And since, in a way, that’s a real concern and you will surely be inundated with stuff about central banks warping yield curves/ safe asset hunts, this from Deutsche earlier in the week might be useful.
First an estimate on what’s already been bought:
The interesting take-away from this analysis relative to the static analysis is that some of the shorter dated bonds which are currently ineligible for purchases would have been eligible for purchases for reasonable periods of time since the start of the QE programme in March 2015. Therefore, a static analysis highly understates the eligible universe of bonds while a dynamic analysis would more accurately capture the evolution of eligibly universe of German government bonds since the start of the ECB’s PSPP.
We do not have precise information about the size of purchases of bonds which had been eligible in the past but are no longer eligible. At an extreme 33%, i.e. the current issue limit, of these bonds amounting to EUR 156bn in market value terms (which is greater than the estimated amount of German govt. bonds bought so far) could have been bought.
We consider more realistic scenarios where we assume that (1) upto 25% of these bonds (original issue limit) were bought, (2) 33% of bonds which were eligible for 10 months or more have been bought (i.e. a monthly rate of 3.3% during eligible months capped at 33% cumulatively) or (3) 25% of bonds which were eligible for 10 months or more have been bought (i.e. a monthly rate of 2.5% during eligible months capped at 25% cumulatively). Under these more realistic scenarios we find that EUR 85-119bn of bonds which are no longer eligible might have been bought (see Figure 5 and Figure 6).
This implies that of the current eligible universe EUR 15-50bn of bonds have been bought leaving EUR 159-193bn of eligible bonds available for purchases. The monthly run-rate of 15.4bn of German PSPP (o/w 70% is in govt. bonds) implies purchases of 10.8bn of German govt. bonds a month and implies that we will run out of bonds in 15 to 18 months (Sep-17 to Dec-17) and not before the end of the announced QE programme in March-17 (see Figure 6)
Then on the ECB’s ability to widen its available universe (and yes that the ECB would might like Germany to also issue more should also maybe be in our header) as DB admit “the size of purchases of German govt. securities relative to the eligible universe is reasonably large especially if the QE programme is extended beyond March 2017 as is widely expected”:
First, our analysis does not take into account the impact of additional issuance which will increase the eligible universe of German government securities. Over the course of the rest of 2016 Germany will be issuing EUR 53bn of 5Y+ securities which would have the impact of increasing the eligible universe by ~18bn (33% of 53bn). This is to some extent offset by the reinvestment of proceeds upon maturity of bonds. However, based on our estimation we would conclude that reinvestment needs from German govt. bonds could be limited to as little as EUR 2.5bn for 2017.
Second, the ECB could increase the issue limit of 33% to 50% for bonds which do not have collective action clauses (approximated as bonds issued before 1- jan-2013) or even possibly all bonds. This could increase the eligible universe by 66bn in the former case and by 108bn in the latter case.
Third, the ECB could remove the yield floor on purchases which would increase the current eligible universe by 84bn. If this move was accompanied by an expansion of the maturity window to include all bonds with maturity greater than 6 months (as is the case for the CSPP) it would increase the current eligible universe by a cumulative amount of EUR 200bn.
Fourth, the ECB could also expand the list of PSPP of eligible securities to include some more local government debt and increase the allocation of purchases of local and regional government to further alleviate the scarcity issue as far as German government debt is concerned.
Finally, and probably the least likely the ECB could deviate away from the capital key distribution of the PSPP purchases. This could allow a smaller share of purchases of German PSPP securities and increase the allocation to other countries where there are a plenty of bonds available for purchases (for example France, Italy & Spain).
Comments