Covered bond market participants have been struggling to make sense of the announcement of a third covered bond purchase programme by the ECB yesterday (Thursday) afternoon, with Caffil potentially the first issuer to test the new environment after announcing a mandate today.

Caisse Française de Financement Local (Caffil) has mandated BNP Paribas, Crédit Agricole, Deutsche and SG as leads for a five year public sector obligations foncières benchmark for next week, subject to market conditions, and a deal for another French issuer has also been mentioned. Abbey has meanwhile announced a roadshow for Monday and Tuesday, with Commerzbank, Natixis, Santander and UniCredit involved.

A syndicate official on the Caffil mandate said that it had been in the pipeline for a couple of weeks and that it was coincidental that the announcement came just after the European Central Bank move.

He said that a July 2019 Caffil benchmark was trading at 2bp/5bp through mid-swaps. The French issuer’s last benchmark activity was a Eu500m tap of an October 2028 issue at 37bp over mid-swaps in April.

Asked whether the ECB move would help the deal, the syndicate official said: “Yes, of course.”

But in stark contrast to the glee with which the announcement of CBPP1 was received – CBPP2 having already met with a more measured response – reactions to yesterday’s announcement were generally negative.

“It’s not good news for us,” said a covered bond syndicate banker. “Nobody will go short this market, so you will have no offers. Most are already bid only.

“Personally I don’t see why they are doing that,” he added.

There was apparent unanimity on the positive effect on spreads going forward.

Peripheral names were said to have already tightened 10bp yesterday afternoon and a further 5bp-10bp more this (Friday) morning, according to the syndicate banker, with the widest core spreads more stable after tightening 5bp yesterday and the tightest core spreads a few basis points tighter.

“This is just a start,” said Joost Beaumont, fixed income strategist at ABN Amro, “and given negative net supply and poor liquidity, a further sharp tightening is on the cards, although this will not be based on actual flows.”

The syndicate banker cited the example of an investor this morning looking to buy core names from across jurisdictions this morning in anticipation of such tightening.

However, almost no details of the covered bond purchase programme (and the ABS programme) were announced yesterday, with these due after the next monthly meeting of the ECB’s governing council, on 2 October.

The most important missing “detail” is the size of CBPP3, and, again, also the ABS programme. A Eu500bn figure cited in a Reuters report yesterday morning in relation to ABS purchases ahead of the official announcement yesterday afternoon has been widely discussed, but this was not confirmed by the ECB, which gave no figures.

Market participants have therefore sought to anticipate what the ECB might announce by looking at past experience and the implications of what Draghi did announce.

An investor – not dedicated to covered bonds – noted that Draghi sees the various current programmes (TLTRO, ABS purchase programme and CBPP3) as roughly Eu1tr in size, and the investor said: “We take from that that the latter two could achieve a maximum size of c. Eu400bn, but with the majority of that coming from CBPP3 given the relative sizes of the markets.”

This would imply CBPP3 totalling more than Eu200bn – far more than the Eu60bn of CBPP1 or the Eu40bn earmarked for CBPP2, let alone the Eu16bn spent.

Market participants have struggled to understand how, in light of the already strongly negative net supply of the benchmark covered bond market, the ECB can achieve anything near what it might seek to achieve, even if it targets an amount much more in line with the two previous programmes – particularly given its CBPP2 underspend.

“The size of CBPP3 is likely to be around Eu40bn-Eu50bn, assuming that the covered bonds on the ECB’s balance sheet might go back to 2012 dimensions (it bought a total of Eu76bn of covered bonds in the CBPP1&2),” said Beaumont at ABN Amro. “However, we think that it will be a big challenge for the central bank to really buy such an amount, given negative net supply, and also bearing in mind that the ECB only bought less than half of the targeted amount under CBPP2.

Many market participants have turned to the previous programmes for clues as to how the ECB might conduct CBPP3 after the modalities are announced a month hence. This could suggest purchases in the primary and secondary markets, possible minimum ratings and issue sizes, and compliance with UCITS or “similar safeguards”.

However, Bernd Volk, head of covered bond research at Deutsche Bank, suggested that such rules could be dispensed with and a much wider range of securities considered eligible for CBPP3.

“If they really want to bring back their balance sheet to 2012 levels, they need to significantly change their approach compared with CBPP1 and CBPP2,” he said, “and may not even be able to afford to stick to UCITS-compliant covered bonds.”

The ECB could, for example, include bonds collateralised by SME loans issued under a Banque de France scheme launched earlier this year, said Volk, although it is questionable if the ECB wants to do that given it could blur the boundaries of what it considers a covered bond, he added.

Draghi highlighted at yesterday’s press conference that CBPP3 would be different from its predecessors.

“Then we have the starting again of a covered bond purchase programme,” he said as part of a response to a question. “As you all know, it’s not a new thing, but the purposes of this programme are very different from the previous programmes.”

Several market participants have also suggested that in light of the volume of euro benchmark covered bonds outstanding issued by euro-zone banks – Eu588bn in the iBoxx euro benchmark index, according to ABN Amro’s Beaumont – retained issuance, whether currently repo-ed or not, could be targeted by the ECB.

“The ECB needs to either pay up very aggressively or directly target retained covered bonds (with the latter potentially reducing ECB repo),” said Deutsche’s Volk. “We estimate the volume of retained covered bonds to be at least Eu150bn.

“However, also in the case of retained covered bonds, it seems unclear if the banks are actually willing to sell as liquidity may not be needed and selling does not provide any capital relief.”

Meanwhile, Jeroen van den Broek, head of debt strategy and research at ING, suggested that the tightening of spreads resulting from CBPP3 could result in peripheral banks relying less on ECB funding and tapping the public covered bond market – thereby making ING’s Eu125bn covered bond forecast for 2015 seem “rather conservative”, even if the availability of mortgage assets and asset encumbrance issues could constrain volumes.

“Our reading for the surprise decision to buy covered is that it creates the opportunity for the peripheral banks to issue covered when retained collateral from the expiring LTRO is freed,” said van den Broek. “As we have indicated before, the peripheral banks can draw Eu150bn under the TLTRO, but are still drawn to the tune of Eu350bn under the LTRO.

“The Eu200bn excess was likely to be refinanced under the shorter term ECB programmes (MRO) but now with covered spreads across the board compressing, it could be refinanced in the open market, switching the old retained deals to new covered supply in 2015.”