Percentage of non performing loans in P2P trading portfolio?

This will vary widely by platforms used and especially Investment strategy


I have 0% on Lendy, Estateguru, Lendix and nearly 0% on Linked Finance. On the other end I have a high percentage on Bondora and I would have a high percentage on Fellow Finance if I would not sell them on the secondary market?
Then there are the platforms with buyback guarantees. Obviously the percentage is always 0%, until Desaster would strike (platform or loan originator inability/failure to honor this 'guarantee')

Now what can you deduct from this? Because I am actually reducing Lendy at the Moment, Lendix is not that interesting (scaling, no autoinvest), while Bondora is one of the platforms delivering me best ROI in my portfolio.
In my view non Performing loans are just part of the deal. What counts is the ROI after defaults, recoveries (and tax if you cannot offset Default losses against interest)

I've been doing secured direct lending (mainly bridges, but some SME loans etc) since 2002. These are typically quite risky lends with yields of 15-18% and LTVs of 50-70% (so equivalent in risk to loans you see on P2P platforms at 10-12%). My rule of thumb is that about 10% of my portfolio will be genuine NPLs (so recovery action required, not a technical default) once the portfolio has reached some sort of equilibrium.

I would expect NPLs on P2P platforms to be somewhat higher (normalizing for equivalent risk say, so RS would not be as high as SS clearly). First, it takes some time for P2P platforms to reach the equilibrium point. As P2P platforms tend to refuse to write down NPLs unless they are absolutely forced to, this risks an eventual day of reckoning where the defaults that should have been recognized in prior year's loan cohorts have to be accepted. So the NPL rate can overshoot the equilibrium for a period to make up for a low NPL rate early on. Second, P2P platforms tend to want to "treat the customer" fairly. "Customer" means borrower, not lender. So they are rather slow going after NPLs and no where near ruthless enough to generate good recoveries. Third, (and rather obviously) the process of writing down NPLs reduces the percentage of NPLs in the book. If I have a £1mm portfolio with 20% in NPLs and I write those NPLs down by 50%, I am left with £900k and £100k of NPLs, just 11% on the book.

I'm of the opinion that P2P platforms should be forced to write down every loan to some degree using an accounting standard somewhat like IFRS9. That would require every loan at the point it is originated to be provided with a credit charge based on expected losses, and then these losses adjusted to reflect future economic conditions and the specifics of the loan (so once it's an NPL, the default probability is 100% and it's all about recovery). If self-select type investors saw these loss provisions in their portfolios, it might stop some of the hysterics when loans default and investors are somehow shocked they actually lose money.

It would come from the lender’s own capital and interest, and remain part of their account as their own “bad debt provision”. So a lender investing £10k in a range of 12% P2P loans, rather than being told they have £10k of loan parts expected to earn 12% interest over the year, would see that maybe 4% of their total was now reserved for projected bad debts, giving a likely 8% net return.

If projected returns worsened (defaults rose) then that 4% might be raised to 5%, 6%, ...

Or, looked at another way, automatically make prudent provision for the almost inevitable losses a P2P portfolio will incur. The point is that the headline 12% is an illusion. How often do we read on here yet another irate diatribe from a naive lender shocked to discover that the reason his P2P loan paid a high rate of interest is that it involved a high risk of default / loss?