The ECB’s reinvestment policy: Issues and options

5 min
  • All eyes on the reinvestment policy: Monthly net purchases under the ECB’s QE program are expected to end come January. The pledge to reinvest maturing bonds will help cushion the impact on financial conditions by ensuring that the ECB maintains a hefty presence in eurozone bond market.
  • Implementation bottle necks & tighter financial conditions: Maintaining the current degree of monetary policy accommodation would require reinvesting principal in notably longer-dated sovereign bonds. However mounting implementation constraints – particularly in low-debt core countries – would likely force the ECB to tolerate at least a gradual rise in long-term interest rates.
  • Desperately seeking flexibility: To ensure the reinvestment policy’s smooth implementation, we expect the ECB to soon opt for more leeway around its modalities. With the exception of temporary deviations, the capital key constraint should remain sacrosanct, but more flexibility regarding the timing of reinvestments and the composition of purchases should already suffice to do the trick.
  • The bottom line for eurozone yields: We expect the rise in long-term rates in 2019 to prove more pronounced in non-core countries due to relatively less favorable stock as well as flow dynamics. Italy stands out as the country with the least favorable prospects judged by demand and supply factors underpinning its sovereign bond markets.

All eyes on the reinvestment policy

The ECB has committed itself to reinvesting the principal from maturing securities purchased under the Asset Purchase Program (APP) since 2014. But with monthly net purchases expected to end after an almost four-year run come January, the reinvestment policy will now take center stage in determining financial conditions in the eurozone until at least September 2019 when we expect the ECB to implement its first timid rate hike.

Reinvestments: A cushion for 2019 and beyond

Whereas net purchases will stop next year, thanks to the reinvestment policy bonds bought under the PSPP program remain on the ECB’s balance sheet for an extended period of time. As maturing principal is reinvested – about €165 billion in total for 2019 – the overall size of the ECB’s Public Sector Purchasing Program (PSPP) portfolio remains constant in euro amounts for as long as the reinvestment policy is fully in place.

In some months in 2019 PSPP redemptions are expected to exceed the estimated average pace of 2018 net purchases (Chart 1). Nevertheless the overall monthly flows will prove much more volatile compared to current net asset purchases – ranging all the way from an expected € 4 to 21 billion – with the volume of principal reinvested dictated by the securities maturing around that time. For 2019 as a whole PSPP gross purchases will more than halve compared to 2018 (Chart 2).

Maturity matters most now for monetary policy

With the ECB still absorbing a big chunk of the interest rate sensitivity (i.e. duration) risk from bond markets that would otherwise have to be borne by private investors, term premia – the extra return that investors require to compensate them for the risk associated with a long-term bond – will remain compressed. The key indicator to watch in order to gauge the degree of monetary stimulus in the eurozone is the PSPP portfolio’s residual maturity (Chart 3). The longer the maturities targeted by the ECB, the more pronounced will be the downward pressure exerted on term premia and hence on eurozone long-term yields.

For the ECB to maintain the current degree of monetary policy accommodation over the course of 2019 would prove highly challenging. As time passes, the PSPP portfolio leaks duration as the securities bought by the ECB mature. To counter the declining residual maturity of the PSPP portfolio - and the de facto tightening in financial conditions - the ECB would have to notably extend the maturities of PSPP reinvestments. A practical example helps illustrate why: In 2019 the ECB will reinvest maturing securities to the tune of €165 billion – equivalent to approximately 8% of its PSPP portfolio by end-2018. Due to this relatively small share of reinvestments, maintaining the residual maturity over the course of 2019 at the current level of 7.5 years (Chart 3) would require the ECB to reinvest in eurozone debt securities with an average maturity of 12.7 years, notably above the current average residual maturity of the PSPP portfolio. For the purpose of comparison: since early 2015 the maturity of monthly PSPP purchases has ranged between 8 and 11 years. Due to implementation constraints, however, extending the maturities of all 2019 PSPP reinvestments to 12.7 years a la the Fed’s ‘operation twist’ is not a feasible option. 

Eurozone financial conditions to tighten in 2019

Market neutral implementation of reinvestment policy not feasible

We expect financial conditions to tighten gradually over the course of 2019 for the simple reason that maintaining the current degree of monetary policy accommodation over the course of 2019 would run into mounting implementation constraints.

In contrast to the US, a key challenge for the implementation of the ECB’s reinvestment policy is that it has to be rolled out across 19 individual bond markets. The going assumption is that reinvestments are to be split across eligible eurozone jurisdictions according to the ECB's capital key as has been the case during the period of net purchases. 

The heterogeneity of eurozone bond markets – notably in terms of liquidity across the curve and weighted average maturity – will severely complicate the implementation of the reinvestment program along the lines of the capital key criteria. Germany stands out once again as a bottle neck due to two complicating factors:

  1. High share of redemptions: The German PSPP portfolio boasts one of the lowest residual maturities with 6.4 years compared to 7.5 years for the eurozone as a whole. Hence we expect total 2019 PSPP redemptions to be skewed towards the eurozone’s largest economy with German redemptions making up 30% notably above the country key of 26.3% (Chart 5).
  2.  Liquidity / maturity constraints: Assuming that the 33% issuer limit– the maximum share of an issuer’s outstanding securities the ECB is allowed to hold – is within reach in Germany (Chart 4), reinvestments are dependent on about 33% of new supply. The average maturity of new issues however is around 8 years notably below the 12.7 years that is needed to keep the PSPP portfolio’s maturity constant. An expected negative net issuance of medium to long-term (MLT) debt securities in Germany in 2019 (Chart 9) as public debt declines, reinforces scarcity issues particularly at the long end of the German government debt curve. Extending the maturities of German PSPP reinvestments would probably see the ECB hit the issue(r) limit with the curve so sparsely populated at the long-end. From January 2019 onwards the ECB may decide to conduct QE reinvestments according to slightly revised capital key figures. The recalculation – which takes into consideration the growth of each eurozone country’s economy and population over the last five years –would likely see Germany’s share in the QE program rise. German liquidity constraints would hence become even more pronounced.

Since a marked increase in the maturity of German reinvestments is not feasible, the remaining 70% of PSPP reinvestments will have to overcompensate by targeting non-German eurozone securities with an average maturity of more than 14 years to offset the portfolio ageing effect.

The implications are that if the ECB wants to maintain the current degree of monetary stimulus, the reinvestment policy cannot be applied in a “market neutral” manner across eurozone countries. For one, the ECB would have to pursue very different strategies for high-debt/non-core vs low-debt/core countries by buying much longer-dated securities of high-indebted countries to make up for buying relatively short-dated debt in eurozone countries with healthier debt levels. In addition, buying notably longer than the average maturity of outstanding government bonds clearly runs counter to the ECB’s guiding principle to act as “market neutral” as possible in order to minimize the potentially distortive effects of purchases on the functioning of the financial markets.

To help alleviate the issue of non-neutrality we expect the ECB to accept a gradual rise in long-term yields over the course of 2019. For instance, a decline of residual duration by 0.34 year over the course of 2019 to 7.2 years (broadly in line with the decrease in residual maturity observed on average from 2017 to 2018 YTD) would call for an average reinvestment maturity of 9 years – a much more feasible task to accomplish, in line with recent ECB communication.

Desperately seeking flexibility 

In addition to tolerating a gradual tightening in financial conditions we expect the ECB to soon be forced to tweak its reinvestment policy to ensure a smooth implementation in the context of mounting liquidity and issue(r) limit constraints.

Fully addressing the bias towards German redemptions in 2019 PSPP reinvestments would essentially require dropping the capital key constraint which from our perspective remains a political no-go. We see three feasible policy tweaks that the ECB could apply that would help smooth the implementation of the reinvestment policy:

  1. More leeway around the timing of reinvestments: Whereas the going assumption is that the same rule will be used as applied during the period of net purchases extending the period from three to at least six months would help smooth the reinvestment flow. Such a tweak should be relatively uncontroversial from a political point of view. However by itself it might not be sufficient to allow the ECB to stay below the 33% issue(r) limit in countries like Germany where the limit is already within reach and 2019 net issuance is negative.
  2. More flexibility regarding the composition of purchases in addition to more leeway around the timing of reinvestment would in our view already be sufficient to tackle key implementation constraints. For instance national central banks could receive even more leeway when deciding between purchases of central government securities and securities of regional governments and certain agencies established in their respective jurisdiction in times of heightened liquidity and/or capital key constraints.
  3. Temporary deviations from the capital key constraint with the view of at least keeping the stock of PSPP purchases constant: This option may provide some much needed – albeit limited - flexibility. After all 80% of the PSPP program is implemented at the discretion and at the risk of national central banks. We deem it unlikely that the German Bundesbank will load up on non-core country securities due to the German liquidity headache. Any flexibility would hence be reserved for the remaining 20% for which the ECB itself carries the risk. This tweak should be kept in mind as an emergency back-up to counter any undue tightening of financial conditions in a scenario where the implementability of the reinvestment policy would otherwise be threated. Politically however any deviation from the capital key constraints – even if limited in size or duration – would be very controversial.

The bottomline for Eurozone yields

How significant will be the increase in benchmarks yields in 2019?

Due to the discussed obstacles that stand in the way of a full-fledged ‘operation twist’, we expect the ECB to tolerate a gradual decline in monetary policy accommodation – and hence the downward pressure on government bond yields – over the course of 2019. Still very powerful stock effects will keep the upward rise in yields contained, this is particularly true for countries where the issuer limit is within reach thanks to a relatively low debt burden compared to their capital key. Germany is of course the most prominent example in this category. The ECB estimates the bond market free float – the fraction of outstanding bonds that is held neither by the Eurosystem under the public sector purchase program nor by foreign central banks as part of their foreign exchange reserves – to have dropped below 20% from around 50% before the start of QE.[3]

Meanwhile flow effects are also supportive of keeping a lid on German long-term yields thanks to a negative net issuance of medium to long-term bonds (Chart 9) and the skew towards German redemptions in 2019 (Chart 5). Thanks further to rising core inflation and higher US yields we expect 10yr Bund yields to reach 0.7% by end-2018 and 1.25% at end-2019.

United they fell, divided they will rise – in particular Italy

While we expect the increase in the eurozone benchmark yield to be rather muted based on a stock-flow analysis – leaving (Italian) politics aside –the rise in long-term borrowing costs for more highly-indebted eurozone countries should prove more pronounced. For one, stock effects are relatively less powerful in highly indebted countries with the stock of PSPP purchases making up a lower percentage of total debt (Chart 4). Meanwhile flow effects should also prove less favorable with the market having to absorb larger shares of securities in 2019 as PSPP net purchases end (Chart 7) coupled with positive net issuance of sovereign bonds in the Eurozone (Chart 6).

Italy stands out as one of the non-core countries that faces the least favorable stock as well as flow effects – and even more so when factoring in an expansionary fiscal policy (Chart 8). We think that any maturity extension achieved for PSPP reinvestments will be outweighed by these unfavorable stock and flow effects for non-core countries and most notably for Italy.