepared in accordance with the legal requirements designed to promote the independence of investment research, and that it is not subject to any prohibition on dealing ahead. This report is marketing communication and not investment research and is intended for professional and eligible clients only.

Covered-Bond-Watch-No-tapering-of-covered-bonds.pdf (382 KB)

Will 26 October be the D-day for CBPP3?

This week all eyes will shift to the ECB Governing Council meeting on Thursday 26 October. The meeting is likely to provide more details about the future of the central bank’s asset purchase programmes, including the third covered bond purchase programme (CBPP3). Our base scenario is that ECB President Draghi will announce that the central bank will reduce total monthly asset purchases from EUR 60bn to EUR 30bn, while extending the length of its asset purchase programme to September 2018 (see more detail here). We prefer to call this scenario a QE extension at a slower pace rather than tapering. Recently, Bloomberg reported that some ECB officials had leaked a similar scenario.

In any case, it is widely expected that the amount of monthly purchases will be reduced but that the programme will be lengthened. What is more, we expect that the slowdown in monthly purchases will mainly be targeted towards the central bank’s purchases of government bonds and agency debt (the PSPP programme), reflecting the fact that this programme is reaching its limits (see for more detail here). In this note, however, we will have a closer look at what Thursday’s announcement would mean for the outlook of the Eurosystem’s covered bond purchases.

Current CBPP3 practice

The total portfolio of covered bonds that the Eurosystem has bought under CBPP3 now stands at EUR 235bn. One third of the purchases have been bought in the primary market, while two-thirds of the purchases were carried out in the secondary market. Furthermore, the CBPP3’s share of the Eurosystem’s total quantitative easing (QE) package called the Asset Purchase Programme (APP) has dropped to around 5%, down from roughly 20% at the start of 2015 (when the PSPP started). Indeed, the monthly amount of net covered bond purchases (total gross purchases minus reinvestments) has declined to an average of around EUR 4bn in 2016 and 2017, compared to EUR 10bn in 2014 and 2015. The decline was mainly due to the start of the PSPP in 2015, and that of the CSPP in 2016.

Net purchases expected to remain steady

Looking forward, we expect that the net CBPP3 purchases will remain on a steady path, despite a 50% reduction in the total QE amount. This would imply that the Eurosystem would continue to buy around EUR 3bn to EUR 4bn of covered bonds per month. This, in turn, would continue the current trend that started at the beginning of last year. Furthermore, the fact that the monthly amounts of covered bond purchases have already declined during the course of the programme also limits the room to reduce CBPP3 purchases much further. More importantly, the PSPP programme will reach its limits.

Reinvestments will keep Eurosystem dominant market player in any case

On 3 December 2015, the ECB announced that it would reinvest the principal payments on the securities purchased under the APP as they mature, and for as long as necessary. This will imply that the Eurosystem will remain a dominant market player, even if net purchases have been reduced to zero. Indeed, the central bank’s covered bond purchases under CBPP3 equal EUR 234bn, which is some 42% of all euro benchmark covered bonds in the iBoxx index that are CBPP3 eligible. Consequently, this implies that reinvestments across just the CBPP3 programme will total some EUR 30bn annually.

But how long will the Eurosystem remain active in the market?

This is difficult to say, but in general we think that the Eurosystem will keep reinvesting principal repayments well after it has reduced net purchases to zero. But this is of course largely dependent on the outlook for the economy and inflation. If we base our outlook for Eurosystem reinvestments on that of the US Federal Reserve Bank, it might take a couple of years after the end of QE, before the ECB starts reducing its balance sheet. It therefore seems reasonable to assume that the Eurosystem will remain a key player in the covered bond market for a while.

What about the capital key and CBPP3?

The Eurosystem has never explicitly indicated that it needs to stick to the capital key when conducting covered bond purchases, unlike the PSPP. As a result, the central bank has a lot of flexibility when implementing the CBPP3 programme. However, if we look at the country breakdown of reinvestments that will be due next year (see graph below), it should not come as a surprise that most principal repayments stem from the biggest covered bond markets, such as France, Spain, and Germany, followed by Italy and the Netherlands. This raises the question whether the Eurosystem will stick to the country breakdown for its reinvestments.

We think that the country breakdown of the CBPP3 portfolio is not a top priority at the central bank. Moreover, we expect that it takes a rather practical approach when reinvesting maturing principals. This means that it will buy covered bonds that are available at that time (in the primary as well as secondary market) rather than reinvesting, for instance, maturing Spanish covered bonds like for like.


In 2018, reinvestments are concentrated in the first half of the year

Now that we have discussed the broad outlook for CBPP3, we take a closer look at the CBPP3 reinvestment profile in coming years, starting with 2018. We assume that the central bank owns on average 42% of every CBPP3 benchmark covered bond. Although simplistic, we think that this is the best way of getting a feeling for the reinvestment profile, given that we only have high level information about the CBPP3 portfolio.

The graph above show that we estimate that the Eurosystem will need to reinvest around EUR 27bn of covered bonds in 2018, of which EUR 19bn in the first half of the year. This leaves ‘only’ EUR 8bn of reinvestments for the second half of the year. On average, reinvestments will equal EUR 2.3bn per month. But most interesting is that reinvestments will be concentrated in the first half of the year in 2018.

Will the Eurosystem smooth reinvestments?

This raises the question whether the Eurosystem will directly reinvest maturing bonds or that it will smooth reinvestments over a number of months in order to keep the monthly reinvestment amounts roughly similar? Currently, the ECB’s weekly financial statement shows that the central bank immediately reinvests maturing covered bonds. We therefore think that the Eurosystem will prefer to continue this behaviour, but only when market conditions allow.

In June 2018, we forecast, for instance, that the central bank needs to reinvest some EUR 5bn, which would imply that it needs to buy around EUR 8-9bn on a gross basis that month. This might prove difficult in case of low activity in the primary market. Therefore, the Eurosystem will probably be forced to shift part of these reinvestments to July. As a result, it is likely that volatility of the changes in the outstanding balance of CBPP3 will increase. In our view, this will be mainly related to practical impediments with the implementation of CBPP3 rather than to changes in the central bank’s purchase policy.

2019 and 2020 reinvestment profiles

The reinvestment amounts of 2019 and 2020 appear to be larger than that of 2018. In 2019, we currently estimate that the Eurosystem needs to reinvest around EUR 29bn, while this amount will increase to almost EUR 33bn in 2020. This equates to monthly averages of EUR 2.4bn and EUR 2.7bn, respectively.

The reinvestments will also be more equally spread over the year. In 2019, some 64% of reinvestments will be during the first half of the year, while this will be 46% in 2020. This will probably imply that volatility of changes in the outstanding CBPP3 balance will be less in the two years following 2018.

 What will be the market impact ?

Overall, the market impact of any ‘tapering’ announcement should remain modest in our view, as the Eurosystem’s CBPP3 purchase policy is unlikely to be affected significantly. The central bank will continue to be active in the primary market, giving issuers less reason to increase new issue premiums (so keeping the current favourable conditions in the primary market). Meanwhile, the ECB will keep on buying covered bonds in the secondary market, which will also support spreads. Therefore, should spreads begin to widen significantly, we see this as a buying opportunity.


Having said that, the reinvestment profile in 2018 suggests that the central bank might be less active in the market in the second half of the year. This could be interpreted as that the Eurosystem is pulling back from the market, resulting in some widening in spreads. However, it could well be that spreads start tightening again early 2019 when the central bank steps up covered bond purchases again according to our reinvestment profile.

Spreads to continue to be supported by central bank

Overall, though, spreads are currently at record low levels since the outbreak of the financial crisis. This also leaves limited room for spreads to continue to tighten much further. This is especially true for spreads of peripheral covered bonds, which have benefitted the most from the purchase programme. Therefore, we think that these are the most vulnerable for any ‘tapering’ talk. As such, it might be worth considering to switch out of the periphery into the (semi) core, which we deem less vulnerable. We also take into account that the spread pickup of the periphery versus the core has reached historically tight levels, so that a switch can be executed with only a small give-up.

Finally, we think that non-CBPP3 paper will outperform CBPP3 paper in case the Eurosystem reduces its presence in the market. Therefore, we stick to our view to be overweight non-CBPP3 paper. This paper is mostly trading with a pickup versus CBPP3 paper, while it offers also some ‘taper protection’ in our view.