Dan Rasmussen is the founder and portfolio manager of Verdad, a firm that invest in leveraged small value stocks in the public markets; he is a super smart dude and I tend to read / listen when I am lucky enough to find anything about him. Unfortunately broke chaps like me cannot invest in his fund, so the best return I can get from him is via learning.
In a recent article, Rasmussen referred to the phenomenon of chasing credit from blue-chip Private Equity deals as “Fool’s Yield”. Everyone seems to know these deals are junk — whether it’s the banks who aren’t funding them, or the ratings agencies who regularly label them non-investment grade — bar institutional investors. Rasmussen pointed to some great data on credit which just goes to show how poor the returns are if you decide to lend money at a seemingly usurious yield. As an example, he took the following data is from Lending Club; take a look at how poorly its credit has performed, relative to the yields offered:
Since 2007 the highest yielding credit — rating FG — has offered an average interest rate of 26.09 per cent, but only returned 3.11 per cent to investors. Indeed, despite lending at an average rate of 13 per cent, the average return from a Lending Club loan has been just 5.4 per cent. Higher interest rates do not seem to compensate for the higher risk and defaults have a huge impact on the return investors take home.
It was great to finally have the chance to look at this type of analysis on p2p loans, then this question popped in my mind: what about European platforms? I always avoided to invest in the most risky rating categories, maybe it is a good time to check if there is a rational support behind my behaviour.
Bondora
Bondora was the first place to check, since at least they are good at something: they provide data about their loans’ performance. The way data is presented is not 100% transparent, so it takes a little experience to read between the lines and find the truth. In following tables, white and blue shaded % means Bondora issued a relevant number of loans (white means more then blue), while orange (it is orange? my wife always jokes that I am colourblind…anyway you get it) means not relevant. I took data only up to 2017 because Bondora has a weird way to report defaults: if a loan defaults on the first payment, Bondora does not consider 100% of that loan as defaulted but only the part accrued up to the point; this means that more recent figures overstate return and inflates numbers. You should concentrate on the ‘Actual’ returns, ‘Target’ are reported just to confuse you (or embarrass Bondora, you choose).
Finland
Estonia
Spain
As you can see, returns are quite random, there is no actual premium in lower ratings. In Estonia, the only country where Bondora experiences persistent positive returns, E and F ratings seems to yielded more than A and B but the over-performance is not statistically significant. I am sure Bondora stans will use this data point as a proof that the overall rating system works, here a picture of one of them:
Fellow Finance
Here I reorganised their data, ordered by loan rating from 1 star to 5 star, and used the same colours as Bondora, light blue for high number of loans and orange for low:
Fellow Finance reports only the loan volumes issued in the last 30 days, so the colour distinction is quite arbitrary. As you can see, only for Finnish business loans higher risk means higher returns, for all the other markets the final investor yield is close to random. If you are not familiar with the platform, Finnish loans are FF bread&butter; Sweden, Poland and Denmark are new markets so I expected to see this kind of volatility, figures should stabilise once FF issues a statistical relevant number of loans.
FinBee
There is not a lot of stats from FinBee and it is not clear which period the below figures cover (my assumption is last year). That said, you can see that in this case higher risk was compensated with higher return:
Mintos
Mintos is the biggest disappointment for me, because loan originators on their platform are rated by their internal model but it is not possible to see how those ratings performed over time. They have troves of data but they offer only a static picture of how your portfolio is performing today, no past analysis is possible, for example if a loan originator was upgraded or downgraded and why. Buyback guarantee also distorts investors risk picture; without bb, an investor can see (and be subject to) a default distribution, while bb morphs risk in a digital option: either the loan originator is solvent, and investor has 0 risk, or the originator defaults, but there is no way to monitor how close each originator is to the tipping point.
Linked Finance
LF does not report lot of data, they offer only some graphs, which at least seem realistic based on my experience so far (my personal default rate is 2,8% while the reported platform one is 1,2%). Keep in mind that their estimated returns are ‘recession adjusted’, meaning that since 2013 Irish economy did quite well and as expected actual returns are above estimates; we will have a real proof after the first economic downturn. If you exclude the Y grade, which has a negligible number of loans, the other ratings indeed offer more risk-adjusted returns when you go from left to right.
Conclusion
I based this (fast) analysis only on the platforms I use and I have experience with; I did not include DoFinance and ViaInvest because returns there are more or less fixed and there are no risk-based choices to make, Viventor does not rate the loan originators on its platform.
FinBee seems to be the only place where it makes sense to invest in loans with interest rates higher than 20%: one instance out of multiple platforms makes it more like the exception that confirms the rule; in other words, I would not suggest you to fall for the allure of ginormous yields, like HR Spanish loans in Bondora or 1 star Polish loans in Fellow Finance.
Does your investment experience offer you a different conclusion? Share it with us!
What I am reading now:
Follow me on Twitter @nprotasoni