Por Matthias Melms, NordLB, Franz Rudolf, UniCredit and Maureen Schuller, ING BANK
This article is from the 2017 edition ofECBC European Covered Bond Informationwhich can be accessedHere.
The Covered Bond Purchase Program (CBPP), initiated by the ECB in September 2014 and later integrated into a broader asset purchase programme, is currently scheduled to run until the end of 2017. It has been one of the main drivers of the mortgage bond market. titles covered since its inception. The article deals with the criteria and mechanism of the purchase program, the development of the ECB covered bond portfolio after termination. The article concludes with an analysis of the scenario after the reduction/completion of the program, including the potential impact on spreads.
COVERED SECURITIES PURCHASE PROGRAM 3: THE FACTS
On Thursday, September 4, 2014, the European Central Bank (ECB) announced its plan to purchase mortgage bonds. This mortgage bond purchase program (CBPP) took the markets by surprise at the time and was the third mortgage bond purchase program alongside CBPP1 (July 2009 to June 2010) and CBPP2 (November 2011 to October 2012). . CBPP3 purchases began in late October 2014. The CBPP3 program was originally scheduled to run through October 2016. However, in January 2015 it was integrated into a broader asset purchase program (APP) which it included government bonds and international and supranational debt agreements. Institutions and agencies with a target monthly volume of €60 billion. In December 2015, the ECB's bond purchase program was extended until March 2017 and the monthly volume increased to €80 billion in March 2016, including corporate bonds from June 2016. After a tenure of two years, the ECB announced in December 2016 that from April 2017, monthly purchases under the APP will be reduced from the previous €80 billion to €60 billion, while the duration of the program will be extended to the following year. end of 2017.
The ECB's rationale is that the CBPP3 together with the Public Sector Program (PSPP), the Asset-Backed Securities (ABS) Purchase Program (ABSPP), the Corporate Sector Purchase Program (CSPP) and TLTRO) promote the transmission of monetary policy, it will facilitate lending to the eurozone economy, generate positive spillovers to other markets and thereby ease the eurozone economy. The ECB's monetary policy will help bring inflation rates back to levels close to 2%.
Purchases are uniform and decentralized in the primary and secondary markets, ie Eurosystem central banks purchase eligible covered bonds from eligible counterparties. To qualify for purchase under the program, covered securities must meet the following eligibility criteria:
Be eligible for monetary policy operations in accordance with point 6.2.1 of Annex I to Guideline ECB/2011/14 (eligibility criteria for marketable assets) and also meet the conditions set out in point 6.2 for acceptance as collateral for own use. 3.2. (fifth paragraph) of Annex I to Guideline ECB/2011/14.
> issued by credit institutions in the euro area; or in the case of multiple certificates by special purpose entities based in the Eurozone.
> denominated in euros and held and settled in the euro area.
> Have underlying assets that include exposures to private and/or public entities.
> Minimum premium rating of credit rating grade 3 (CQS3; BBB- or equivalent) from at least one rating agency.
> Covered bond programs that currently do not reach CQS3 in Cyprus and Greece will be required to have a minimum asset rating at the level of the maximum attainable covered bond rating defined for the jurisdiction, provided the Eurosystem credit quality threshold does not apply collateral eligibility requirements for negotiable debt instruments issued or guaranteed by the Greek or Cypriot government, with the following additional risk mitigation measures: (i) monthly reports on pool and asset characteristics; (ii) minimum committed over-guarantee of 25%; (iii) currency hedging with at least BBB-rated counterparties for non-euro denominated loans included in the program hedging portfolio, or alternatively, that at least 95% of assets are denominated in euros; AND
(iv) loans must be made to borrowers domiciled in the euro area.
> Covered bonds issued by suspended issuers of Eurosystem loans are excluded for the duration of the suspension.
> Eligible counterparties to participate in the CBPP3 are eligible counterparties for monetary policy operations of the Eurosystem and all counterparties that the Eurosystem uses to invest its euro-denominated portfolios.
> The Eurosystem will apply an issuance quota of 70% per ISIN (joint holdings under CBPP1, CBPP2 and CBPP3), except in the case of covered bonds issued by issuers in Greece and Cyprus that do not meet the CQS3 rating requirement; A frontend load limit of 30% per ISIN applies to these Pfandbriefe.
> Covered bonds held by their issuer can be purchased under CBPP3, provided they meet the specified eligibility criteria.
Furthermore, the Governing Council decided to make its CBPP3 portfolio available for lending. Lending is voluntary and takes place either through securities lending facilities provided by CSDs or through matched repo transactions with the same eligible counterparties for CBPP3 purchases. Unlike CBPP1 and CBPP2, the current purchase program (CBPP3) did not stipulate a specific minimum size or term for purchased mortgage securities.
PREVIOUS COVERED SECURITIES PURCHASE PROGRAMS
In June 2009, the ECB announced its first €60 billion mortgage bond purchase program (CBPP1), with purchases between July 2009 and June 2010. The program was purchased with a face value of €60 billion and a total of 422 different bonds fully withdrawn, 27% on the primary market and 73% on the secondary market. The Eurosystem mainly purchased covered bonds with maturities of three to seven years, resulting in a modified average portfolio duration in June 2010 of 4.12. In November 2011, the ECB launched its second covered bond purchase program (CBPP2). 40,000 million euros and Pfandbriefe eligible until October 2012. However, accumulated purchases only reached a volume of 16,400 million euros, of which 36.7% on the primary market and 63.3% on secondary markets.
On 23 June 2017, the ECB reported 222.84 billion euros of covered bonds under the CBPP3 at amortized cost in the primary (33%) and secondary (67%) markets. In addition, outstanding inventories from terminated mortgage bond purchase programs amounted to €7.9 billion in CBPP1 and €5.4 billion in CBPP2.
PRIMARY AND SECONDARY PURCHASES
After the Eurosystem started purchases of public sector securities under the Public Sector Purchase Program (PSPP) in March 2015, purchases under the CBPP3 were gradually eased. The ECB's decision, in March 2016, to increase monthly purchases under its asset purchase program (APP) from €60 billion to €80 billion per month did not change this trend. The decrease in the volume of mortgage bond purchases continued after the reduction in monthly purchases to EUR 60 billion per month announced in January 2017. An example of this is the drop in the share of CBPP3 in total asset purchases. That share has averaged 15% in 2015, 7% in 2016, and 4% in 2017 so far this year.
More subdued buying activity was more evident in the secondary market. In October-September 2015, gross secondary purchases (not adjusted for CBPP3 redemptions) averaged €7.5 billion and declined to an average of €3.3 billion in the first five months of 2017. Secondary market purchases steadily declined and markets became more dependent on primary market purchases. This resulted in higher volatility in the primary vs. secondary, depending on the new qualified offer available in each period, particularly evident since December 2016 (see Figure 3).
In 2016, Draghi announced that the Eurosystem would in future reinvest assets in the market to keep the ECB's balance sheet stable. We understand the statement to mean that past due purchases would be reinvested in the same asset class. This means that the ECB will remain active in the market even after the end of the purchase program and will support it with reinvestments.
This requires an assessment of when these purchases are likely to be due. Compared to PSPP and CSPP, the mortgage bond purchase program has different conditions. The Eurosystem can operate in the primary and secondary markets. Furthermore, the limit per ISIN is at a high level of 70% and therefore also allows for considerable accumulation of individual bonds. Unlike the other programmes, no information is provided on purchased ISINs, so the composition of the Eurosystem's CBPP3 portfolio is unknown. Through purchases in the primary and secondary markets, the portfolio size of CBPP3 at the end of May was €219.9 billion, of which €71.8 billion in the primary market and €148.1 billion in the secondary market. The average number of purchases per month shows that the Eurosystem's core market activities remained constant between EUR 1.7 billion (2014) and EUR 2.6 billion (2015). In contrast, the average monthly volume of purchases on the secondary market was only EUR 1 billion in 2017 and is therefore only a fraction of the value at the start of the purchase program (2014: EUR 8.1 billion).
Lack of transparency in accountability
Not only is there no information on the volume of purchases made (only in aggregate form), but there is also no data on the maturity structure of the portfolio. As a result, there is no transparency in this regard. Therefore, we seek a better understanding of the Eurosystem's portfolio based on the available information, in order to have at least an idea of the structure. Therefore, we analyze both primary and secondary market activity and then build an overall portfolio from this information. Where data were not available, we worked on assumptions, which are explained where appropriate.
According to the ECB, 71.8 billion euros were purchased on the primary market under the CBPP3 until the end of May. This corresponds to 32% of the total volume of purchases. However, information about the specific ISINs invested and the amount invested is not publicly available. To better understand which covered bonds from which countries and with which maturities were invested, we evaluated all available transaction reports (282) for benchmark transactions (benchmark = issuance volume of at least EUR 500 million, investment grade, public placement , at least three joint). Engaged Leads) of issuers from Eurozone countries that have qualified under the program's criteria since the effective start of the program in October 2014. Post-deal statistics of investors who participated in the settlement will be provided by the Joint Leaders. These transaction statistics do not expressly identify investors by name, which means they will not contain information about the holdings of the ECB or the national central bank responsible for conducting the transaction. However, investors are divided into groups. These are typically banks, fund or asset managers, insurance companies, pension funds, and also central banks/SSAs or official institutions.
Significant increase in central bank allocation
While the central bank/SSA investor group allocation averaged 10.4% across all Eurozone issuer transactions between 2011 and September 2014, participation increased to 36.1% in the October 2014 period. 2014 to May 2017. From this, we can assume the working hypothesis that, of all securities eligible for the purchase program, the allocation attributable to CBPP3 is less than 36.1% due to additional central bank demand determined prior to program implementation. Based on this working assumption, the CBPP3 results in a nominal value of €69.7 billion, which is allocated to the Eurosystem in the primary market. Taking into account that the volume actually reported by the Eurosystem is EUR 68.9 billion, we are confident that our estimated values sufficiently represent what is actually happening in the primary market. Based on these data, French bonds account for €20 billion (28.6%) of primary market activity, followed by €17.2 billion of German issues (24.7%) and Spanish banknotes with a volume of 11.8 billion euros (16.9%). ).
The maturities of operations in the primary market are distributed over the years 2018 and beyond, with our estimates of the final maturities of the portfolio in 2037 (ABNANV 1 3/8 01/12/37 andACACB 1 1/2 02/03/37). Although only a very small volume of transactions in the primary market will mature in 2018 (1 billion euros) and 2019 (0.8 billion euros), the volume in 2020 will increase significantly to 9.6 billion euros and will reach at least 4, 9 billion euros. EUR ( 2024) annually until 2026. In 2027, the volume due drops to 3.7 billion euros and then remains low in the following years.
The reported volume that the Eurosystem has purchased on the secondary market since the start of the program amounted to €148.1 billion at the end of May. However, it should be noted that, as we understand it, this is a reported net volume and therefore includes reinvestments of past maturing securities and quarterly portfolio adjustments. Additionally, volume includes all secondary market purchases, regardless of bond size. To get an idea of the maturity structure of the secondary market portfolio, we limit our analysis to the benchmark bonds. Based on this, we create a monthly sample portfolio of all covered bonds that qualify for the purchase program (ECB Eligible Portfolio = EEP). The example portfolio in October 2014 had a total volume of 591 billion euros. French mortgage bonds accounted for 34.6% (204 billion euros), followed by Spanish Cédulas Unica with 16.6% (98 billion euros) and German Pfandbriefe with 15.5% (92 billion euros) . Noting that the composition of the EEP is not consistent with the stakes that national central banks have in the ECB and that we believe that the percentage stake should continue to reflect the composition of the EEP, we also account for the portfolio. composition, therefore, we do not expect large differences in maturity comparisons between an unadjusted and an adjusted portfolio for percentage holdings in the ECB. We compare the calculated EEP for each month with the reported ECB purchases, therefore we assume a distribution that is aligned with the EPP distribution.
Significantly shorter terms from 2023 onwards
The estimate thus obtained for the secondary market share of the CBPP3 portfolio covers a portfolio of EUR 140 billion, and the maturity structure indicates that covered bonds between EUR 15 and EUR 20 billion will be repaid annually until 2021. €13 billion remains to be repaid in 2021. The amount will drop significantly in the following years and will fall below €2 billion in 2026. After all, only small amounts will be due in individual years after 2030.
Peak times between 2020 and 2022
To provide an overview of maturities for the entire portfolio, we have summarized the maturity structures of assumed primary and secondary market purchases. Overall, we managed to accumulate a volume of around 210 billion euros, which roughly corresponds to the portfolio volume of 220 billion euros disclosed by the ECB at the end of May. Under this framework, between €19 billion and €26.6 billion matures between 2018 and 2023, with a peak of €26.6 billion in 2022.
This would result in monthly purchases of around EUR 2 billion to EUR 2.5 billion over this period. However, this analysis did not take into account that reinvestments themselves lead to an expansion of the portfolio structure. By current estimates, a reinvestment in year t0 would increase maturities for the next six years (t1 to t6) by 11% of the final volume in t0.
For example, reinvested maturities of EUR 19.4 billion in 2018 would add EUR 2.1 billion annually to the respective maturities between 2019 and 2024. This would increase the maturity profile for the following years to 2031 by EUR 1,000 million annually. This reinvestment effect means that estimated maturities today would increase from €25.9 billion to €30.4 billion in 2020. This effect will persist as long as the Eurosystem reinvests maturities in the covered bond market.
The simulation of the maturity structure of the CBPP3 portfolio shows that the maturities span a very long period of more than 20 years, although 97% of the estimated portfolio matures in the next ten years. Therefore, an average of between EUR 2 and 2.5 billion will be reinvested monthly in the covered bond market until the reinvestments are completed, providing some stability to the market. Furthermore, the reinvestment effect will only amplify it to the extent that the ECB decides to reduce its balance sheet total or reinvest maturities in other asset classes. In general, however, the volume of expirations and, therefore, extensions is lower than what is often speculated in the market, as the volume is spread over a longer period of time.
QE TAPERING: A SCENARIO ANALYSIS
As mentioned in part above, the ECB has taken several extraordinary steps to implement monetary easing in recent years:
1. Reduction in interest rates (0% principal refinancing rate, -0.4% deposit rate);
2. Asset Purchase Programs (CBPP3, PSPP, ABPP, CSPP); AND
3. Unlimited access to central bank funding through refinancing facilities (including multitranche LTROs).
Based on the US experience, the most likely course of action for the ECB to reverse this monetary easing phase is to gradually reduce (i.e., phase out) its net asset purchases before proceeding with restrictive measures such as loan maturities for new debts or hiking fees.
The previous section provided an overview of the support that the covered bond market is likely to continue if the ECB continues to reinvest bonds maturing after net purchases have been reduced to zero. In this section, we discuss three different hypothetical write-down scenarios with reference to the acquisition of liquid assets.
Scenario 1: Abrupt end without bottleneck
In Scenario 1, the ECB will stop adding new bonds to its portfolio under the asset purchase programs until the end of 2017. Past due bonds continue to be reinvested. Reinvestments are assumed to be made within the same asset class as the maturing bonds. Given the proportion of debt already held by the ECB in each bond market, the APP will most likely target new issues for reinvestment either directly in the primary market (where permitted) or later in the secondary market.
Scenario 2: Rapid decline with no change in program scope and focus
Scenario 2 is based on the notion that restrictions on the public sector asset purchase program will force the ECB to slow down its pace of asset purchases in the coming year compared to the current monthly target of €60 billion. We expect the ECB to already own 75% of the eligible debt it can buy under the government purchase program before reaching the 33% ISIN restrictions of that program. This gives the ECB scope to execute the asset purchase program at the current monthly net purchase rate of €60 billion (of which 85% is spent on public sector assets) until the end of March 2018. program should terminate abruptly at this point. At the same time, we expect the ECB to reduce its monthly net asset purchases from €60 billion to €30 billion from January 2018, before the central bank's net purchases to zero at the end of June 2018 reduced.
Scenario 3: Step-down facilitated by expanding program scope
In Scenario 3, the ECB broadens the scope of its asset purchase program to facilitate a gradual reduction in purchases. For example, if the ECB were to broaden the scope of its purchase program to include shorter-term government bonds with maturities of less than one year, the central bank would still have a few months, i.e. until the end of 2018, to reduce asset purchases to zero.
Figure 10 shows the monthly purchase path for the three scenarios. While ECB policymakers have hinted that they might even increase the size or duration of the asset purchase program if economic conditions deteriorate, the number of options on the table for actually doing so has dwindled. Therefore, Scenario 3 also assumes a reduction and only a slight extension of the program duration until the end of 2018. In none of the above scenarios do we consider it likely that the ECB will start replacing a reduction in public sector purchases with an increase in investment in your other shopping programs.
WHAT ARE PURCHASES AT CBPP3?
As mentioned above, the ECB is already a significantly less active buyer of mortgage bonds than it was when the program started. Figure 11 provides an overview of the monthly pace of net and gross purchases (including reinvested redemptions) of CBPP3 on a rolling three-month basis since 2015. Net and gross purchases gradually declined over the year as the asset expanded . purchase program. Since the beginning of this year, the drop in net and gross purchases has been more gradual than in the last two years. In addition, the reinvestment impact of reimbursement became more significant, resulting in a greater difference between gross and net monthly purchases.
This year's flatter net buying trend (shown in gray in Figure 11) would not see the end of net buying until November next year. With the more pronounced downward trend in purchases observed in 2015-2016 (shown in blue in Figure 11), this would occur as early as the second half of this year. The continuation of this year's flatter gross buying trend will result in CBPP3 monthly gross purchases of over €4 billion for most of next year. This is not a realistic assumption, even in our most optimistic exit scenario (Scenario 3).
In fact, phase-out scenario 3 for the APP estimates that net purchases will end before 2019 (thin dashed gray line in Figure 11). If the share of CBPP3 within the APP remains constant at around 4.5%, the drop in net purchases of mortgage securities in Scenario 3 will be a little later than indicated by the net purchase trend for this scenario. year (represented by the thin gray line in Graph 11). Therefore, in Scenario 3, net purchases of mortgage securities exceed the estimated trend of net purchases in 2018.
The situation is quite different for gross purchases. Even if CBPP3 main payments are fully reinvested in covered bonds next year, gross buying activity in Scenario 3 (thick gray dashed line) will tend to decline, and at a more robust pace than the current trend (raw gray line). . Under Scenario 3, gross purchases would be around €1.5 billion in Q4 2018, compared to €4 billion based on this year's gross purchases trend.
REINVESTMENT REIMBURSEMENT GET FULL ECB COMMITMENT
CBPP's total covered bond portfolio increased to EUR 233 billion at the end of May 2017, c. 37% of Eurozone benchmark euro-denominated bonds. We estimate that the total mortgage bond stock will exceed EUR 250 billion by the end of this year, before our tapering scenarios come into effect (Chart 12).
In December 2015, the ECB announced its intention to reinvest redemptions of securities purchased under the APP for as long as necessary. Over the past five years, eurozone banks have issued an average of €83 billion a year in collateralised euro-linked bonds. Reinvestments of mortgage bond redemptions, when undertaken through primary issuance, would coincide with an annual absorption of 23-33% of the CBPP3 in new businesses with similar offering volumes. This compares to the average acceptance of 31% of eurozone primary trades during CBPP3.
Figure 12 provides an overview of the total potential covered bond holdings under the three scenarios described above. If repayments were fully reinvested, the ECB's covered bond portfolio would still stand at between €245 billion (scenario 1) and €260 billion (scenario 3) within five years. If the ECB were to stop buying covered bonds by the end of this year and not reinvest redemptions, the central bank's covered bond portfolio would shrink to just under €135 billion by 2022.
If the ECB continues to reinvest redemptions, CBPP holdings of EUR-backed bonds in the eurozone would still be 35% (Scenario 1) to 37% (Scenario 3) five years from now. Without reinvesting redemptions, that percentage would drop to 19% by 2022. This is still well above the 10% share of CBPP holdings at the start of CBPP3 (Figure 13). This reinforces our expectation that the ECB will remain a major player in the covered bond market for a long time to come, unless the central bank eventually sees the economic case for actively reducing its covered bond holdings through the sale of its portfolio.
EVALUATE THE IMPACT ON SPREADS
The analysis of the historical evolution of covered bond spreads throughout CBPP3 can give us important clues about the potential impact of the gradual narrowing of spreads. To do this, we analyze how the development of the covered bond spread interacts with:
• Support for secondary purchase of CBPP3;
• primary market activity;
• General spread dynamics in the financial sector;
• bond versus swap trading levels; AND
• Renditeniveaus (Swap-Taxa).
However, before proceeding with our analysis, it is important to note the broad-based impact on performance of four different ECB monetary policy announcements over the last three years, discussed in the second paragraph of this article (see Figure 14):
1) 4 September 2014: The ECB announced its intention to start buying covered bonds under the CBPP3
2) January 22, 2015: The ECB announced that it would expand its asset purchase program to include government bonds and SSAs. The overall size of the program was calibrated at €60 billion.
3) March 10, 2016: the announcement of the extension of the asset purchase program to corporate bonds, a new round of TLTRO, the reduction of the deposit rate to -0.4% and the increase in the amount of
the program from EUR 60 billion to EUR 80 billion.
4) December 8, 2016: The ECB announced its intention to extend the duration of the program until the end of 2017 (from March 2017) and reduce the size of the program to €60 billion from April 2017,
and allow purchases at a rate lower than the -0.4% deposit rate.
The mere announcement of these policies has proven to be far more important to performance than actual implementation. This suggests that any ECB tightening proposal will have a more definitive impact on spreads than the actual final pullback in buying.
Secondary buy of CBPP3 and spreads
In addition to the positive reaction from the covered bond market to the various QE announcements, covered bond spreads were well supported by the proven presence of CBPP3 in the secondary market through the summer of 2015. 2015 coincided with a notable widening of covered bond spreads for up to the ECB meeting in March 2016.
Primary market activity and spreads
The slowdown in secondary purchases in the second half of 2015 coincided with a strong recovery in the supply of covered bonds in the euro area (chart 15). This abundance of new shares led to the purchase of CBPP3 moving from the secondary to the primary market at a time when the increased supply was already putting pressure on mortgage bond spreads. This suggests that if tapering were to occur, a drop in covered bond purchases would likely have to coincide with another negative performance, such as a high bid, to trigger widening of the spread.
Diffusion dynamics in finance
Figures 16 and 17 show the relationship between senior unsecured and secured bond spreads since the beginning of 2014 for France (conforming CBPP3) and Sweden (non-compliant). Indeed, CBPP3-backed covered bonds, in particular, barely expanded in the first half of 2015, while senior unsecured bonds reflected verifiable expansion. The ratio of spreads between senior unsecured and covered bonds prior to CBPP3 shows that further widening of covered bond spreads would have been possible without support from the purchase program. This was also evident in the Swedish covered bond market, where spreads on covered bonds increased significantly in line with spreads on senior unsecured bonds.
In the second half of 2015, senior unsecured bonds started to decline, while covered bonds continued to expand. The historical (ie pre-CBPP3) spread relationship between senior unsecured bonds and covered bonds has been reinstated, although the ECB continues to buy covered bonds. This confirms our view that the additional support for CBPP3 covered bond spreads is now gone.
However, the charts illustrate that spreads on non-CBPP3-backed covered bonds exceeded their 2014 pre-call program highs in 1Q16, while CBPP3-backed bonds did not. While CBPP3-backed covered bonds are expected to be as resilient as ever in response to a general deterioration in credit market conditions, the yield outlook for non-CBPP3-backed jurisdictions could still become more favorable once the ECB has completely disinvested from the market. covered bond market.
Government bond trading levels
Figure 18 provides an overview of the relationship between covered bonds and Bund asset swap spreads. Higher Bund trade levels versus swaps are constructive for tighter covered bond spreads. Therefore, as Bunds begin to underperform swaps in a tapering environment, we are likely to see covered bond spreads increase at the same time. This may be more important to covered bond performance than the actual drop in net purchases to zero below CBPP3.
Underlying achievement levels
However, in addition to buying less cheap government bonds, we would likely need to see underlying yield levels rise to support a widening of covered bond spreads to levels seen before the ECB's covered bond purchase announcement in September 2014 (Figure 19).
LESSONS FROM THE FED
Figure 20 provides an overview of the impact of QE tightening on US Treasury and USD-backed bond spreads. The Fed's initial tip and subsequent anticipation of markets tightening had a greater impact on spreads than the actual beginning of the tightening process in December 2013, when the Fed first tightened its $85 billion monthly purchases for the US $75 billion. The subsequent $10 billion cut for each Fed meeting had minimal negative impact on spreads. In fact, spreads have narrowed.
The greatest spread expansion occurred after the end of the tapering process, when the market began to anticipate the Fed's first interest rate hike. After the first 25 basis point increase in late 2015, Treasury spreads have remained relatively stable. This suggests that real monetary tightening, which the Fed says will involve a gradual reduction in reinvestment later this year, is likely to have a bigger impact on spreads than a gradual reduction in asset purchases.
Both the impact of the QE announcements on spreads and the tapering experience in the US suggest that the spread impact on mortgage bonds may be more pronounced after the ECB's initial indications of tapering than during the tapering process itself. of no reduction will have a stronger ripple effect than a slow reduction process. Once the phasing out of asset purchases is complete, the market will likely wait for the central bank's next move in the form of financial tightening. This could trigger a significantly stronger amplification response than any of our reduction scenarios discussed here.