Podcast: Play in new window | Download
On the show today I am delighted to be joined by John Goodall CEO and co-founder of Landbay, a member of the P2PFA of leading UK P2P/marketplace lending platforms.
Landbay connect retail and institutional lenders with loans assets in the UK residential buy-to-let mortgage market. They have a well-secured “belt and braces” approach as not only do they have a first charge (mortgage) over an asset (“the building” :-D) but also the debt is serviced by a rental income stream from the tenants.
Stress testing is a very important topic across Finance, though, as all too often one that is poorly understood.
As one of the unique things about finance is that it almost always relates to the future, the unknown future. You buy a pint of beer now you drink it now. You eat a cheese sandwich now you know what it is like now. But invest money – and well, as Yoda put it, “always in motion the future is”.
Stress testing basically amounts to “what would happen if the shit hits the fan”? Would investors lose 1%, 10% or 100%?
In crowdfunding early-stage equity this is an easy calculation – the answer is obviously 100% – you’d lose the lot.
But what about in P2P (debt) in a well-diversified portfolio?
Well as we will discuss its not a trivial question. In essence as one has to decide how much of what type of shit this what fan. As I never cease to try and explain to the non-FS crowd who have lept on the Fintech bandwagon “in FS the devil is always in the detail”.
Not only that but in an uncertain world even so-called experts can’t be trusted. A couple of years ago the ECB (European Central Bank) had European banks “stress test” their portfolios. A few months later many banks had already experienced losses beyond those worse case expectations. So more press test than stress test and yet more FS BS.
John is an ideal person to talk to about this subject as Landbay have just spent time and money getting their portfolio independently stress-tested. So he is well placed to talk to the challenges, value and limitations of the methodology, as well as I hope reveal some of their results.
Topics discussed include: – if P2P was a schoolboy which year would be it in? A memorable soundbite from John on where the market leader would be 🙂
– John’s background having formed and sold a headhunting firm; lessons learned, and phase changes in the growth curve of a company
– what does Stress Testing mean? Doomsday scenario? Bad weather? Really bad weather?
– the disconnect between normal market events and extreme market events
– the extreme case is, by definition, always a disaster
– “single number” answers to stress tests versus “surfaces” depending on key parameters; the sensitivity to this parameter or that and how that feeds back into deeper understanding of the business model (which can be counter-intuitive)
– the big difference between a stress test done internally versus using external specialists
– the real difference in quality of the results depending on the in-depth expertise of those conducting the stress test
– furthermore the important aspect of a data-rich marketplace (such as mortgages) versus less well understood (data-wise) markets eg SME lending
– how to cope with restrictions of being a young P2P and hence not having a long history oneself
– how does one stress test non-credit issues and less-modellable risks such as eg cyber-attack? “Qualitative stress tests”
– Landbay had two motivations – (a) to help inform lenders of the risks that they are taking; (b) re the Landbay provision fund and its adequacy
– the very process of going through the process has informed them as to how to tune their operational model eg underwriting/collections policy
– how one factors in the timescale to sell a property and managing that via rental cover; immediate “firesale” discounts can be (according to historical data) as much as 25-50% of the nominal value); good rental coverage allows one to have a longer sale period and narrow any discount from this factor
– LTV mostly around 75% with some up to 80% (where they would be looking for better rental cover); book average around 63%
– using the Bank of England stress scenario and applying it to the Landbay portfolio they come out with about 0.5-0.6% losses in the portfolio; much more sensitivity to unemployment (via rental coverage) than house price falls
– why are potential stressed losses so small in absolute terms?
– why on earth are regulators not insisting on P2Ps doing external stress tests?
– whether a market in “lax” external stress testers might then arise (after all who would volunteer to hire a “tough” external stress testing company?!)
– Landbay’s £250m Keystone mortgage deal where it has match-funded a book that it is taking over from Aldemore, the challenger bank
– Landbay’s principal focus on retail investors
– Landbay’s product innovations in the market of gret import to P2P and innovative buy-to-let mortgage products
– they will retain their low credit risk approach
And much much more 🙂