Deals from Virgin Money and Paragon Bank last week laid bare the absurdity of the uneven playing field between RMBS and mortgage-backed covered bonds.
Virgin’s RMBS is backed by prime owner-occupied mortgages and carried the ‘Simple, Transparent and Standardised’ securitization hallmark. It came from Virgin’s Lanark Master Trust, which has over £2bn of outstanding paper, with issuance dating back to 2007.
Paragon, meanwhile, was making its covered bond debut, with a £500m deal, backed by about £1bn of buy-to-let mortgages, a variety of housing debt generally considered to carry more risk.
There are clear reasons to prefer Lanark, even with the covered bond also offering recourse to Paragon, rated BBB+ by Fitch, rather than simply to the pool of underlying assets
Investors agreed, with Lanark landing at 47bp over Sonia and Paragon’s deal at 60bp, but mystifyingly the regulations take a different view. By hitting benchmark size, Paragon achieved eligibility for the highest quality level in the Liquidity Coverage Ratio for its trade, while Lanark languishes at Level 2B. The higher the level, the cheaper the bonds are to hold for certain investors.
Perhaps one justification might be that there is a bigger buyer base for covered bonds and therefore they are more liquid, but the track record doesn’t back up that claim.
For most of 2024, sources suggested issuers were considering RMBS as covered bond investors filled up their credit lines. In addition, securitization proved its liquidity in the liability-driven investment (LDI) crisis of late 2022. There were €6.9bn of bids won in competition (BWICs) in just three weeks and although spreads widened, the vast majority of bonds were traded — in a time of panic, plenty of paper changed hands.
All of this is in the context of the regulations being completely slanted against RMBS. If the playing field were level, the liquidity of securitizations would be even greater.
It makes a mockery of the STS label. If regulators are going to distinguish securitizations that meet a certain standard, essentially regulating bad practice out of that section of the market, then they need to reflect that properly in how they treat those deals.
That Paragon has gone down the covered bond path must be a wake up call to securitization's regulators. It was a regular RMBS issuer until 2019, and would have been welcomed back, likely printing around 70bp over Sonia in the current market.
Clearly for Paragon there is some benefit in appealing to new investors, but the main reason to issue a covered bond was likely that it could print 10bp tighter for far less work to get a deal away. Those economics mean others might well follow.
Yet from an issuer’s perspective, RMBS is the more efficient structure for buy-to-let assets. Paragon’s cover pool was nearly double the size of its bond, while Shawbrook needed just 13.25% of credit enhancement for a triple-A rating on its most recent Lanebrook RMBS.
All the disclosure required for securitization means regulators should also prefer buy-to-let mortgages to be funded in the RMBS market, by investors that are used to underwriting such loans.
STS buy-to-let RMBS has been issued by two banks, Shawbrook and Charter Court, but it has not caught on because of the double punishment of extra work and poor LCR treatment leading to minimal pricing benefit.
It does not make sense to drive buy-to-let mortgages into the covered bond market. It is time to make things more proportionate between RMBS and covered bonds and there will never be a better example to prove it than Paragon's covered bond.