As we and many others have noted on any number of occasions since the official launch of QE in Europe, Germany faces the very real possibility of not being able to fulfill its asset purchase targets under PSPP. This is in part due to the fact that the ECB will not purchase bonds that carry yields below the deposit rate of -0.20%, making shorter-dated German paper ineligible. This is a problem for a number of reasons, but at the perceptual level, just about the last thing the market needs is for the Bundesbank to admit it can’t hit its targets. This would amount to a forced tapering [6] and would also mark the point at which the market simply cannot suffer further asset purchases proving, once and for all, that central bank fantasy will eventually run square into reality triggering the infamous “taper tantrum.”
Hilariously, this undesirable outcome has been made a bit less likely by the fact that ECB issue limits apply to nominal amounts meaning that, to the extent NCBs can buy bonds at a premium to par, they can effectively buy less bonds than they otherwise would have and still hit their purchase targets. In other words, if you overpay, it’s easier to stay under the issue cap when supply is scarce in eligible paper. We argue that buying €1.1 trillion in bonds at 124% of par when yields on some of those assets could potentially soar in the event the market’s faith in the supposedly indissoluble nature of the EMU is shaken is probably a bad idea [7], as it effectively means that the minute redenomination risk gets priced into sovereign spreads you are insolvent from an accounting perspective, but then again, when has the fear of operating from a negative equity position ever stopped the ECB?
Given all of this, market value becomes a key factor for the ECB when eligible assets are scarce (as they are in Germany). Here’s BNP:
Germany has the highest ECB capital key (18%) determining the amount of the country’s assets to be bought, the lowest annual net supply of bonds (EUR 4bn) and the lowest yields. Thus it is the only EU member state where the implementation of QE has become a real concern with yields at record lows…
Here’s a look at spread compression since the start of PSPP and the breakdown of eligible vs. ineligible bonds (based on the depo rate floor):
[8]
More from BNP:
Only EUR 551bn of German bonds remain eligible for QE purchases at the moment in terms of maturities and yield. We estimate that, until September 2016, the QE programme will purchase EUR 168bn of German government bonds (assuming a market cap split between government bonds and agencies and a continuation of the current monthly pace of purchases of covered bonds and ABS). This implies that QE purchases will amount to 30.5% of the current universe of QE-eligible bonds (EUR 168bn purchases of EUR 551bn of QE-eligible bonds). On the face of it, this exceeds the 25% limit on ownership per ISIN. However, two critical factors need to be taken into account: i) the price effect and ii) the future bond supply effect.
The ECB’s document “Q&A on the public sector purchase programme” clearly states that the 25% per issue and 33% per issuer limits should apply to nominal rather than market values. Therefore, the purchase price of securities is of critical importance as we need to translate the EUR 168bn of purchases of German bonds from market value to face value. ECB member Coeure confirmed this point in a recent speech when he said, “Note, however, that the average weighted price of bonds for the 2- to 30-year maturity range is well above par, which should mitigate concerns about the reduction of net issuance to some extent, as the purchase target refers to settled (rather than nominal) amounts.”
This works out great if you’re the Bundesbank because given dirty prices for bunds ….
[9]
...EUR 168bn of cash QE purchases will acquire a total nominal value of EUR 132bn!
Of course purchases will have to skew towards longer-dated bunds to avoid hitting the -0.20% floor, and longer-dated bunds are trading at even greater premiums, meaning that in the end, “the weighted-average price of the remaining bonds is 132%, 5% higher than the weighted-average price on all 2-31y bonds.”
Here’s the takeaway from BNP:
Despite the fall in German yields which cause more short-term bonds to become ineligible, when we take into account both the face value adjusted purchases by the ECB and the upcoming bond supply over the next 18 months, the 25% maximum ownership per issue will not necessarily be exceeded. However, for this to be the case, purchases of longer-dated bonds will need to be higher than implied by the simple “weighting by nominal outstanding amounts” of the eligible 2-31y range. This explains the strong flattening of the German curve that we are witnessing this week...
Therefore, for the time being, the problem is one of a scarcity of bonds – where it is difficult to find the all the bonds across the curve that need to be bought – but, if yields keep falling further; eventually it will become a problem of shortage of bonds due to the 25% threshold in terms of ECB ownership – where there is not enough bonds for the ECB to buy the required amounts.
***
What all of this means is that at the end of the day, depending on how low the ECB manages to drive yields, it could indeed come to pass that a shortage of purchasable assets materializes, leading to what we again emphasize will be a de facto taper. This will effectively prove that central bank ambition must at some point be constrained by market realities and that realization, dear friends, will be a bitter pill for Mario Draghi to swallow.
