|
Post by propman on Apr 7, 2016 7:25:38 GMT
Thanks for the update. I appreciate that any estimate will be wrong, but a couple of observations:
1) my personal experience of 5 year loans was an increase in the loans defaulting in their last year although I am not talking about a statistically meaningful sample. Of course the amount of default decreases as the amount on loan decreases but that is a different point.
2) I am concerned that the assessment appears to be based upon loan performance 6-18 months ago. While this is reasonable at the moment, this does leave a worrying gap between your experience of increased defaults in a downturn and adjusting the expected defaults. I have heard that the current performance against expectations is a surprisingly good assessment of the same cohorts future history. Doesn't RS build in any feedback from either more recent performance or economic indicators?
3) Please confirm that you do not intend to build in expectations of future recoveries and thereby understate the likely size of the PF available to meet future defaults.
- PM
|
|
adrianc
Member of DD Central
Posts: 9,045
Likes: 4,841
|
Post by adrianc on Apr 7, 2016 7:41:57 GMT
1) my personal experience of 5 year loans was an increase in the loans defaulting in their last year How do you know they defaulted, rather than being repaid early by the borrower?
|
|
alender
Member of DD Central
Posts: 957
Likes: 647
|
Post by alender on Apr 7, 2016 8:23:43 GMT
Is there any breakdown on the type of loan as I believe property loans are more likely to default in the latter part of the loan period.
|
|
|
Post by westonkevRS on Apr 8, 2016 17:57:39 GMT
Is there any breakdown on the type of loan as I believe property loans are more likely to default in the latter part of the loan period. The changes being made in terms of default curvature over the loan lifetime is only being made to retail loans. Your thoughts on property are probably correct, and especially (obviously) for bullet loans. Kevin.
|
|
|
Post by propman on Apr 11, 2016 7:57:23 GMT
1) my personal experience of 5 year loans was an increase in the loans defaulting in their last year How do you know they defaulted, rather than being repaid early by the borrower? The loans I am referring to were on Zopa. ratesetter made its first 5 year loans in February 2012 so does not have any that have gone full term yet.
|
|
|
Post by westonkevRS on Apr 11, 2016 8:13:11 GMT
Thanks for the update. I appreciate that any estimate will be wrong, but a couple of observations:
1) my personal experience of 5 year loans was an increase in the loans defaulting in their last year although I am not talking about a statistically meaningful sample. Of course the amount of default decreases as the amount on loan decreases but that is a different point.
This isn't my experience of consumer loans, neither from a statistical basis now a common sense viewpoint. Why default in year 5 after 4 perfect payment years? Customer has shown both intention and ability. Perhaps you experience coincided with an economic downturn with a cohort of loans that all started a few years prior. Perhaps your experience is more property focussed ;-) The time-scales are standard industry scorecard model performance periods, as well as being in line with regulatory expectations (not that we have to comply) such as Basel II. Any cohort more recent might be more in line with recent underwriting and economic environment, but doesn't have the time to demonstrate accurate performance. Old cohorts the reverse, less reflective of current underwriting but better statistical robustness. It's an eternal issue, but we've used what feels the best time-frame.
That said, this isn't a a homogeneous segment. Expected loss calculations are determined within multiple segments (e.g. home-owners), term and by risk credit bands. For example an 'A' grade 5-year home-owner customer will be more stable in terms of expected loss expectations; you'll just get more or less people falling into the consistent 'A' grade segment that the actual analysis cohort.
Correct, this is not considered. Any return we get on already defaulted loans is purely "fortuitous recovery". Although recognise that the expected loss calculation is a two stage process to get to the bad debt, it is a mix of the likelihood of initial default and the monies actually recovered after a fixed time period (we use two years). Kevin.
|
|
jimc99
Member of DD Central
Posts: 284
Likes: 115
|
Post by jimc99 on Apr 14, 2016 4:37:44 GMT
We've updated the methodology for the Provision Fund estimation of losses and therefore the coverage ratio. This will be changed on the 13th April 2016. It's blogged in summary here: www.ratesetter.com/blog/article/improving-the-accuracy-of-the-provision-fund-coverage-ratio-2With more detail in the members area when logged in: members.ratesetter.com/noticeboard/improving_the_accuracy_of_the_provision_fund_coverage_ratio_explaining_the_methodologyThe main change is: " The main update we are making is for loans to individuals, where we will move the likelihood of default from a ‘straight line’ over the term of a loan to a ‘curve’. So, for example, rather than a 60 month loan with a perfect repayment record being assigned a fixed likelihood of default over its whole term, our data suggest that the likelihood of default declines after 12 months, so we will apply that lower rate for the rest of the loan term." We always knew this would be the case. Anyone that works in consumer lending is well aware of this dynamic delinquency feature of maturing loans. We just wanted to wait until we had enough data to model it for ourselves. Kevin. I see the expected default rate has dropped a little today from 2.3% to 2.2% and the cover of the PF increased from 1.3 to 1.4 as a result. Presumably as a result of the above change to the calculation of the anticipated future defaults. Kev, you have previously indicated that you and the majority of people working in Ratesetter would feel better if the all important PF were stronger (this back when the PF stood at 1.5 cover). Very sad to see it strengthened today not from extra funding but through a jiggling of the math).
|
|
|
Post by westonkevRS on Apr 14, 2016 6:52:33 GMT
I find the last message very disparaging and depressing. I can only presume that if you don't trust RateSetter or think we are somehow "jiggling the numbers" you do not lend with us. That would be my hope expectation anyway.
We gave a clear message and relatively detailed bop and notification message of what changed. We also have 7 days notice. Are we only allowed to make changes that make things look worse? Every month we calculate the expected loss input variables and these often result in a reduced coverage ratio, and I've not heard you complain about us jiggling numbers then. How detailed an explanation do you need, a live feed of the analytics with peer review by the P2P FA?
Or perhaps you are one of those lenders that would prefer us to look much riskier than we are as to discourage lenders and keep rates high? This is my favourite, often mentioned at lenders drinks!
Anyways, as I've said before. You are welcome to come into the office anytime and I'll show you the numbers.
Kevin.
|
|
jimc99
Member of DD Central
Posts: 284
Likes: 115
|
Post by jimc99 on Apr 14, 2016 7:19:14 GMT
I invested when the PF coverage was at 1.7 and this gave me quite a bit of confidence. To sell out my holdings today I would pay a 3% fee and be £600 worse off.
I make no apologies for losing confidence in the PF. Despite previous assurances that RS were looking to strengthen it and that it was the No1 selling point of the platform in my opinion it is weakening over time.
|
|
|
Post by dualinvestor on Apr 14, 2016 8:48:44 GMT
. Very sad to see it strengthened today not from extra funding but through a jiggling of the math). Posted by westonkev about an hour ago I find the last message very disparaging and depressing. I can only presume that if you don't trust RateSetter or think we are somehow "jiggling the numbers" you do not lend with us. That would be my hope expectation anyway. Although the language may be a little perjorative jimc99's point is entirely valid. The Provision fund coverage ratio has not been increased because of the addition of any funds, it has increased because of a mathematically based reduction in the calculation of potential default. To take exception to the original comment indicates to me a certain amount of over sensitivity on the part of the platform. Personally I make no comment on the validity, or otherwise, of the adjustment, nor of the overall validity of claims of potential default based upon the current benign interest rate regime, which although it has presisted for some years is not a predictor of the future. Investors in P2P receive a multiple of interest many times that available in ordinary deposits for similar amounts, to believe that the provision funds of the various platforms give any security of capital is niave when one considers he fundemental fact, risk and return are inversely correllated. Basically we get up to 6% compared to 1 or 2% because there is a risk.
|
|
toffeeboy
Member of DD Central
Posts: 507
Likes: 363
|
Post by toffeeboy on Apr 14, 2016 11:01:25 GMT
Posted by westonkev about an hour ago I find the last message very disparaging and depressing. I can only presume that if you don't trust RateSetter or think we are somehow "jiggling the numbers" you do not lend with us. That would be my hope expectation anyway. Although the language may be a little perjorative jimc99's point is entirely valid. The Provision fund coverage ratio has not been increased because of the addition of any funds, it has increased because of a mathematically based reduction in the calculation of potential default. To take exception to the original comment indicates to me a certain amount of over sensitivity on the part of the platform. Personally I make no comment on the validity, or otherwise, of the adjustment, nor of the overall validity of claims of potential default based upon the current benign interest rate regime, which although it has presisted for some years is not a predictor of the future. Investors in P2P receive a multiple of interest many times that available in ordinary deposits for similar amounts, to believe that the provision funds of the various platforms give any security of capital is niave when one considers he fundemental fact, risk and return are inversely correllated. Basically we get up to 6% compared to 1 or 2% because there is a risk. As you it is the language used that is the problem, the way it is worded, the use of the word jiggling, suggests that the figures have been manipulated to increase the coverage rate which I think it is fine for someone connected with Ratesetter to raise an objection to. I assume, perhaps wrongly, that the expected default figure has been changed based on the historical data that Ratesetter has of it's past loans so to imply that the figures are being fiddled (let's be honest that is what was implied by the word jiggled and we all know it) is unfair and deserves a retort.
Regarding the coverage rate, I haven't read the small print on the PF but is there anywhere that states that they will keep the coverage rate high. I would assume that as long as they estimate bad debts to the best of their abilities and the coverage rate is above 1:1 then I don't see a problem as the PF has enough to cover the expected bad debts, yes it is nice to have a lot of extra cash sat around in case of a crash again but it is not imperative.
Once again the post is from someone that seems to want increased protection which will lead to reduced interest rates although they are looking to leave the platform anyway so expect everyone else to accept lower rates so their money is protected (which it more than is anyway).
|
|
pip
Posts: 542
Likes: 725
|
Post by pip on Apr 14, 2016 11:02:30 GMT
Posted by westonkev about an hour ago I find the last message very disparaging and depressing. I can only presume that if you don't trust RateSetter or think we are somehow "jiggling the numbers" you do not lend with us. That would be my hope expectation anyway. Although the language may be a little perjorative jimc99's point is entirely valid. The Provision fund coverage ratio has not been increased because of the addition of any funds, it has increased because of a mathematically based reduction in the calculation of potential default. To take exception to the original comment indicates to me a certain amount of over sensitivity on the part of the platform. Personally I make no comment on the validity, or otherwise, of the adjustment, nor of the overall validity of claims of potential default based upon the current benign interest rate regime, which although it has presisted for some years is not a predictor of the future. Investors in P2P receive a multiple of interest many times that available in ordinary deposits for similar amounts, to believe that the provision funds of the various platforms give any security of capital is niave when one considers he fundemental fact, risk and return are inversely correllated. Basically we get up to 6% compared to 1 or 2% because there is a risk. Hi, I don't think anybody on here would not agree that there will always be some risk in P2P, the risk is the reason why you get a premium on your interest and if you want zero risk you should go for a product backed by the FSCS. However the point is not that people can't accept there is some risk, it is that the risk of rs is directly linked to the provision fund. As the coverage goes down (I prefer to use my owns calcs to avoid all the moving goalposts), the safety net is less and risk rises. The question for investors therefore is whether the premium you get in interest is worth the risk. I concluded before, with higher coverage for me it was, I conclude now it is not. That's why I am gradually winding down my money. It's up for other investors to make their own decisions.
|
|
|
Post by Deleted on Apr 14, 2016 14:47:52 GMT
Without more information on thing likes recovery rates, collateral values on secured loans etc, the coverage ratio is not much use anyway, except as a vague ballpark number and a bit of a marketing gimmick
There has definitely been an increase in the 'loans not on time' though, and the 2014 loans have also had an uptick in the 'bad debt fund usage' metric over the last couple of months
|
|
alender
Member of DD Central
Posts: 957
Likes: 647
|
Post by alender on Apr 14, 2016 16:43:25 GMT
Regarding the coverage rate, I haven't read the small print on the PF but is there anywhere that states that they will keep the coverage rate high. I would assume that as long as they estimate bad debts to the best of their abilities and the coverage rate is above 1:1 then I don't see a problem as the PF has enough to cover the expected bad debts, yes it is nice to have a lot of extra cash sat around in case of a crash again but it is not imperative.
Once again the post is from someone that seems to want increased protection which will lead to reduced interest rates although they are looking to leave the platform anyway so expect everyone else to accept lower rates so their money is protected (which it more than is anyway).
Not sure how good a coverage ratio just above of 1:1 in benign times will be when the going gets tough. You say in case of a crash, if history teaches us anything a crash is inevitable, it is just a case of when and how deep. However given 1. The difficulty understanding the risks with the loan book. 2. The money in the PF has been used in ways that seem risky to some of us. 3. The lack of legal separation from RS. 4. A question on how much money is in the PF that is lent by other parties and cannot be withdrawn. 5. The lack of legal rules governing what the PF money can be used for/invested in, except for bad loans. 6. The changing calculations of the coverage ratio which makes comparisons difficult. Given the above I question the usefulness of the PF except as a marketing tool for RS, I have no idea how well it will perform when things get tough. As risk is spread around all investors perhaps it may be better to lose some interest off the headline rate to cover bad loans, this obviously varies on each payment but should be small. In the end we will be no worse off or perhaps better off as there is no cost to pay administering the PF, also by monitoring loss in each payment due to non performing loans we will have a clear idea whether the risk are increasing or decreasing.
|
|
|
Post by dualinvestor on Apr 14, 2016 18:03:46 GMT
You say in case of a crash, if history teaches us anything a crash is inevitable, it is just a case of when and how deep. I believe it was Oscar Wilde who first said "I can predict everything but the future," although I wholeheartedly agree with the comment above it is not a given. However the former head of the Finacial Services Authority (before it became the FCA) is on record as believing losses in P2P lending will make the bankers of last decade look like the most prudent professionals. Therefore to believe that the Provision Fund, or the other various names on other platforms, are a protection against the risk is, at the very least optimistic.
|
|