Yesterday for the first time, the various central banks of the Eurosystem disclosed which bonds the ECB had bought under its CSPP program. Specifically, we broke down the purchases of the Bundesbank, which revealed some of the most prominent public company debt issuers in Europe. However, we were curious to get a more detailed look at what Mario Draghi's trading desk was spending their time BWICing all day. For that we went to the undisputed master when it comes to tracking what the ECB does in the bond realm (because the ECB is not buying equities just yet), BofA's Barnaby Martin.
Here is the big picture as revealed in his report today titled "CSPP: Buying Frenzy" - "in just over a month of the Corporate Sector Purchase Programme, the ECB have bought 458 bonds, with virtually no stone left unturned.With the monthly run-rate of buying hovering around the €8.5bn mark, our conclusion for CSPP is, bluntly, that it is too big, too powerful and ultimately too bullish for spreads."
But the best part was Martin's answer to the key question: "So what did they buy?" His answer: "In short, almost everything."
But we don’t think this should be particularly surprising – the weekly CSPP buying numbers continue to suggest a monthly run rate for corporate purchases of between €8bn-€9bn (with the buying pace picking up over the last week). Yet, we think this is too big an amount for a market devoid of supply and thus we continue to believe that CSPP is a very bullish technical for non-financial spreads.
What caught our attention on the ECB’s shopping list?
- They bought “topical” credits such as VW, Glencore and EdF.
- They bought “high-yield” credits such as Telecom Italia and Lufthansa.
- They bought “foreign” credits such Bunge and Schlumberger (US), Nestle and ABB (Swiss).
- They bought long-dated bonds (2036s), but they also bought plenty of short-dated bonds. In fact, 35% of the bonds that they bought are negative yielding.
- The Central Bank of Italy seems to be the most “active” central bank thus far, purchasing 43% of their eligible universe (55 out of 127 bonds).
- The Banque de France seems to be relatively “lagging” at the moment, purchasing just 29% of their eligible universe (115 out of 399 bonds). In addition, the BdF has shied away from buying issuers such as Alstom and Legrand.
- The ECB has been involved in primary – buying recent supply from Bunge, Repsol, ASML, Iberdrola, Tennet, Total and Air Liquide.
Ironically, the most popular name seems to be Deutsche Bahn where 12 bonds from the issuer have been purchased. And yet, Deutsche Bahn bonds have some of the most negative yields in the Euro IG market (DBHNGR 18s yield -25bp).
They bought (almost) everything!
Since June 8th the ECB has bought 440 corporate bonds, which is around 35% of our estimated universe. By issuer, we find that the ECB has bought bonds from 158 different corporates.
The most popular bonds appear to be Deutsche Bahn (12 bonds bought), Telefonica (11 bonds bought), BMW (10 bonds bought), Daimler (9 bonds bought), ENI (9 bonds bought), Orange (9 bonds bought), Air Liquide (8 bonds bought), Engie (8 bonds bought), Iberdrola (8 bonds bought), Total (7 bonds bought) and Enel (7 bonds bought).
But in relative terms, we find that the ECB has bought relatively more of Snam (7 out of 10 eligible bonds have been purchased), Deutsche Post (5 out of 8 bonds), Repsol (5 out of 8 bonds), Total (7 out of 12 bonds), Telefonica (11 out of 19 bonds), EDP (5 out of 9 bonds), RWE (5 out of 9 bonds) and Deutsche Bahn (12 out of 24 bonds).
A glance through the individual names shows that there is very little that the ECB have held back on. In particular:
- Names with “event risk” have been bought, such as VW, Glencore, EdF and Repsol.
- Plenty of BBB3 rated credits have been bought, such as RWE, Metro, EdP, Renault, A2A, Pernod and REN (55% of names purchased were BBBs). Although not every BBB was bought – note that KPN and Alstom have yet to be purchased.
- Foreign issuers have been bought due to their issuing entities being Dutch, for instance. For us, this underscores the point that we made a few months ago that CSPP is really “QE for the world”. Plenty of Swiss credits have been bought (such as Nestle, Novartis and Adecco) as have UK credits (Unilever) and US credits (Schlumberger and Bunge).
- The ECB even dipped into high-yield, buying bonds from Telecom Italia (Fitch rating of BBB-) and Lufthansa (S&P rating of BBB-).
The next chart shows the total number of bonds purchased by country. Issuers are classified based on their ultimate country of risk (rather than issuer entity, such as a Dutch SPV, for example).
- The ECB seems to be lagging a bit in its buying of French credits, with only 115 bonds bought so far. This compares to 122 German bonds bought. Yet our previous work suggested that the ECB would weight CSPP purchases 30% towards French names and just 25% towards German ones.
- Likewise the ECB seems to be a little bit ahead in its buying of Swiss credits, and behind in its buying of US credits.
- The chart also suggests that the Bundesbank and the Banque de France are buying smaller clips, on average, than the central banks of Italy and Spain.
Is the ECB mostly focusing on the highest-rated issuers? Nope. As Martin puts it, "No “softly, softly” for CSPP"
Chart 5 shows the percentage of eligible bonds that the ECB has bought to date, by rating. It’s clear that the ECB has not shied away from taking credit risk. Already they have bought between 36%-52% of eligible BBB bonds, but only 16%-36% of eligible single-As. The ECB has also bought 31% of eligible BB bonds
Chart 6 uses the same methodology and shows the ECB’s purchases by duration. The ECB has bought across the curve, although it does appear that they have tended to stay in the 2-10yr region, and have avoided going very long (10+yr).
Chart 7 shows ECB purchases split by issue size. The ECB has, by far, focused their efforts on “big issues”. For instance, they have already purchased 86% of eligible bonds that have an issue size greater than €2bn. This tells us that a) the ECB is going for the most liquid bonds first, and b) the ECB is cognizant of its desire not to distort asset markets.
Chart 8 splits purchases by yield bucket. We thought it was interesting as it somewhat contradicts chart 5 and suggests that the ECB have been a little tentative in their buying. It highlights that the ECB have bought relatively more of low or negativelyyielding bonds as opposed to higher-yielding bonds. For instance, the ECB has already bought 162 negative yielding bonds (€130bn), meaning 35% of their total purchases thus far have been negative yielding.
What did the ECB not buy?
While the ECB has bought a lot, there are some names which are yet to show up on the shopping list. Chart 9 is a graphical description of what the ECB has bought and not bought, ranking eligible issuers’ debt from highest to lowest. While not every ticker is visible on the x-axis, it does appear that the ECB has tended to avoid buying the smaller names (perhaps they have struggled to buy them?), but has tried its best to buy most of the larger names.
For those eager to frontrun the ECB, Table 1 shows some of the high profile issuers that are yet to appear on the ECB’s shopping list. Note that some of these are French, which ties in with the analysis in chart 4 that French credits have been under-purchased so far.
Finally, here is why - as we hinted last week - corporate yields are set to go even lower, perhaps even dipping below their respectively sovereigns: simply, because the ECB is completely price and yield indescriminate, and it will prioritize volume and liquidity (over price and yield). As a result, we would bet lots of money that in the coming months, Deutsche Bahn will end up trading "through" the matched-maturity Bund, as corporations end up being "safer" than governments, and all thanks to the ECB's trading desk, first profiled here.
The growth of negative yielding corporate debt has been dramatic post the announcement of CSPP earlier in March. As of today, €445bn of Euro IG corporate bonds trade with a negative yield (note that our universe here includes less than 12m bonds).
While there is only around €5bn of IG bonds yielding below -40bp, we calculate a 20bp rally in Euro IG spreads would push this number up to around €55bn, for instance. The significance of this, in our view, is that as spreads rally further, the ECB will be forced to take more duration risk as part of the CSPP.
Because nothing screams financial stability than the European central bank buying not only government bonds, but corporates at negative yields, in the process assuring that virtually all corporate bonds will trade with negative yields.