Improvements to the Stats page - what do you want to see?...

A few quick thoughts:



- Some disambiguation is required. What qualifies a loan for 'Default' and 'Bad Debt' status, respectively on your platform? Compared to AC, for example, you appear to be applying rather loose interpretations of those terms.



- Specifically, what is the status of (1) the C**** B***** L******** L** loan (Winding up order for the borrowing entity, Personal Bankruptcy hearing for the Guarantor was to have been heard on 11th June but which has now been postponed, Multiple unmet repayments in arrears) and (2) the P******* I********* P********* L** loan?


- You have purposefully not included any information about achieved levels of recoveries (are these negligible then?) and I think this is an important piece of information for many prospective and current lenders and an opportunity to differentiate yourselves as a platform if you have performed better than others.



- You have not included data as a % of loan capital outstanding which would be more helpful, e.g. % of Outstanding debt overdue, or of loans value originated e.g. % default rates.



- If you have only 2 'D' category loans how can you have 2 with '<30 Days Late' status and 2 with '>30 Days late' status?



- Can you really expect your stats efforts to be a meaningful guide and taken seriously if you change the risk band for a loan just so lenders can bid at a higher rate, as disclosed in one of your recent communications on 29/05/14: "This week on the auctions, we've split the A******* C********* a** E********** T********* (A**) loan in to two separate tranches. The first is live at the moment, and for the first time, we've moved the business down a risk category to allow lenders the opportunity to bid at a higher rate - up to 17%"?


- Most important of all, it seems that the chosen presentation of the risk bands performance is potentially misleading. Both the C**** B***** L******** L** loan the P******* I********* P********* L** loan are, I think, now portrayed as 'D' risk band loans. However, you know very well that at the time of auction/origination they were not D risk band loans (which seems to be a category created post-launch for known non-performing loans to be shunted into - correct me if I am wrong), e.g. P******* I********* P********* L** was originally another borrowing entity which went into liquidation and was originally rated A risk band. Defaults/bad debts should clearly be reflected in the original risk band allocated by your platform. As it stands your stats do not appear to present a true rand accurate representation of what has actually happened to loans in the given risk bands. FC can elect to change loans to a 'No risk band status' where a credit event has occurred but any lateness, default, recovery is attributed to the original risk band. By effectively concealing from anyone simply looking at your stats page the fact that an A (or any other original allocated) risk band loan is non-performing as a corollary of changing it to another risk band, which is the result of that action even if not the primary intention, it potentially imo brings the platform into disrepute.

Hi beastie, wysiati,

Thanks for the posts. Good point about the 'D' ratings.

I've edited these to be what they were when the loans were listed.  We're looking into ways that we can use more live information to better inform users of the secondary market as to the creditworthiness of businesses because this can quickly change.  I'll let you know once we have something.

Thanks for the suggestion to show the % of capital outstanding, I noticed this has been adopted by the P2P Finance Association's standardised format for 'Bad Debt debt over time', which is something we plan to adopt.

I take the point about relaying information about recoveries. I believe our loan book is proportionally more secured than some other platforms. However we have yet to successfully recover assets from defaulting borrowers because the process can take a long time.  ThinCats forecast a recovery of circa 50% and I think that's about right for the industry, however it depends on the cooperation of the courts.

We do communicate periodically when there is something to say about the progression of defaulting loans.

Thanks for bringing the communication of the 29/5 to my attention. I'll look into this. I think questions relating to the underwriting process warrant a different thread.

Its not our intention to 'conceal' a rating given to a business, however we do need to come up with a way that reflects the change in creditworthiness (for users of the secondary market). I'm not sure that 'declassifying' the rating is the best solution. All suggestions on how we might achieve this are welcome...

If you took the FC one as a template (it'd be nice to see both absolute values or percents on the display), I think the thing its missing is the ability to slice by time period - ideally as user selectable, or by year of origination. Better still would be additional optional slicing criteria - like industry type, type of security taken, even region (though I don't think that's held at present). As the loan book matures, this would allow investors to explore the real levels of risk for potential new loans. The one thing this table doesn't show is the vintage style analysis (which FC have as charts though they aren't great as they don't split by risk level) which is useful for identifying when in their lifetime loans are going bad - this would need to be a separate chart/table but I would say its of secondary importance to the basic table.

I don't know how feasible this is, but I'd like to see what % of the money invested in the platform is 'real' money as opposed to borrowed.
I'd hope there's almost no borrowed money being invested in REBS, but who knows!

Now I can foresee two problems with collecting/presenting this data straight away:

1. Data protection and similar laws will probably make it illegal
2. You may not even know whether a debit card payment is in fact a credit card payment in disguise, and even if you do, there's nothing stopping people having a loan paid into their debit card current account

So maybe that suggestion's doomed from the outset. Don't waste any time on it!


Second idea: pay interest in bitcoin? ^^

danraj I have a few suggestions but you have given no indication that you or your organisation listen. Your statistical reporting will always be poor unless you make it possible to link loan number to loan title.
Are you looking for advice to enable you to baffle potential investors using too much information like another site has just done?

Keep things simple. Call a spade a spade. Give clear definitions of headings. Be honest.

Good luck.