|
Post by Butch Cassidy on May 15, 2018 9:57:07 GMT
After 6 months of platform operations I thought it might be a good time to review the progress made so far & give an opportunity to highlight areas of concern or that need further improvements;
Loans so far 601, totalling £151k (£114k still active), averaging 102 days with those 7/14 days late being 39% & 32% against a target level of 10%. PF has now dropped to £23k against £50k starting fund (all visible on statistics page).
It now appears that a 10% default target is clearly too optimistic & 30-40% is much more realistic - I would be interested to hear from nsiam whether this level is seen as sustainable & still profitable going forward or that now a decent chunk of raw lending data is available the AI lending algorithm needs to be refined to reduce this level to nearer target? Personally I think 10% will be unachievable so perhaps 25% might be a better, more realistic target.
This has obviously impacted on the PF which is just under half it's initial value - What plans are in place to bolster the PF going forward & when will it become "self sustaining" either from recoveries or profitability?
Personally I have been satisfied with my own profitable lending experience but would need further assurance about the future growth trajectory & how the model is performing against expectation before committing any further funds.
|
|
|
Post by sayyestocress on May 15, 2018 15:50:08 GMT
It now appears that a 10% default target is clearly too optimistic & 30-40% is much more realistic - I would be interested to hear from nsiam whether this level is seen as sustainable & still profitable going forward or that now a decent chunk of raw lending data is available the AI lending algorithm needs to be refined to reduce this level to nearer target? Personally I think 10% will be unachievable so perhaps 25% might be a better, more realistic target. I do share your concerns, though if I understand correctly they call it a default 90 days after non-payment, so the pf is buying off lenders (after 30 days max.) before they reach that point. I think (and therefor may be wrong) that the 10% target relates to 90 days overdue, rather than 7/14/30. We don't have the 30 day overdue repayment stats (yet?) to judge non-repayments beyond 14 days overdue. It may get close to 10%, alternatively it may stay roughly the same. I don't think the non-paid loans will be old enough just yet for the (currently) 0 % 90 day overdue stat to have any value, but if this is around or below the 10% mark once the loan book matures then we can expect the pf to be sustainable under current market conditions and hopefully whatever change during the highly political not-too-distant future comes to pass. I had hoped for better returns at this point, they are definitely below the expected return table presented on the website but they are still satisfactory given this is a new endeavor. Hopefully the lending criteria is being tightened during this 'soft launch' and things will pick up and we can have more faith in the longevity of the provision fund.
|
|
|
Post by nsiam on May 19, 2018 16:26:05 GMT
After 6 months of platform operations I thought it might be a good time to review the progress made so far & give an opportunity to highlight areas of concern or that need further improvements;
Loans so far 601, totalling £151k (£114k still active), averaging 102 days with those 7/14 days late being 39% & 32% against a target level of 10%. PF has now dropped to £23k against £50k starting fund (all visible on statistics page).
It now appears that a 10% default target is clearly too optimistic & 30-40% is much more realistic - I would be interested to hear from nsiam whether this level is seen as sustainable & still profitable going forward or that now a decent chunk of raw lending data is available the AI lending algorithm needs to be refined to reduce this level to nearer target? Personally I think 10% will be unachievable so perhaps 25% might be a better, more realistic target.
This has obviously impacted on the PF which is just under half it's initial value - What plans are in place to bolster the PF going forward & when will it become "self sustaining" either from recoveries or profitability?
Personally I have been satisfied with my own profitable lending experience but would need further assurance about the future growth trajectory & how the model is performing against expectation before committing any further funds. Hi Butch Cassidy , thank you for the post. It is good to hear that you're are happy with your Welendus experience to date. We aim to further improve Welendus for our users for even better returns and experience. Very much appreciate your feedback. "It now appears that a 10% default target is clearly too optimistic & 30-40% is much more realistic - I would be interested to hear from nsiam whether this level is seen as sustainable & still profitable going forward or that now a decent chunk of raw lending data is available the AI lending algorithm needs to be refined to reduce this level to nearer target? Personally I think 10% will be unachievable so perhaps 25% might be a better, more realistic target." - Just to clarify as highlighted by sayyestocress , the 10% is the default rate where the 7/14 days stats are missed payment rates. Borrowers who miss their payments do not necessarily default unless if the payment is not made in 90 days from the original payment due date. But I agree with you on the rest. We are actually working on refining the credit model on daily basis. With that in mind, Our lending model is only getting better and stronger with the increase in lending activity. We are expecting the missed payment rate to start dropping soon as a result of the changes were are making. With that being said, the number will fluctuate around the current values for sometime before starting drop. " This has obviously impacted on the PF which is just under half it's initial value - What plans are in place to bolster the PF going forward & when will it become "self sustaining" either from recoveries or profitability?" - Our target is to get the PF to a self sustaining state. This will need some optimisation and we will do some capital injection into the fund (if needed & when required) until it balances out. With regards to the model, it is performing close to where we expected. We were aware that we will need to do some optimisation and tweaks with real life customers but overall all is within our expectations. With regards to the growth trajectory, I cannot say much here but we will have some great news to share very soon. Hope this answers your questions. Best, Nadeem
|
|
|
Post by nsiam on May 19, 2018 16:35:17 GMT
It now appears that a 10% default target is clearly too optimistic & 30-40% is much more realistic - I would be interested to hear from nsiam whether this level is seen as sustainable & still profitable going forward or that now a decent chunk of raw lending data is available the AI lending algorithm needs to be refined to reduce this level to nearer target? Personally I think 10% will be unachievable so perhaps 25% might be a better, more realistic target. I do share your concerns, though if I understand correctly they call it a default 90 days after non-payment, so the pf is buying off lenders (after 30 days max.) before they reach that point. I think (and therefor may be wrong) that the 10% target relates to 90 days overdue, rather than 7/14/30. We don't have the 30 day overdue repayment stats (yet?) to judge non-repayments beyond 14 days overdue. It may get close to 10%, alternatively it may stay roughly the same. I don't think the non-paid loans will be old enough just yet for the (currently) 0 % 90 day overdue stat to have any value, but if this is around or below the 10% mark once the loan book matures then we can expect the pf to be sustainable under current market conditions and hopefully whatever change during the highly political not-too-distant future comes to pass. I had hoped for better returns at this point, they are definitely below the expected return table presented on the website but they are still satisfactory given this is a new endeavor. Hopefully the lending criteria is being tightened during this 'soft launch' and things will pick up and we can have more faith in the longevity of the provision fund. Hi sayyestocress, thank you for the clarification. To answer your question on the return, we are expecting the returns to start improving as we build the loan book. Plus, we will soon be applying the bonus PF acquisition returns for Q1 which will improve the earnings. Best, Nadeem
|
|
|
Post by Butch Cassidy on May 20, 2018 7:46:30 GMT
I should definitely have been clearer with my use of the available statistics; they only represent early stage (7/14 days) late payment borrowers which is not the fully distilled 90 day+ default level, which will by definition be lower, but it is still not clear how much lower & I would stand by the claim it is likely to be higher than the 10% target & may remain stubbornly so going forward IMO. Currently my cash drag has increased to approx. 15-20% as late payers are refunded by the PF & new lending rates have slowed considerably but if that results in better AI targeting of lower risk borrowers that will be good news all round. That said I am still confident with my overall experience so far to suggests that the Welendus model can be sustainable & successful into the future & look forward to the "great news to share very soon" on growth eluded to above by nsiam.
|
|