Following a quiet 2020, the UK residential mortgage backed securities (RMBS) primary market sprung to life in the first month of 2021. This brings some focus to an area of the bond market that I think looks cheap – not something you hear said about bonds every day. We even saw the first ESG UK RMBS issued (Gemgarto) just last week.
The exceptionally low interest rates offered by the Bank of England to the high street lenders explains the absence of issuance from those institutions last year. That has been the case so far this year too. All the deals we have seen have been in the non-conforming (NC) or buy-to-let (BTL) sectors, where originators don’t have access to the Bank of England facilities.
As one might deduce, NC and BTL mortgages are a riskier proposition than investing in prime mortgages (loans written by the high street banks). But how much riskier, and are we being paid for that additional risk? Currently I think we are.
When we analyse an asset backed security there are three principal conditions that must be satisfied before we buy.
1. Are we comfortable with the collateral?
As the Great Financial Crisis proved, no amount of financial engineering can make up for bad collateral. We need to be at least reasonably confident that the loans on the underlying collateral, be that houses or cars, will generate the cash flows we expected. And, if for some reason they don’t, that we will be insulated from losses when that collateral is sold.
2. Will the structure of the deal protect us?
Generally we invest in the AAA tranches of these deals, which would be the last to take losses if the future expected cashflows weren’t to materialise. Our team of analysts run deep stress tests on all tranches, but essentially we want to be confident that there are enough investors below us in the capital stack to absorb any losses that may occur. We consider other metrics – prepayment rates, coupon step ups, arrears and more – but I think credit enhancement (the proportion of the deal that can take a loss before your bonds do) is by far the most important factor in avoiding a loss to capital.
As a very quick aside: no AAA tranche of any European ABS has ever taken a capital loss (assuming it were held to maturity) since the inception of the market in the early 2000s. The structures worked in Europe. Losses were suffered on AAA tranches of US RMBS but this was primarily due to the collateral (sub-prime, non-recourse, adjustable rate mortgages) being so bad that the losses ate through the whole stack.
3. Are the bonds good value?
We assess this versus other similar securities and against corporate bonds more broadly. As you can see below, secondary market credit spreads on AAA-rated NC and BTL bonds are trading in line with BBB corporate spreads – and much wider than corporate bonds with a similar rating (we’ve used AAs in the chart below due to the scarcity of AAA corporates for comparison).
Unsurprisingly this, coupled with the lack of prime RMBS issuance (due to the BoE lending facilities), has caused a strong technical in the wider market with demand outstripping supply. In the last week or so we’ve seen new issues tighten 10-15bps in the secondary market and issuers have been able to use this strength to offer tighter spreads on subsequent deals. Even with this tightening we think these assets still offer better value than most traditional short dated corporate bonds.
While I’m singing the praises of the asset class, I should also mention that these instruments are typically floating rate notes (FRNs). This means that they are a good hedge to any increase in interest rates, as coupons are reset quarterly with reference to SONIA. As interest rates rise so do your coupons, helping to preserve the capital value of the bonds.
Another interesting feature of FRNs is that, when short term interest rates are negative, there is no mechanism for the lender to pay the coupons to the borrower. It doesn’t matter how negative SONIA goes – the minimum coupon the holder of an FRN will receive is zero. As strange as it sounds, the more negative rates go, the more attractive an income of 0% becomes and this will be reflected in the price of the bonds. In other words: at the zero bound, FRNs turn in to fixed rate securities as yields are falling. That is a very helpful and attractive asymmetry for bond investors to take advantage of.
We are of course very fortunate to have a large and experienced ABS analyst team to help us navigate the sometimes complex collateral and structures of these deals. But with this in mind, I think selling BBB unsecured risk into AAA secured for a pick-up in credit spread, while lowering interest rate risk, makes a lot of sense. Especially right now, when many investors have the desire to reduce the risk in their portfolio more broadly.