Riveting year ahead for covered bonds in QE-warped market
Covered bonds performed well in 2019, but yields finished in negative territory and spreads ended at their tightest for the year. The implication is that, despite higher than expected ECB covered bond purchases and a renewal of its ultra-cheap TLTRO facility, investors will struggle to match 2019’s returns in 2020, writes Bill Thornhill.
Aggressive covered bond buying by the European Central Bank since November 2019 and speculation of a further 10bp cut in its deposit rate boosted sentiment into the final few weeks of 2019 leaving the market on a high. These key factors will continue to provide support at the beginning of 2020, suggesting spreads could still tighten and yields may yet fall further.
But, with the ECB running out of monetary policy ammunition and a supply splurge likely in the first quarter, a sustained performance matching that seen in 2019 seems highly unlikely. Indeed, it won’t take much for a setback to be triggered.
The combination of central bank and bank treasury buying should mean there is an almost “infinite amount of cash chasing assets, as long as covered bond spreads stay noticeably positive against three month Euribor,” says Florian Eichert, head of covered bond and SSA research at Crédit Agricole CIB in Frankfurt. However, the moment covered bond spreads get close to the spread equivalent of the ECB’s deposit rate, “banks will quickly switch off buying”. And since it’s highly unlikely any other type of investor would step at that point, a correction would be expected to follow.
Despite a 35bp rise in outright 10 year Bund yields in the final four months of 2019, the market finished around 50bp below where it started the year, leaving almost 90% of euro covered bonds with negative yields. At the same time spreads tightened sharply, with Caffil printing a 10 year in November more than 20bp inside where Crédit Agricole and Crédit Mutuel CIC issued 10 year deals in January.
The steep move down in yields combined with a substantial spread tightening provided ideal investment conditions causing the Markit iBoxx covered bond Total Return Index to jump from zero in 2018 to 3.65% in the 12 months to early November 2019. “I’m struggling to see how investors will repeat 2019’s strong performance in 2020,” says Zuzana Stockton, head of covered bond EMEA syndicate at Deutsche Bank in Frankfurt.
Conditions became almost too good — when yields fell to their decade-low in early September 2019 borrowers increasingly avoided the covered bond market as deal execution became less certain. Deeply negative yielding covered bonds issued at that time by Berlin Hyp, at minus 0.588%, and later by MünchenerHyp emerged just when rates were at their low point.
With around 30 to 40 investors on board, these issues and others that came around the same time attracted roughly half the average number of buyers typically expected. Worse still, some deals, such as those issued by Deutsche Pfandbriefbank and Hypo Tirol, were not even subscribed. Fearing the worst, borrowers opted to move down the capital structure and issue higher yielding senior preferred deals.
However, as outright yields subsequently corrected higher, banks were successfully able to issue long-dated covered bonds with only slightly negative yields. With a reoffer yield of minus 0.053% Caisse de Refinancement de l’Habitat brought in 85 investors for its €1bn 10 year issued on October 1, and with a reoffer yield of minus 0.069%, BPCE hauled in 84 buyers for its €1.25bn seven year issued on October 30.
Then, in early November, the ECB stepped up purchases, raising the size of its order from 5% of a deal’s size to 40%. The ECB’s eightfold increase in buying was well beyond expectations and gave strong a boost to sentiment. But in the event outright yields head lower and spreads continue to tighten the positive mood could quickly change.
“I think it is likely spreads will grind tighter from here,” says Armin Peter, head of EMEA syndicate at UBS in London. “But investors are sure to step back from books in the event subscription ratios fall to 1.2 times or less and the ECB gets close to a full allocation.”
Deutsche’s Stockton warns that market fatigue is often seen towards the tail-end of the first supply wave in January, saying “there may well be a few bumps of volatility through the year.” She doesn’t expect covered bonds to post anything close the near 20bp widening seen in the second half of 2018.
Apart from the scale of supply, negative rates and tightening spreads, covered bond bankers will also have to contend with relative value considerations. Covered bonds offer a good enough spread against other high quality liquid assets, but this margin has come down to the point where interest could quickly disappear. “The real pinch point would come if covered bonds lose relative value at the same time as yields becoming deeply negative,” cautions Peter.
At the start of 2019, Pfandbriefe were delivering a pick-up of 55bp-60bp over Bunds, but the differential fell to around half that level by November. At around 35bp over Bunds Pfandbriefe are still attractive, but getting closer to levels where investors have previously deserted the market. Two years ago, when the Pfandbrief pick-up was only 15bp-20bp over Bunds and German Pfandbriefe traded inside agencies such as KfW, covered bonds were sold.
For this reason Matthias Melms, head of covered bond and SSA research at NordLB in Hanover, is “bearish on covered bonds compared to SSAs, especially at the long end”.
All eyes on the ECB
Conversely, scope for the relative performance of covered bonds against other asset classes could improve in the unlikely but possible event the ECB steps up purchases to an even higher level. The ECB is limited to buying a maximum of 33% of any sovereign, supra or agency bond, which is likely to become insufficient to meet its overall asset purchase target of €20bn a month.
As such, at some point in the first few months it will probably need to raise this limit. If not, it will be obliged to look elsewhere, probably by increasing purchases among all its private sector progammes that include covered bonds, securitizations and corporate bonds.
“Should the ECB not manage to raise the ISIN limit on the Public Sector Purchase Programme, it might be forced to step up private sector purchases and [covered bond] spreads could be squeezed back toward the mid-swaps minus 25bp to minus 30bp levels,” says Eichert.
The ECB is able to buy up to 70% of any eligible covered bond so it could easily step up purchases, returning to the peak buying levels seen in 2017. “The ECB have to fulfil their full quota and, if there’s no new issue supply they’ll buy in the secondary, so there’s likely to be a squeeze of one asset class versus the other,” says Stockton.
In such circumstances, the Eurosystem would need to buy covered bonds in much larger size, in which case the market could potentially trade inside agencies such as KfW and tighten even further than 15bp-20bp against Bunds. While this is possible, Eichert doesn’t believe it’s likely.
Added to this quagmire of considerations, covered bond practitioners must take account of the ECB’s third Targeted Longer-Term Refinancing Operation (TLTRO). This three year funding should in theory be supportive for covered bonds because it subtracts from banks’ overall funding needs.
“After maxing out on the TLTRO and fulfilling relatively expensive MREL funding, I expect issuers to turn to covered bonds,” says Peter, who thinks this should help lower their overall average cost of funding.
Analysts in Barclays’ research team forecast a much bigger take-up in the December TLTRO of around €300bn, as funding in this operation could benefit from a further potential 10bp cut in the ECB’s deposit rate to minus 60bp.
Barclays’ analysts say the third TLTRO has been transformed by the ECB from a generous backstop to an attractive funding tool and, at a gross €650bn, they expect almost the entire amount borrowed under the second TLTRO to be rolled into the third.
Italian and Spanish banks are by far the largest TLTRO users with BBVA research analysts estimating the volume of retained covered bonds used by Spanish and Italian banks for ECB liquidity operations at €65bn and €62bn respectively.
Critically however, this TLTRO borrowing is not expected to increase or fall from 2019, which means there will be no net impact on the expected supply of covered bonds in 2020. Eichert says the TLTRO is there merely as a contingency funding tool and he doesn’t believe it will push down covered bond issuance in 2020.
In fact, given heavy covered bond redemptions next year, a number of new entrants to the market and solid mortgage asset production both he and Stockton are optimistic that 2020’s covered bond supply will increase from 2019.