blender
Member of DD Central
Posts: 5,719
Likes: 4,272
|
Post by blender on Apr 2, 2015 18:41:25 GMT
I suppose it's worth paying a small premium for safer loans.
Are you aware ratrace that the C- loans started about August 2013 (plus or minus a month) and so we do not yet know their statistical performance much beyond a year. I used to buy C- when they started and sold before the fourth repayment.
|
|
chrisf
Member of DD Central
Posts: 224 
Likes: 67
|
Post by chrisf on Apr 2, 2015 18:42:02 GMT
That is an original scheme. After how many repayments does a C- have the same risk profile of a new A loan? Do they then become increasingly safe? l feel l have already said too much, l don't want to risk having you guys crowding me out of my little niche in the market.  No, please carry on, I have just put some of my old C- loans up for sale, there are some great bargains there.
|
|
|
|
Post by ratrace on Apr 2, 2015 19:33:06 GMT
I suppose it's worth paying a small premium for safer loans. Are you aware ratrace that the C- loans started about August 2013 (plus or minus a month) and so we do not yet know their statistical performance much beyond a year. I used to buy C- when they started and sold before the fourth repayment. l don't try to trade, l just buy and hold when l feel happy with the risk level. lts not hard really, its just a case of understanding that the risk of default lowers as loans get older and then spreading your risk over many loans. Because through lending on Zopa when it had risk bands l found that lending to C- band was lower risk then l expected. The current bad debt rates at FC are B 2% C 2.5% C- 2.4% which would bare this out.
|
|
chrisf
Member of DD Central
Posts: 224
Likes: 67
|
Post by chrisf on Apr 2, 2015 19:59:00 GMT
I suppose it's worth paying a small premium for safer loans. Are you aware ratrace that the C- loans started about August 2013 (plus or minus a month) and so we do not yet know their statistical performance much beyond a year. I used to buy C- when they started and sold before the fourth repayment. l don't try to trade, l just buy and hold when l feel happy with the risk level. lts not hard really, its just a case of understanding that the risk of default lowers as loans get older and then spreading your risk over many loans. Because through lending on Zopa when it had risk bands l found that lending to C- band was lower risk then l expected. The current bad debt rates at FC are B 2% C 2.5% C- 2.4% which would bare this out. I totally applaud the sentiment of ratface, he/she may have spotted a gap in the market for low-rate C- loans, and who knows, this may be the way to go, C- grade loans have not been around long enough for us to collect long-term data. I reiterate for the benefit of fatface: I have some good C- loans up for sale now, which I am reluctantly selling due to stuff.
|
|
|
|
Post by goldservice on Apr 3, 2015 7:35:30 GMT
I think I understand the issues being raised here and I have found ratrace's approach to be original and thought provoking and so I welcome his posts even though I would be nervous about following that approach myself. Let a thousand flowers bloom!
|
|
blender
Member of DD Central
Posts: 5,719
Likes: 4,272
|
Post by blender on Apr 3, 2015 8:47:04 GMT
I appreciate Goldservice's approach here. In the spirit of wishing to assist ratrace's development of lending strategy, I would point out that the only definition we have of what is a C- loan on the FC platform is the predicted loss rate. The current C- population is not mature enough for the statistics to say how the loans placed in that band will match the predicted loss rate, but the predicted annual bad debt rate for the band is 5% of principal (statistics page). So for every £1000 of principal held in C- loans, FC predict that we will lose £50 per annum. If FC decide that the loss performance of C- is going to be substantially better than that, then they will be able to relax their (secret) criteria to let more risk into the band - as we think they have with the other bands.
Looking at the graphs of bad debt performance on the FC statistics page, the one striking feature is that the losses in the first six months are very low - although of course it generally takes 90 days from a late payment for a default to be declared - say four months from the start even if not a penny is paid back. But analysis reported long ago on this forum, looking at when unrecovered lates first happen I believe, seemed to show that the risky time was in the second half year of the loans - though it varies by band.
The conventional wisdom is that the safest period is the first six months, though of course if you choose the loans who have established a good payment history and who you believe have the determination and ability to stick with it, then the statistics are only a guide. Given that many people are trying to sell at six months, there can be good offers on the SM at that stage.
There are plenty of more counter-intuitive strategies used on FC which give good results.
|
|
|
|
Post by uncletone on Apr 3, 2015 20:04:34 GMT
I totally applaud the sentiment of ratface, he/she may have spotted a gap in the market for low-rate C- loans, and who knows, this may be the way to go, C- grade loans have not been around long enough for us to collect long-term data. I reiterate for the benefit of fatface: I have some good C- loans up for sale now, which I am reluctantly selling due to stuff. A small concern has been expressed. I am sure that your typing of "ratface" and "fatface" were a result of fat finger problems. The "f" and "r" keys are very close. One would not be impolite in these hallowed halls.
|
|
|
|
Post by ratrace on Apr 4, 2015 13:33:45 GMT
I appreciate Goldservice's approach here. In the spirit of wishing to assist ratrace's development of lending strategy, I would point out that the only definition we have of what is a C- loan on the FC platform is the predicted loss rate. The current C- population is not mature enough for the statistics to say how the loans placed in that band will match the predicted loss rate, but the predicted annual bad debt rate for the band is 5% of principal (statistics page). So for every £1000 of principal held in C- loans, FC predict that we will lose £50 per annum. If FC decide that the loss performance of C- is going to be substantially better than that, then they will be able to relax their (secret) criteria to let more risk into the band - as we think they have with the other bands. Looking at the graphs of bad debt performance on the FC statistics page, the one striking feature is that the losses in the first six months are very low - although of course it generally takes 90 days from a late payment for a default to be declared - say four months from the start even if not a penny is paid back. But analysis reported long ago on this forum, looking at when unrecovered lates first happen I believe, seemed to show that the risky time was in the second half year of the loans - though it varies by band. The conventional wisdom is that the safest period is the first six months, though of course if you choose the loans who have established a good payment history and who you believe have the determination and ability to stick with it, then the statistics are only a guide. Given that many people are trying to sell at six months, there can be good offers on the SM at that stage. There are plenty of more counter-intuitive strategies used on FC which give good results. With my own strategy l wait until at least 16 to 20 (depending on loan term) repayments have been made before l buy. By which time many of the weakest payers should have been weeded out. Then with what's left l choose to buy what l think will be the lowest risk while still offering 11%+ rate. Waiting this long does mean you have fewer loans choose from, so it does take longer to build up your spread of risk.
|
|
chrisf
Member of DD Central
Posts: 224
Likes: 67
|
Post by chrisf on Apr 4, 2015 17:10:54 GMT
I totally applaud the sentiment of ratface, he/she may have spotted a gap in the market for low-rate C- loans, and who knows, this may be the way to go, C- grade loans have not been around long enough for us to collect long-term data. I reiterate for the benefit of fatface: I have some good C- loans up for sale now, which I am reluctantly selling due to stuff. A small concern has been expressed. I am sure that your typing of "ratface" and "fatface" were a result of fat finger problems. The "f" and "r" keys are very close. One would not be impolite in these hallowed halls. Ah yes, apologies to ratrace for any offense caused by my clumsy fingering, and sorry if I was uncharitable about your novel investing strategy.
|
|
blender
Member of DD Central
Posts: 5,719
Likes: 4,272
|
Post by blender on Apr 4, 2015 19:45:26 GMT
Waiting 16 to 20 months is an interesting strategy. That may well be after the bulk of defaults. However we have no statistical knowledge of C- performance after than time. Please keep us in touch ratrace.
|
|
|
|
Post by GSV3MIaC on Apr 4, 2015 21:01:13 GMT
It's probably a great strategy for 12 month loans. 8>.
That's the problem trying to data-mine FC .. too many different loan terms (many of which have not existed for the whole life of FC). When I last looked months 13,14 and 30 were all significantly worse than you'd expect, but that is based on the whole universe of loan durations.
|
|
|
|
Post by loanstar on Apr 4, 2015 21:21:51 GMT
If I may make a small observation. The typical default rate graph is S shaped. So for a 36 month loan the number of defaults is low for the first six months. The rate of defaults then picks up between six and 20 months approx. Then the graph will flatten out for the remainder of the term. There are two problems I see with buying older loans. Some will be repaid early. The last payments will be mostly capital. (please correct me on the last point if my maths understanding is incorrect). The longer the loan in years, the longer the late period of low defaults.
|
|
|
|
Post by ratrace on Apr 4, 2015 22:09:15 GMT
If I may make a small observation. The typical default rate graph is S shaped. So for a 36 month loan the number of defaults is low for the first six months. The rate of defaults then picks up between six and 20 months approx. Then the graph will flatten out for the remainder of the term. There are two problems I see with buying older loans. Some will be repaid early. The last payments will be mostly capital. (please correct me on the last point if my maths understanding is incorrect). The longer the loan in years, the longer the late period of low defaults. The early repayment and the fact my money returns quicker in the later payments are not really a problem as l reinvest the money into newer loans. Also if you are worried about defaults then its good to have your money returning to you quicker.
|
|
mikeb
Posts: 1,072
Likes: 472
|
Post by mikeb on Apr 5, 2015 20:35:28 GMT
The early repayment and the fact my money returns quicker in the later payments are not really a problem As long as you didn't pay a premium for the loan part, otherwise the chances are you will lose out -- the premium is definitely gone, and you didn't get to hold the loan part long enough, to get interest to outweigh that premium
|
|
|
|
Post by ratrace on Apr 5, 2015 21:46:37 GMT
With my loan parts l get most of them at par or 0.3%+.
Am only willing to pay between 0.5%+ to 0.9%+ if the rate is very good or if am keen on gaining loan parts in the business.
|
|
