Speaking earlier on Monday, ECB governing council member Ewald Nowotny said that despite growing market concerns, the ECB "sees no need to set a timetable to end bond buying" adding that "the question is not when but how to continue. That will depend on the economic projections for 2018, which we will have in the fall [...] It’s not about an abrupt halt, but about registering that we are no longer confronted by such an acute crisis as we were when we implemented the measures. I consider it wise to step off the gas slowly."
In other words, just like all other central bank pronouncements, this too was meant to instill confidence in the economy. There is just one problem: the question which Nowotny tried so hard to ignore is precisely the one that matters as we most recently explained in "Both ECB And BOJ Are Just Months Away From Running Out Of Bonds To Buy." The question is even more relevant considering it has been the ECB's purchases of corporate (and government) bonds that has led to a record drop in European credit spreads as we showed yesterday.
But back to the math, because apart from the macro economic rationale for tapering the most important, if not only limiting factor which will result in a moderate tapering of the QE programme starting in the coming months is the lack of eligible government bonds, Bunds in particular, for central bank purchases.
As a reminder, the current eligibility criteria (which admittedly can be changed) for government bonds is that they should be euro denominated, have a remaining maturity of 1Y to 31Y and an issue and issuer limit of 33% applies. The issuer limit is different from the issue limit as it takes into account central banks' holdings of government bonds outside of the asset purchase program as well.
Using ECB data, one can estimate the remaining eligible universe taking into account the reinvestment needs and the impact of gross issuance on the eligible universe. This is what Deutsche Bank has done recently, assuming that 70% of German PSPP purchases are in central govt. bonds with the remaining 30% in regional government bonds and local agencies. The bank then estimates that the remaining eligible universe of bonds is €114bn. Further assuming gross issuance for the remainder of 2017 and 2018 to be € 69bn and € 150bn respectively, I.e. a total of EUR 220bn, the eligible universe increases by 33% of this amount which is €73bn. This takes the total eligible universe by the end of 2018 to approximately €185bn.
At the current pace, Bund purchases until the end of the year should amount to €50bn. Should the ECB continue monetizing debt at the current pace it will not have enough eligible bonds by the end of 2018.
This is where the taper comes in: at an aggregate QE pace of €40bn per month from Jan-18 onward, a €20bn reduction of the current monetization pace, total QE purchases of German govt bonds would amount to €67bn. Additionally, estimating that reinvestment needs until the end of 2018 would amount to €40bn, this takes total QE purchases to €157bn which is comfortably below the available eligible universe of €185bn, however virtually no eligible bonds remain going into 2019.
Summarizing DB's calculations, if the ECB were to reduce the pace of QE to €40bn per month starting from Jan-18, the ECB should not run out of German government bonds to purchase until early 2019. On the other hand, if it keeps the current pace of QE, it will run out of paper by late 2018, and even with a downward revised €40bn monthly total, the ECB will have almost no German bonds left to buy in early 2019. Furthermore, even tapering to €20Bn in late 2018 or 2019 will only extend total QE by just a few more months at best.
There is a last resort: either by then Germany starts running a huge budget deficit - obviously a very touchy political issue - which the ECB will be delighted to fund, or Greece will finally be eligible for QE and will be delighted to step in to Germany's shoes, allowing the ECB to monetize Greek bonds, although it will take some very imaginative Goldman financial engineering to allow the ECB to monetize the (defaulted) Greek bonds that it already owns...