|
|
Post by meledor on Dec 9, 2016 9:33:12 GMT
The biggest platform with a provision fund is obviously RS but personally I think that RS do a decent job in disclosing the details of their provision fund, how it works and the impact of increased defaults. Some of the the smaller platforms however provide very little detail on their PF and in some cases their PF is so laughably small (in % terms) that frankly they should be banned from advertising it's existence ... other platforms give no/insufficient detail on how their PF is deployed and should therefore also be banned from advertising it
Arguably the smaller the size of the provision fund (in % terms) the less risk introduced.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Dec 9, 2016 9:48:50 GMT
This is interesting for ablrate , MoneyThing and quite likely others: " 3.43 There is existing guidance in our rules to indicate to firms the types of disclosure we expect to
see in order to satisfy the rules. This includes information on expected and actual default rates and how firms assess creditworthiness". ( page 19) I'd say that neither platform comes close to meeting the requirements to disclose expected default rates - is there even any at all? - and neither really does as much as seems appropriate in disclosing how creditworthiness is assessed. They undoubtedly aren't alone in this. One area for both is the disclosures about the backer of a loan where that backer is providing a guarantee of some sort, so the abilities of the backer as well as of the borrower need to be disclosed to determine the ability to meet the commitment. |
|
rick24
Member of DD Central
Posts: 244
Likes: 138
|
Post by rick24 on Dec 9, 2016 10:40:55 GMT
Sounds like a good idea to standardize mortgage approval criteria (avoiding regulatory arbitrage; achieving government policy goals relating to - i.e. damping down - buy to let frenzy). No point in having rules intended to stabilize the financial system if p2p can circumvent them.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Dec 9, 2016 10:54:23 GMT
Sounds like a good idea to standardize mortgage approval criteria (avoiding regulatory arbitrage; achieving government policy goals relating to - i.e. damping down - buy to let frenzy). No point in having rules intended to stabilize the financial system if p2p can circumvent them. That depends which end of the deal you're looking at. For a consumer lender the constraints could limit their ability to get their money back on any predictable schedule. That in turn would presumably require an increased risk premium and might make the area non-viable. Just imagine a risk disclosure saying "we may be required to let the borrower keep the mortgage until their death, perhaps 20 years after the planned end of the loan date, and settle it out of their estate if the estate has any money after paying for any care costs they may later have".
|
|
bigfoot12
Member of DD Central
Posts: 1,817
Likes: 816
|
Post by bigfoot12 on Dec 9, 2016 11:48:39 GMT
No point in having rules intended to stabilize the financial system if p2p can circumvent them. Again part of the original point of P2P is that the financial system wasn't at risk so loans would be made that banks would refuse. Rules that exist to protect the financial system shouldn't apply to P2P (as long as neither P2P investors nor platforms borrow from the financial system).
|
|
|
|
Post by SophieThing on Dec 9, 2016 14:20:18 GMT
This is interesting for ablrate , MoneyThing and quite likely others: " 3.43 There is existing guidance in our rules to indicate to firms the types of disclosure we expect to
see in order to satisfy the rules. This includes information on expected and actual default rates and how firms assess creditworthiness". ( page 19) I'd say that neither platform comes close to meeting the requirements to disclose expected default rates - is there even any at all? - and neither really does as much as seems appropriate in disclosing how creditworthiness is assessed. They undoubtedly aren't alone in this. One area for both is the disclosures about the backer of a loan where that backer is providing a guarantee of some sort, so the abilities of the backer as well as of the borrower need to be disclosed to determine the ability to meet the commitment. Thanks for your comments James. You missed out the remainder of this clause which states 'Within this high-level framework, we leave flexibility for firms to determine what information to disclose and how best to disclose it, taking account of their offering and target customer base.’ The context of this clause is in the presentation of loans, so that information is presented in a balanced way. We’ve spent some time looking at this and we always aim to present a balanced picture of each loan, the risks as well as the rewards. We have a financial promotions policy that we follow for each loan. In regards to general information about loans across the platform, I agree we could do better and indeed it is in our plans to do so, quite soon. We will also ensure that we remain up to date with FCA guidance and any new regulations that are implemented within the sector. To date, we don’t plaster the fact that we have had no defaults on the platform as we are a young platform and we do not want to mislead new lenders into thinking that this will always be the case. We will shortly release our new website plan which will include this fact on it, with some wording to balance it out. It is also in our plan to release more statistics on our loan book, but we have a couple of significant development projects that take priority over this at the moment. Kind regards, Sophie
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Dec 9, 2016 20:50:32 GMT
This is interesting for ablrate , MoneyThing and quite likely others: " 3.43 There is existing guidance in our rules to indicate to firms the types of disclosure we expect to
see in order to satisfy the rules. This includes information on expected and actual default rates and how firms assess creditworthiness". ( page 19) I'd say that neither platform comes close to meeting the requirements to disclose expected default rates - is there even any at all? - and neither really does as much as seems appropriate in disclosing how creditworthiness is assessed. They undoubtedly aren't alone in this. One area for both is the disclosures about the backer of a loan where that backer is providing a guarantee of some sort, so the abilities of the backer as well as of the borrower need to be disclosed to determine the ability to meet the commitment. Thanks for your comments James. You missed out the remainder of this clause which states 'Within this high-level framework, we leave flexibility for firms to determine what information to disclose and how best to disclose it, taking account of their offering and target customer base.’ The context of this clause is in the presentation of loans, so that information is presented in a balanced way. We’ve spent some time looking at this and we always aim to present a balanced picture of each loan, the risks as well as the rewards. We have a financial promotions policy that we follow for each loan. The framework requirement is to disclose information on expected and actual default rates. The piece you quoted doesn't remove the framework's requirement to disclose anticipated and actual default rates but rather is subordinate to it and gives flexibility in how to do that - it's "within" the framework not modifying the framework. And of course that makes sense since investors need both interest rates and anticipated default level to estimate returns and compare platforms. Unless, of course, you have more specific guidance from the FCA that says that disclosure of anticipated default rates is not required.
|
|
ablender
Member of DD Central
Posts: 2,204 
Likes: 555
|
Post by ablender on Dec 9, 2016 22:26:28 GMT
Thanks for your comments James. You missed out the remainder of this clause which states 'Within this high-level framework, we leave flexibility for firms to determine what information to disclose and how best to disclose it, taking account of their offering and target customer base.’ The context of this clause is in the presentation of loans, so that information is presented in a balanced way. We’ve spent some time looking at this and we always aim to present a balanced picture of each loan, the risks as well as the rewards. We have a financial promotions policy that we follow for each loan. The framework requirement is to disclose information on expected and actual default rates. The piece you quoted doesn't remove the framework's requirement to disclose anticipated and actual default rates but rather is subordinate to it and gives flexibility in how to do that - it's "within" the framework not modifying the framework. And of course that makes sense since investors need both interest rates and anticipated default level to estimate returns and compare platforms. Unless, of course, you have more specific guidance from the FCA that says that disclosure of anticipated default rates is not required. how would you calculate the expected default rate for a platform that has never had a default? The way I see it is 0% would be false because there will be a default at some point in the future. Similarly any other figure can (at least potentially) be false as they are invented figures (some kind of educated guess at best). What is your suggestion?
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Dec 9, 2016 23:48:05 GMT
james how would you calculate the expected default rate for a platform that has never had a default? The way I see it is 0% would be false because there will be a default at some point in the future. Similarly any other figure can (at least potentially) be false as they are invented figures (some kind of educated guess at best). What is your suggestion? I see no way for a lending-based platform to conceivably be treating its customers fairly without first having made some sort of estimate for potential default rates and returns as part of its initial and ongoing business plan. Loan performance is critical to the success of the business and the viability of its offers to investors so it's an important factor in such a plan. Nor do I see much potential for treating customers fairly if there is not some estimate made for each loan offered to investors as part of establishing the terms of the offer and how to present it, since platforms have to describe the offer in compliance with the financial promotions rules, a critical part of which is telling investors the information they need to assess the offer. It's no surprise that the FCA repeatedly expressed significant reservations about the compliance with promotions requirements in its interim report. At MoneyThing the football stadium loans illustrated the way MoneyThing does in fact have at least some view on risk and modifies deals accordingly over time. ablrate did something similar with a radically revised proposal for the Scottish holiday park loan based on initial feedback, as previously Saving Stream had done in deciding not to do a previous version of the deal. So the value established before going into business as a platform and modified in light of the changing nature of the loans and actual bad debt experience as it happens could be the one to use. And given lack of actual bad debts, an observation that there is significant uncertainty and some sort of range would perhaps also be appropriate. With discretion "within" the requirement to at least do it to some degree. Personally I think that MoneyThing is broadly doing a good job in this area compared to many, so it's just the lack of even an estimate that I thought needed a mention. Of course all platforms will have more direct feedback available from the FCA on these matters than just the consultation response, and given the FCA's concerns in this area I expect that we're going to see improvement in every platform in these areas, not just the two I mentioned, which I use in part because I think that they do a better than average job in these areas - they are good who could do better rather than bad meriting special FCA attention.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Dec 10, 2016 1:49:22 GMT
I suppose I should give an example. It might be something like this: "Broadly we expect that a well diversified investor using our platform might see n-o% loss to defaults and the per-loan variation depending on the individual loan properties we target is from p% to q%". Use real numbers instead of n, o, p and q and that sets out some sort of expectation, the potential per-loan range and the need to diversify to achieve the platform-level expectation. Could be 1, 4, 0.5 and 8 say.
|
|
|
|
Post by dualinvestor on Dec 10, 2016 6:18:10 GMT
Whatever the correct anticipated default rate the FCA is clearly of the opinion that current information published by platforms in this regard is inaccurate and sometimes misleading particularly where they make statements along the lines of "not a penny lost" or "backed by a provision fund." In at least one case Receivers have been appointed, the property is being marketed for less than the loan and the platform has not "defaulted "the loan and states that the loan will be fully repaid.
The unease is I believe caused because the FCA are of the opinion that the lack of meaningful information about defaults and the existence of provision funds combine to create a false sense of security and false market.
|
|
hazellend
Member of DD Central
Posts: 2,363
Likes: 2,180
|
Post by hazellend on Dec 10, 2016 6:47:57 GMT
Any anticipated default rate is a guess itself.
Surely just a disclaimer saying you could lose some or all of your money.
In equities you are generally advised that you may see drawdowns of up to 50% in a crash (if you are a broad indexer).
Personally, I wish the FCA would **crude language redacted**
|
|
ablender
Member of DD Central
Posts: 2,204
Likes: 555
|
Post by ablender on Dec 10, 2016 8:13:05 GMT
Thanks for all the replies relating to my previous post.
|
|
ablender
Member of DD Central
Posts: 2,204
Likes: 555
|
Post by ablender on Dec 11, 2016 12:06:10 GMT
james how would you calculate the expected default rate for a platform that has never had a default? The way I see it is 0% would be false because there will be a default at some point in the future. Similarly any other figure can (at least potentially) be false as they are invented figures (some kind of educated guess at best). What is your suggestion? There is a large amount of data on defaults and recoveries going back over 80 years in some cases. Various players from rating agencies, such as Moodys and S&P, investment banks, private equity companies, bridge lending companies and the likes of Experian have databases on this. This is not freely available since it takes considerable effort to collect and maintain these types of datasets. Clearly there is always going to be some subjectivity but it would be unfair to say that it's just an educated guess. I would argue that it's possible to put some sort of cap and floor on likely LGDs (losses given default) for most types of loan classes (SMEs, consumer loans, bridges etc). The main difficultly is that it's not clear that the dataset being used to calculate LGDs will be mirrored by the specific platform's loan population. So for example AC are using rating agency data (Moodys I think) for their LGD predictions (it drives their default probability and recovery assumptions). Personally, I would question whether AC's loan population is of the same quality as the dataset (i.e I think defaults might be higher and recoveries slightly lower on AC's loans). However, AC do deserve considerable credit for at least making the effort to think about likely LGDs and also putting that in writing on their website. Moreover, while I suspect that their realized LGD numbers might be somewhat higher than their forecast LGDs, their numbers do demonstrate that they can afford to be wrong by some margin and still offer decent returns. I do not understand why you object to the term "educated guess" when you say that they are subjective, the data set might (I would say will definitely) be different from those of any particular platform and that the quality of the loans is different. What I meant with that term is that it is definitely not a science and I do not see anything up to now to make me think differently. The only thing that I am getting is that there are "standard" ways accepted by the industry to arrive at a number which might or might not turn out to be true. Please note that I am not saying that there should not be any attempt at quantifying this. What I am saying is that without actual defaults on a particular platform I cannot trust that they will reflect the circumstances of that platform, their logic in accepting loans etc. In the meantime I think that there are more pressing issues that FCA should be looking at such as the structure of SM of some platforms such as LC where lenders are not allowed to buy the best available rates, possibly to the benefit of some lenders which lend at much reduced rates in order to help the platform get more loans through the door. I argue that such practices attempt to make the platform seem more successful than it actually is.
|
|
ablender
Member of DD Central
Posts: 2,204
Likes: 555
|
Post by ablender on Dec 11, 2016 13:47:19 GMT
I do not understand why you object to the term "educated guess" when you say that they are subjective, the data set might (I would say will definitely) be different from those of any particular platform and that the quality of the loans is different. What I meant with that term is that it is definitely not a science and I do not see anything up to now to make me think differently. The only thing that I am getting is that there are "standard" ways accepted by the industry to arrive at a number which might or might not turn out to be true. Please note that I am not saying that there should not be any attempt at quantifying this. What I am saying is that without actual defaults on a particular platform I cannot trust that they will reflect the circumstances of that platform, their logic in accepting loans etc. In the meantime I think that there are more pressing issues that FCA should be looking at such as the structure of SM of some platforms such as LC where lenders are not allowed to buy the best available rates, possibly to the benefit of some lenders which lend at much reduced rates in order to help the platform get more loans through the door. I argue that such practices attempt to make the platform seem more successful than it actually is. To be fair, you're correct. I think "educated guess" is ok. I'm actually conflating what you said with someone else saying that the default rate was a guess. However, "educated guess" still implies a single hypothesis or point estimate. In reality the dataset is much richer than a single point estimate would convey. We have an idea of means, variances and kurtosis. We have some idea of how defaults and recoveries vary over the cycle. So it much more of a phase space than a point guess. The relevance is two-fold in my view. First, that most platforms tend to repeat the mantra that they "price to risk". Well if that is really the case (I'm a skeptic since I think they price to liquidity) then they can demonstrate that by giving the lenders an idea about what default probabilities and recovery rates should be observed. Lenders can use this to benchmark the realized performance of a platform's loans against those expectations. Second, an understanding of the platform's expected LGD allows lenders to calibrate their own expectations without either unrealistic optimism or doom-mongering. Lenders, even of this forum, are still too shocked when defaults happen and recoveries aren't at par ("but it was secured ..."). Perhaps if more lenders were educated about the likely distributions of return outcomes then they could make more intelligent investment decisions. Perhaps even provision funds (essentially an expensive crutch for those who are too fearful of defaults, dragging down IRRs and moving risks rather around than actually reducing them) might be less popular! Thanks for your reply. I agree with you. Re "price to risk" vs "price to liquidity", my opinion is that we are currently seeing this happening in SS. Looking at the sizes of the particular loans and the number of lenders participating in these loans versus others at 12% will quickly reveal that there is a core of lenders that will bid on a loan even if secured against sea water. (Excluding obviously someone being prohibited from creating salt from sea water - this person would value sea water more than I am doing now, and rightly so). Re Lenders are "still too shocked when defaults happen" - It depends what defaults you refer to. If you include loans that default before even making a single payment or after a few payments, than I think that platforms should be held responsible for this. Personally I will feel shocked when this happens even though I would expect some defaults to happen but I will not expect or condone platforms not doing their part in the DD process when they are the ones who should have the information to carry out such DD. After all it is the platforms who decide what information will pass to the lenders. Keep in mind that we should not be contacting the borrowers for more information (apart from through the platform). There was a case, not too long ago when a borrower pulled out of a loan for being contacted by a lender, or so we were told. Re provision funds: If removing provision funds will actually result in an increase in the interest rate (gross) that we get, then I am up for it. If removing the provision fund means that that money goes to the platform then I will be less inclined to like it, even though as SS puts it, that money is the platform's money, I am inclined to disagree at least to a certain point.
|
|
