The European Central Bank has purchased almost €1.5 trillion of securities since it started buying bonds in March 2015. Most of these are government bonds but starting in June 2016, the ECB began purchasing corporate debt as well, and now holds €50 billion of it.
The logic behind the corporate sector purchase programme (CSPP) was an attempt to jumpstart eurozone growth by lowering the borrowing costs of non-financial companies, encouraging them to invest more in the real economy.
According to ECB president Mario Draghi, the CSPP has already worked its magic. “Corporate bond issuance increased markedly”, he told a press conference in September. “The corporate bond purchase programme has been a success beyond our expectations.”
But how much is cause-and-effect versus correlation? Since the end of 2015, bond yields for euro-denominated investment grade corporate issuers have fallen by about 60 basis points on average, according to iBoxx data. That’s less than the decline in dollar-denominated investment grade yields, which fell by 70 basis points over the same period, despite the absence of any central bank buying.
Exactly, the ECB might respond. It was only by putting its thumb on the QE scale that corporate borrowing costs were reduced in line with the US. This may be true, but in aggregate, it was not the CSPP that delivered that result, but its far larger government bond sibling, the PSPP. Factoring out the government bond impact, the reduction in corporate credit spreads between December and November was only 20 basis points – half the reduction in dollar investment grade spreads.
To really appreciate the cause and effect question, we need to look at the difference in yields between corporate bonds that the ECB (and national central banks) purchased and those that it didn’t. This information can be gleaned from the lists of bonds available for securities lending, disclosed via their identification numbers published by the ECB and eurosystem central banks. While these disclosures don’t say how much of a particular bond has been bought, they do indicate that the ECB has bought some of it.
By comparing the yields of bonds that the ECB has bought via the asset purchase programme with those that it hasn’t, the question of cause and effect can be examined. We have done this via a bubble chart for two countries – Germany and Italy – using data from the end of October. Bubbles are coloured according to whether bonds are part of the APP or not, and sized according to their notional outstanding amount. Use the control at the top of the two charts to toggle the date from December 2015 to October 2016, to see the impact.
As can be seen in both charts, corporate bond yields for German and French issuers declined during 2016. However, the behaviour of the red dots in the German chart is striking – they drift downwards to cluster below the blue dots. In other words, being included in the APP appears to have an effect on German corporate bonds. For France, the effect is not apparent.
We haven’t completely answered the cause and effect question. For example, it would be interesting to know whether the increase in corporate borrowing mentioned by Draghi in his September press conference was concentrated in Germany. Nor do we know why German bonds were so disproportionately affected compared to other countries (Subscribers have access to data for all countries and additional sectors such as supra-nationals). But the tentative conclusion has to be that the CSPP has had a very unbalanced impact. As the ECB prepares to take a Christmas break from its bond buying, this is worth pondering.
