Covered bond issuers face up to life without ECB handouts

The European Central Bank is finally on its way back to normal monetary policy, having cut purchases in the primary covered bond market from 50% of a deal’s size to 10% and eventually to 0% in December.

But its journey is unlikely to be straightforward. Indeed, no sooner will it have switched off its Covered Bond Purchase programme, than it might feel compelled to offer another round of liquidity, in the form of a third Targeted Long Term Refinancing Operation (TLTRO), to ease peripheral European banks over the looming refinancing cliff edge they face in 2019.

For these reasons many bank funding officials in Europe and their investment bank advisers are relaxed about the outlook for 2019. Covered bond spreads have widened by 15bp or more over the last 12 months and most new issues from many jurisdictions now offer a positive pick-up to mid-swaps.

A return over mid-swaps is widely considered sufficient to ensure crucial demand from the bank investor community which buys bonds to meet its liquidity coverage ratios and form the largest component of covered bond demand.

“Spreads are already at a level where LCR portfolios have been stepping in,” says Armin Peter, global head of syndicate at UBS in London. He expects a small drift wider in spreads at the start of 2019 but it should be “relatively well contained”.

Even so, the ECB has crowded out private demand, and while more investors are returning, the diversity and depth of demand is still far from the levels seen five years ago.

On average around 60 accounts participated in covered bonds in the second half of 2018 compared to five years ago, according to Commerzbank analysts. Many of the biggest investors, typically asset managers, have substantially cut back or withdrawn from the market as spreads are still too tight.

However, an investor survey by Société Générale research at the end of 2018 suggested that almost half of respondents plan to increase their allocation to covered bonds in 2019, up from 18% who planned an increase in their allocation in 2018.  Those aiming to keep covered bond investment unchanged fell from 67% in 2018 to 36% in 2019.

“The tone is constructive,” says SG’s covered bond research director, Cristina Costa but “investors remain cautious for 2019”, citing expectations that the ECB will taper its quantitative easing programme and prospects for a rise in volatility due to “negative headline risk”. 

Even though the ECB’s withdrawal has arguably been priced in, and scope for additional harmful price shocks should be limited, downside risks for spreads “continue to predominate” says Ted Packmohr, head of covered bond and financials research at Commerzbank. Despite the ECB cutting covered bond purchases in the primary market, Packmohr says that it recently compensated by stepping up secondary market purchases. At €3.2bn, October’s gross monthly purchases were down little more than 20% from September, which was a peak month. Gross buying in October was also not vastly different from the start of 2018 when the ECB’s involvement in the primary market was as much as five times higher.

Although the ECB is committed to reinvesting its €261bn covered bond portfolio (CBPP3), it will no longer grow it from January 2019. That means it will only have €22bn to invest in the covered bond market in 2019. For that to happen the monthly average purchasing rate will fall to €1.8bn.

The ECB does have some flexibility to raise or lower purchases in line with supply. But, it is likely to deploy more cash at the start of the year when issuers try to get ahead with funding and take advantage of demand as investors generally have more cash to put to work at the beginning of the year. 

Even so, the surge in issuance at the start of 2019 is likely to exceed the ECB’s demand, suggesting further spread widening is likely. In January the ECB will need to reinvest covered bond redemptions of €2.8bn from the CBPP3 portfolio, and in February it will have a further €2.4bn to buy. But by March the ECB’s purchasing rate will start to slip and will fall to as little as €910m by May — less than one third of what it bought in October 2018. The reduction of CBPP3 purchases is therefore “by no means complete,” says Packmohr.

Assuming the ECB reinvests all covered bond redemptions, and core European issuers replace all their redemptions with new financing, prospective supply would exceed the ECB’s purchase programme demand by a relatively modest €1.31bn in January. Some €4.1bn of core European covered bonds are due to be redeemed, and potentially be replaced with new issuance, compared with a relatively high €2.8bn of purchase programme redemptions (see chart).

However, in February this mismatch between potential supply and ECB demand rises sharply to €6.24bn. The difference between expected supply and potential ECB demand peaks in April at €7.24bn. Over the rest of the year the monthly mismatch is no higher than the €3.7bn.

p1-52TLTRO III?

The implication is that spreads will widen more between February and April, when prospective issuance is most likely to exceed ECB demand. This picture could change if expectations of another round of cheap ECB financing continue to build.

Many bankers believe that the ECB will announce a third version of its TLTROs before June. This could give European banks, particularly those in the periphery, a cheaper alternative source of funding. Even so, the mismatch between expected core European supply and ECB demand means that spreads have the potential to widen by as much as they did in 2018.

In Société Générale’s investor survey, around 70% of respondents expect core European covered bond spreads to widen by a further 10bp in 2019. Around half thought peripheral covered bonds could widen by up to 20bp, and potentially a lot more depending on how the ECB formulates its third TLTRO.

European banks borrowed €740bn over four tranches of the second TLTRO, the majority (62%) by banks in Europe’s periphery. Although the auctions officially ended in March 2017, two members of the ECB’s governing council recently discussed the possibility of renewing liquidity support.

This has fuelled speculation of a third TLTRO in the first half of 2019, when banks face a refinancing cliff. In June the €399bn of four year funding borrowed under the first tranche of TLTRO II will no longer be considered stable funding.

Under the terms the Capital Requirements Regulation’s net stable funding ratio (NSFR), which banks have to meet for the first time in 2019, the residual maturity of all funding must exceed one year. Since this tranche matures in June 2020, it will not meet this requirement from June 2019 and therefore banks will be obliged to consider refinancing it.

Bank treasurers will therefore need to consider the trade-off between their NSFR ratio and liquidity coverage ratio, and some investment bankers, such as Chris Agathangelou, managing director and head of fixed income syndicate at NatWest Markets, expect the covered bond market to “play a very significant refinancing role in this respect”.

Not all of the cash borrowed under TLTRO II was used for funding purposes as many issuers took advantage of the cheap financing to buy higher yielding ‘risk-free’ government bonds and earn a return. As such, it is likely that only a portion of the funding will actually need to be refinanced.

Agustin Martin, head of European credit research, public sector and covered bonds at BBVA, believes TLTRO II refinancing will trigger some €70bn of additional covered bond issuance from banks in Europe’s periphery which will be spread over the next two years. But, he says, much depends on timing. 

The scale of TLTRO II refinancing will pivot on when the ECB announces a new facility. “The longer the uncertainty on future ECB support, the more public supply could hit the market,” says Citi’s covered bond analyst, Michael Spies. If funding costs are too high then borrowers will have to shrink their balance sheets, he says. However, the earlier a third TLTRO is announced, the less public supply is expected.

Issuers with good market access, typically from Europe’s core regions, are less likely to wait for another TLTRO because, if the ECB gauges it correctly, the next operation is likely to be more expensive for them relative to market funding. For those borrowers still short of meeting their funding need for the minimum requirement for own funds and eligible liabilities (MREL), non-preferred senior issuance will be the priority.

However, where MREL issuance is on schedule, borrowers will weigh up the difference in cost of funding between covered bonds, and preferred senior. In March 2018 the spread between the iBoxx preferred senior and covered bond indices was 8bp, but by late September it had moved out to 37bp. As this spread widened, covered bond funding became more attractive, and if this trend were to continue through 2019, covered bond supply would be boosted.

But that increase in issuance volume may result in the market underperforming relative to senior preferred and the spread differential between the asset classes would compress again. In such circumstances “it is plausible to assume that banks would shift from covered to unsecured issuance in 2019,” says Spies.

Otherwise, banks could access other covered bond markets such as dollars and sterling. Spreads in these markets have remained relatively stable, “as a result of either chronic undersupply, favourable basis movements or shifting investor trends,” says Sameer Rehman, director of syndicate at TD Securities. The recent burst of dollar covered bond supply shows that this new trend “has already started to manifest,” he notes.

Sterling covered bonds had “an unbelievably strong 2018,” and Rehman predicts “another good year”.  Gilt supply is likely to fall and sterling swaps have steadily widened, providing a good buffer for bank supply. Moreover, the sterling market is well prepared for the post-Libor world, with recent deals showing that Sonia is now the de-facto reference rate.