Europe's second peer-to-peer securitisation closes
Six months after the closing of the S-BOLT 2016 transaction, Zopa closed last week with final details of its debut peer-to-peer securitisation ‘Marketplace Originated Consumer Assets 2016-1 (MOCA 2016-1).
The capital structure consists of multi-class notes rated by Fitch Ratings and Moody’s. Moody’s pre-sale report has the Class A notes sized at £114m and rated Aa3. They carry a coupon of 1 month LIBOR plus 1.45% p.a. and a par issue price. For comparison, S-BOLT 2016-1 – which had a similar rating from Moody’s but a ‘BBB’ rating from Standard & Poor’s – was priced at 1 month LIBOR plus 2.90%. Fitch has also issued final ratings of AA- on the same notes. According to Fitch’s pre-sale report, the firm has placed a ‘AA’ ratings cap for this particular issuance as MOCA 2016-1 is Fitch’s first sortie into EMEA securitised assets originated on a ‘market-place platform’. It should be noted that this is the highest level cap Fitch has reached globally on this relatively new asset class on account of Zopa’s quality of data. Notwithstanding this ratings cap, Zopa opted to apply for a rating one notch beneath the ratings cap at ‘AA-‘. According to Fitch, this ‘notched’ rating was based on the cash flow model rather than restricted by a credit dependent counterparty.
We understand that EIB and EIF did not participate in this transaction as the underlying assets were ‘consumer’ in nature as opposed to ‘SME receivables’.
Interestingly, Standard & Poor’s chose not to rate any of the notes despite its involvement in the S-BOLT 2016-1. It won’t be possible to check whether Standard & Poor’s was originally approached by Zopa until the company publishes its regulatory disclosures later this year but one might speculate that the ‘AA-‘ rating on the Class A notes might have been a stretch for a rating agency which applied a ‘BBB’ ratings cap for similar assets for Funding Circle earlier this year.
Fitch’s concern (and hence the ratings’ cap) is derived from Zopa’s possible misalignment of ‘economic interest’ in the assets once securitised. This concern is not restricted to Zopa, but more a feature of the marketplace originator model where the bulk of revenue is front end origination and servicing fees (1 per cent in aggregate) derived from borrowers. Moreover, as with other lenders of this type in Europe, loss numbers have been derived in a relatively benign economic environment and – with the exception of originations between 2013 and 2016 – statistically small data sets. Furthermore, Fitch cites Zopa’s recent shift into riskier parts of the lending spectrum (lower average Equifax credit scores, longer terms and debt consolidation at 34% of the portfolio) as a development that makes it difficult to ‘look forward’ as to how the portfolio will behave under stressed conditions.
The other notes used for the portfolio purchase consist of £7.5m of Class B (A/A2) with a coupon of 1 month LIBOR plus 2.90% p.a., £7.5m of Class C (BBB+/Baa2) with a coupon of 1 month LIBOR plus 4.00% p.a. and Class D (BB/Ba3) with a coupon of 1 month LIBOR plus 7.00% p.a. In addition there are £12 million of unrated Class Z notes. All notes are subject to an optional issuer 10 per cent clean-up call provision. As with S-BOLT 2016-1, the ratings of these notes are for the ‘ultimate payment of interest and repayment of principal’ – i.e. do not address ‘timeliness’ of interest. Without seeing detailed cash flow models, it is impossible to say how this type of rating has arisen. One clue might be the liquidity reserve which is sized to cover only the most senior notes outstanding at any point in time. Both reserve funds and up-front costs are funded by an unrated subordinated loan.
Credit enhancement is provided in the first instance from (i) excess spread - beyond the senior costs and interest costs on the Class A to D (calculated to be 3.9% on closing) - crediting losses through the various principal deficiency ledgers notes (ii) subordination of the other notes and (iii) the 1 per cent Cash Reserve Fund (only 50 per cent funded on closing).
The assets were originated through various marketing channels and aggregators including price comparison websites. The portfolio is static in that there is no revolving period. Moreover, all the loans have a maximum term of five years. The loans were represented by the seller, P2P Global, to be free of defaults and delinquencies as at the pool cut-off date. P2P Global has also retained a minimum of 5% of the assets through its holding of the Class Z notes and the provision of the subordinated loan. The fund also holds 100% of the Class D Notes. The underlying assets are unsecured fixed rate loans made to consumers. In order to hedge the excess spread against erosion if interest rate rises, the notes are hedged through an interest rate cap provided by BNP Paribas (A+/A1) with a strike rate of 2%. The swap notional follows the scheduled amortisation of the underlying assets ignoring prepayments and defaults.
Moody’s reviewed the vintage data and observed cumulative default rate of 2.5% against recoveries of 39%. Notwithstanding their observations, they modelled a ‘base case’ cumulative default rate of 7% (higher than peer group) and recoveries of 5% to take account of the relative lack of historical data and that 40% of the pool has been sourced from B and C1 internal credit scores (loans with credit scores of D and E loans were excluded from this transaction). Unlike S-BOLT 2016-1, where Moody’s used a normal inverse default distribution, the agency has used the lognormal approach - more standard for these types of assets (consumer versus SME receivables). In comparison, Fitch used a modelled base case cumulative default rate of 4.7% (coupled with a relatively high AA- multiple of 5.5x) together with stress AA- recoveries of 9.92%.
HSBC (AA-/Aa2) acts both as Account Bank and manages the notes’ liquidity in the event there is a servicer termination event.