Ratesetter lending again?

I look at this more positively - a platform is more likely to attract new lenders (who would buy into the RYI queue) than a platform that has stalled lending.

Your suggestion/request makes no sense, because it prioritises (percentage-wise) the interests of small investors (for whom £1k represents a significant return of capital), in a way not laid down in the original Ts and Cs.
Guaranteed liquidity and capital safety for a small investor are available as FSCS-guaranteed bank accounts, which have lower interest rates. I’m not being snarky here. Honestly, 3% interest is not nearly enough to give up the safety of a classic bank account. But I soon run out of £85k limits. You can get Islamic Finance banks that give 2%-not-interest FSCS capital guaranteed. So, once I’m outside the FSCS limit anyway, that’s when I want some yield to compensate inevitable capital risk, and I look to P2P.

To address the underlying concern that the provision fund might run dry. It might, and there again it might not, it’s an investment risk that we all signed up to. There is a current liquidity shortage, and I can’t see any RYI strategy fairer than queuing requests. Same selling strategy at any broker if you try to sell a lightly traded share after bad news, and same issue. Bear in mind that unlike shares, even the worst-case default ratio in this situation is unlikely to exceed 5-10% over previous predictions. That in itself would be considered a very bad result, and I have no reason to think it would be that bad, but those at the back of the queue are unlikely to lose more than that. This is a liquidity crisis, not a solvency crisis, and it’s nothing like the fiascos over at FS, LY or now FC.

However, it is true that those at the back of the queue may get locked in for up to the underlying loan term, which is 22 months average outstanding. You should get your money back within 2-3yrs worst-case. The term-value Ts and Cs is more my beef with the Access fund in general, rather than the current situation. Which is why I am RYI’ing Access but not 5-yr! I always had in my assumptions that Rolling and then Access could lock up, but I assumed (reasonably) that with an average loan term of 28 months, my term risk was 12-16months. I have been very unhappy to discover that the *average outstanding* term of loans in my personal Access loan portfolio is 33 months. One of the two major parameters for any investment spreadsheet (more important than interest rate) is over double what the company told me. There seem to be “reasons”, but I don’t really care. It’s this fact that makes Access simply un-investable for me, irrespective of market conditions or interest rate. The original 1-yr and 5-yr markets are unaffected by this “term bug”, and would remain investable because they do what they say. For me, Rolling was the runoff from 5yr, and I understood Rolling correctly, but some “minor” changes caused major impacts to Access term structure, and I was too lazy to notice.

I don’t have direct figures to predict historical consumer default rates, but I do have these observations:
A*/A rated consumers have ~0.5% default rates in normal times. That’s “sort of” equivalent to BBB corporate bonds, considered bottom of investment grade or top of junk. That risk mapping is also “sort of” justified by noting that the rates we are getting off RS were similar to the high end of high-yield bond funds dealing in investment grade while avoiding true junk bond funds. My highest yield bond funds (which I got out of with some loss but avoiding the worst) crashed by 20% before partially recovering to stand 12% below pre-Covid. So, the markets are pricing in a 12% default rate excess on that. Those are insane default rates by any standards of the current generation, but probably realistic in the circumstances. Nevertheless, it’s *only* 12% loss of capital.

The paper below has a graph showing that over the past 150 years, US corporate bond defaults have hit 5% a few times, for events like world wars, and 10% just once during the popping of the railroad bubble in the 1870s. 2008 barely flickers to 3% in that perspective.
www.nber.org/papers/w15848.pdf
I know that it’s a different asset class, but ultimately consumers pay their debts when companies can afford to pay their salaries.

RS consumer borrowers below A*/A should have had their interest rate adjusted appropriately; the evidence of whether RS have the technical expertise to map that correctly, is that their estimates of default rates pre-covid have tracked fairly close to the mark.