As we know, on 10 March the ECB announced a new buying programme starting in June called the CSPP, which stands for Corporate Sector Purchase Programme. Since then credit markets and especially corporate bonds eligible for the programme have outperformed all other major fixed-income sectors in the eurozone. In this piece we assess the very first months of the programme’s operations, which began on 8 June, just two weeks before the UK referendum. In particular, we analyse the CSPP’s development in its first weeks of operation, its scope in both quantitative and qualitative terms, its effect on supply and demand for corporate bonds and its final impact on spreads and yield levels. All this in order to assess the opportunities remaining within the EUR corporate bond segment in terms of prospective performance and relative value.
Let’s start with the programme
For the most part, empirical evidence from recent years shows that the announcement effect of new QE measures tends to deliver most of the positive consequences produced on the asset class by the subsequent implementation of these measures, especially when markets are taken by surprise as in the case of the CSPP. The announcement of the CSPP on 10 March immediately tightened credit spreads significantly. However, unlike what happened with other asset classes, the way the ECB actually communicated and managed purchases in June was particularly effective in increasing the downward pressure on both spreads and yields paid by corporate issuers, even in difficult periods such as those preceding and following Brexit.
This effectiveness is certain to be attributed to some factors:
Moving on to the weekly number of purchases, the extreme flexibility of interventions on the corporate sector stands out: graph 1 shows the amount of average daily purchases during the first eight weeks that the CSPP was in operation. In the week of the referendum held on 23 June regarding the UK’s membership in the European Union, as well as during the week preceding the vote, the ECB accelerated its purchases: a peak of 530 million daily on average was reached in the week of Brexit, evidently in order to mitigate the adverse effects on the credit market. Subsequently the interventions slowed and in the last weeks of July they reached a minimum of 273 million daily on average, almost 50 % down from peaks reached just a few weeks earlier. Subsequently, the size of interventions slowed even further, hitting bottom with a daily average volume of EUR 250 million in the second week of August.
The slowdown in volume was certainly also the result of the lack of primary market activity in July due to the referendum’s outcome and the volatility on equity markets. The details released by the ECB about the programme, as we wrote, revealed an impressive number of bonds "touched" by ECB purchases: about 450 bonds issued by about 160 different companies were bought in just a few weeks. This had risen by the end of July to almost 500 bonds issued by nearly 170 companies. Also, in this case there was a sort of "deceleration" in the distribution of purchases: twenty-eight new bonds were added to ECB holdings in the penultimate week of July while only eleven other issuers were added to the list in the final week. New issuers were added in August, but the number was limited compared to the very first weeks: the slowdown was inevitably linked to the traditional decline in activity in the primary market.
A paper published by the ECB in the form of a box in its August Economic Bulletin (“The corporate bond market and the ECB’s corporate sector purchase programme”) highlights some of the points we just focused on. The paper outlined more details about the size of trades under the CSPP and added some colour regarding the sectors represented in ECB portfolio. The ECB underlined that “Average trade sizes under the CSPP are broadly comparable to those under the third covered bond purchase programme (CBPP3) and smaller than trades under the public sector purchase programme (PSPP).” The reduced liquidity of the corporate bond market with respect to the government bond market led the ECB to limit the size of the average trade: “…trades of less than EUR 10 million make up the majority of the volume under the CSPP … and trades are typically larger in the primary market than in the secondary market”. Then, in order “To support market liquidity, the Eurosystem has since 18 July made its CSPP bond holdings available for securities lending via the national central banks conducting purchases.” Of particular interest is the sector breakdown of the CSPP portfolio, which seems to confirm that the weight of the debt market plays a significant role in the allocation of purchases among sectors. At the same time, sector allocation shows that this is not the only factor considered: utilities, for example, are more heavily weighted in major benchmarks than in the CSPP portfolio. Meanwhile, the weighting of telecoms and manufacturing is very similar to the benchmarks’, and consumer goods and sectors with lower benchmark weights appear slightly “overweighted” by the ECB’s CSPP. In our view, this is consistent with the goal of keeping the programme more diversified and “growth” friendly.
Let’s move on to its effect on credit markets
Negative returns do not seem to "scare" central banks that buy corporate bonds within the ECB programme: by the start of August, about 25% of the total bonds purchased were offering negative yields to maturity, with peaks of 50% and 40% respectively in the case of Germany and France. The CSPP announcement, followed by the actual start of programme operation, had an impressive impact on YTM paid by corporates: on 10 March, almost no EUR IG corporate debt offered negative yields while by late July about EUR 250 billion of non-financial corporate debt and EUR 60 billion of financial debt was quoting in negative yield territory. This amounted to 18% of total non-financial debt and 7% of total financial debt, respectively. If calculations are restricted to figures for non financial issuers domiciled in the European Union, the percentage of debt with negative YTM rises to almost a quarter of the total, while 50% of the debt offers very close to zero yield (namely, between 0% to 0.5% YTM). Negative yields are mainly concentrated in debt issued by German and French issuers, together with debt issued by other core countries, but an increasing number of peripheral issuers are joining the club. Even the number of low IG-rated BBBs offering negative-yield debt is growing, next to the AA and A rated names.
The tables and graphs 8 and 9 in this report summarise the average yield to maturity offered by non-financial and financial IG issuers, broken down by rating and curve segment. For non-financial issuers, 90% of the debt with one- to five year terms currently contributes to negative or near-zero yields, while for financial issuers this debt component represents between 60% and 70%. Spreads for this component vary between 50 bp and 90 bp. These spreads are higher than the lows reached in the previous cycle, but absolute yields are not comparable anymore. In the one- to five-year segment, as of writing this, a massive 60% of the yield still available in the market is concentrated in a tiny 4% of the overall debt. This 4% is represented by the BB- and B-rated names of the HY world. Another 24% is offered by a modest 9% of the overall EUR debt market (BBB corporates, especially financials). The inclusion of a percentage of BB-rated bonds in the ECB purchasing plan has significantly contributed to the yield compression in the one to three-year and three- to five-year segments. Currently, the average YTM of BBs in the short maturity segment (one- to three-year) is 2%, rising to 3% in the three- to five-year segment. As a part of the IG universe, non-financial debt fails to offer 1% yield to maturity if both credit risk and interest rate risk are not combined. In fact, only BBB bonds with a remaining term of more than seven years fall into this category. In summary, a close look at the yield curves of corporate bonds shows that non-financial bonds, the target of ECB buying, are characterised by significant flattening. The curves of financial bonds, on the contrary, are steeper not only in the BBB-rated segment but also in the A-rated segment.
CSPP impact on technicals: multifaceted support
The CSPP is the most important supportive technical factor in place for the EUR credit market: however, the support does not come only from ECB direct purchases, but also from other channels. Let’s analyse them: first of all, the announcement acted as a catalyst for private investor demand, spurring remarkable investment flows into funds and ETFs dedicated to EUR IG corporate bonds. From 10 March to the time of writing, there has not been a single week of net outflows from the asset class and the cumulative inflows reached around EUR 15bn in less than five months. Another positive effect induced by the ECB programme is represented by the return of supply, which had fallen to very low levels in the months preceding March. A strong recovery in primary market activity on both gross and net supply levels was a direct result. The re-opening of the primary market was good news for issuers and liquidity, despite the fact that it partly overcame the supportive effect on the demand side. At the same time, the recovery of new issuance in March, April and May finally slowed down significantly in June and July. This slowdown contributed to the post-Brexit tightening of spreads. Despite this, as of writing this, the pace of CSPP purchases somewhat decelerated, too. Another important point to be considered is the fact that most of the slowdown in new issuance in July has been among European names, and particularly in periphery names, while US issuers were much more active: this probably added to the CSPP’s technical impact. Finally, together with CSPP, as we know, new LTROs transactions were announced and in June the first one became operational. The banks’ net demand totalled around EUR 30 billion, mainly from periphery institutions: according to our data, the net issuance of new debt by financials was negative in both June and July. We guess that this could be mainly due to two factors: on one side, the injections of long term financing through the first new LTRO, and on the other side the more challenging market conditions faced by banks after Brexit coupled with the 28 July bank stress tests. In a nutshell, ECB is having quite a remarkable direct effect on demand for corporate bonds. By buying more than 1% of the market outstanding each month, it has also spurred a strong increase in demand for the asset class, and finally it is reducing the recourse of financials to the bond market for their funding needs.
The CSPP and its effect on the correlation of spreads with equity and with equity implied volatility
Some of the charts in this report show the link between credit spreads, equity implied volatility and equity prices. As we know, credit markets tend to lead equity markets in turning phases of the macro and micro cycle, but most of the time spreads are negatively correlated with equity prices. One link which, however, looks quite strong and stable over time (especially in turning phases) is represented by the relationship between credit spreads and equity implied volatility. In this case the correlation is quite positive and this input offers a powerful explanation for regressing credit risk premiums. In severe and long lasting crises and/or in recessionary phases (like the Lehman crisis or the sovereign crisis), equity implied volatility tends to spike together with spreads but then it leads spreads to the downside when the following recovery takes place. Spreads may even lag the VIX or V2X indices by months in recovering after crises peak. On the contrary, when the spike in volatility is not persistent and does not come before an extreme event or a remarkable downturn in economic activity, as has occurred on a number of occasions over the last year, spreads tend to remain well-correlated with VIX and V2X even in the recovery phase. Graph 5 shows the link between the spreads of a good proxy for IG CSPP eligible corporate bonds and V2X over the last year. Chinese devaluation and the return of idiosyncratic risk in the summer of 2015, the sharp fall of oil and energy prices between December 2015 and February of this year and Brexit fears last June all sent V2X to similar peaks in the 40s area. Spreads coincidentally followed or moved in the first two crises. However, after the 10 March ECB announcement, the link seems to have been broken in the very first phase of renewed volatility. If the “Brexit phase” is considered a sort of first empirical test of the post-CSPP link between spreads and equity volatility, we may reach the conclusion that the protection offered by the ECB for corporate bonds worked quite effectively. In a nutshell, the CSPP seems to work in the same way QE worked and still works for periphery government bonds: in this respect, EUR corporate bonds should maintain their more resilient behaviour compared to other risky asset classes in phases of renewed volatility.
CSPP impact on valuations: non-financial spreads become rich compared to fair values
In previous publications we frequently presented the results of our fair value tools: they had shown the extent to which external shocks like China’s devaluation or the dramatic fall in oil prices, together with the temporary return of idiosyncratic risk, pushed both EUR IG and HY spreads to levels not justified by the results of our regression-based fair values. In February this year the gap between market spreads and fair values reached peaks comparable with 2011 crisis levels, namely 170 bp for EUR HY and 50 bp for BBB-rated issuers. Then, starting in March, the gap shrank thanks to the ECB’s CSPP announcement and to supportive macro developments (oil price recovery and better Chinese data, above all). The exclusion of banking-sector debt from the CSPP caused nonfinancials to massively outperform on credit markets. We re-ran our regressions over the last few months, showing that the market spreads of CSPP-eligible bonds finally closed the gap with fair values and fell below them. As of this writing, non-financial BBBs look rich and on average pay a spread that is 20 bp tighter than fair value. This seems to fully confirm the effectiveness of ECB policy on cost of funding for corporates, as well as the strong impact from technicals and the search for yield. In contrast, HY bonds still look cheap by around 30/40 bp vs their respective fair values, together with BBB financials, which are offering an average premium of 35 bp vs their fair values. Relative values for major credit segments, in a nutshell, show that CSPP-eligible bonds are now trading generally below fair value, while other credit instruments still pay a positive premium vs regression factors.
BoE finally following ECB on corporate buying
In August, the BoE finally decided to follow the ECB on corporate purchases and banks’ term financing. The scale of the purchase programme does not seem to be in line with the ECB’s CSPP, when adjusted by the relative market sizes (around one third). However, the combination of direct purchases and term financing for banks is going to produce similar effects in the UK credit market, and will ultimately provide even more support for the search for yield in the eurozone and through USD corporate bonds.
Relative value with USD-denominated IG corporate bonds
The comparison between EUR and USD IG corporate spreads and yields may be summarised in the tables, showing the breakdown by rating and curve segments. As expected, a close look at the numbers underlines that the gap in favour of USD debt is quite widespread and relatively in line for financials and non-financials if the comparison is rating-homogeneous. However, if we compare USD-denominated A-rated debt with EUR-denominated BBB-rated debt, the yield gap is still mostly positive but it looks significantly higher for nonfinancials. The graphs 8 and 9 showing the spread (instead of yield) comparison highlight that USD A-rated bonds are now offering almost the same spreads over corresponding government bonds offered by EUR-denominated BBB-rated issuers. In a nutshell, the same spread has between a 1% and a 1.5% positive gap on absolute yield levels for an average rating that is higher by three notches.
The CSPP became operational just a few months ago, but it has already produced dramatic effects on EUR credit markets. The flexibility of the programme, together with its sustained purchase trajectory, which is spread widely on a very large number of corporate bonds that cover all IG ratings, curve buckets and sectors with issuers from many countries, contributed to the effectiveness of the CSPP. If we consider Brexit a sort of first “stress test” on the programme’s trajectory, the CSPP seems to have worked quite well in keeping spreads from soaring as they usually do. It also triggered a de-correlation from risk aversion indicators like equity implied volatility. Stronger, more stable private demand for the asset class is also one of the results of the programme. However, all this has been achieved at the “cost” of a substantial tightening in valuations offered by the asset class, especially by non-financial issuers. For the first time in history, an important percentage of corporate bonds have fallen into negative yield territory, despite still offering a spread over risk-free bonds that is higher than historical troughs. Furthermore, the average spreads of IG non-financial bonds eligible for the CSPP now look tight compared their fair values: the combination of negative rates and low spreads is particularly true for short-duration segments, where HY bonds are almost the only yield suppliers left. The search for yield is likely to move more on other segments and currency-denominated corporate bonds, thanks also to the latest BoE decision to “join the club” on purchasing corporate bonds. Financials in the eurozone market and USD-denominated bonds are among the new potential targets in the not-too-distant future.
The CSPP announcement
The CSPP is the most
If we consider Brexit
The search for yield is likely