james
Posts: 2,205
Likes: 955
|
Post by james on Feb 20, 2014 21:09:49 GMT
I am keeping those as I had little success in selling them - maybe I did not try hard enough. Maybe. The interest rate and size are both important. I have had little success in trying to sell any loan where the rate for the buyer was below 17% but did sell one on 2014-01-20 with balance 38.56 Euros at about 15.6% after 38 days. The original loan interest rate was 15% and I sold at -2%. Next lowest rate for a sale of this type was over 17%. I tried selling some loans that started at 100 Euros. One sold after more than a month. No urgent reason to sell because they are all paying. Just lower interest rate than I want. But the original interest rates are 17% or 18% and I am not interested in giving a discount to sell those.  Still tried to sell for at least 17% for the buyer. I do not lend 100 Euros now. I want to have more confidence that I will be able to sell quickly when I want to sell. But even if I am using my aggresive investment profiles, I usually get around 4 high percentage interest loans for every low percentage loan, which I feel is a good ratio. Yes, I think that the ratio I am seeing is OK. The trick in my opinion is not to leave aggressive investment profiles running constantly, but to pause it most of the time and only turn it on if there is much surplus cash in the account. Be sure to pause it before 7pm (Estonian time) otherwise it will then bid on all open loans your not already in. If I turn it on in the morning and pause it in time, I only get the fresh loans, hopefully many quickfunding ones. I would do this but I have a practical problem. 7PM Estonian time is some hours before the end of my sleeping time because of the times when I work.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Feb 20, 2014 21:12:20 GMT
No he means premium. You can sell loans at up to 5% premium. After the 1.5% resale fee that means 3.5% profit. I've sold overdue loans at premium, but usually only 1 to 3%. My first reaction was "Who would buy late loans at a premium?" but if IP's collection process has been effective, and the Late fees are bonuses to the lenders then I suppose it isn't that unreasonable. Some IP lenders obviously are gamblers! Yes. Some will take that risk to get the late fees. Time since becoming late may also matter. If a loan is one or two weeks late it is easy to believe that it is temporary. Harder to have that belief if it is 59 days late... At present I am experimenting with keeping late loans for longer to see what happens. I used to sell after about two weeks late. This experiment is part of why I have not had many late loans to try to sell recently.
|
|
|
|
Post by djia977 on Feb 23, 2014 10:34:50 GMT
I'm a bit puzzled as to why there are only 3968 loans detailed in the download. On the investment statistics page IsePankur state that they have made 6102 loans to date. OK, the dataset only starts from Feb 2011 so earlier loans will not be included - however, based on the number of existing customers in Feb 2011 (5842) versus currently (68613) I would be surprised if many more than 600 loans would be missed from the dataset due to loan age.
So where is the data on the other 1600 or so loans?
|
|
|
|
Post by djia977 on Feb 23, 2014 11:07:32 GMT
Here's another curious finding:
Out of the 243 Finnish loans, which started being approved on 23/07/2013, there have only been 4 defaults and only 10 which have at any time been 14 days overdue, and -
since 26th September last year not 1 Finnish loan has been at any time 14 days late
Looks like Finnish loans are suddenly the safest of the safe.
Edit: Also, according to the dataset not one Spanish loan has ever been 1 day late in paying, although I have several.
|
|
|
|
Post by djia977 on Feb 25, 2014 12:21:37 GMT
I sent off an email to partel regarding the incomplete 'in debt ... days' fields and it's been confirmed that these are only complete up to early October 2013. There should be an updated dataset on the IP website soon.
It should be very useful with the Finnish and Spanish loans to see how many have become late in the early days and compare with the Estonian loans experience.
|
|
duck
Member of DD Central
Posts: 2,864 
Likes: 6,895
|
Post by duck on Feb 25, 2014 17:30:56 GMT
Thanks for that djia977, I read your previous post but haven't had time to dig into the figures ..... unless I have been extremely 'unlucky' I can't believe that figure to be true since I have 4 Finnish defaults none of them making even a first payment!
|
|
|
|
Post by djia977 on Feb 25, 2014 19:42:08 GMT
Hi Duck,
Out of those 243 Finnish loans, 215 are in the A1000 category, of which only 63 have gone past their first 2 months of repayments. Of those 63, 4 have defaulted.
and it isn't just your bad luck, i've got all 4 as well!
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Feb 26, 2014 15:08:09 GMT
|
|
|
|
Post by djia977 on Feb 26, 2014 20:05:40 GMT
I've looked at the new dataset which now has complete data on days in debt.
As a sample to test between markets, i've looked at all November 2013 loans in the A1000 category (to find the A loans i've filtered for 'spare cash >250')
For Estonia there were 147 loans and of these just 2 have been 30 days late. (most of Novembers loans haven't had the chance to default yet, so 30 days is what i'm looking at as a quality indicator)
For Finland there were 34 loans and 4 of these have reached the 30 days late point.
For Spain there were 22 loans and 3 of these have reached the 30 days late point.
To test whether this is an improvement over the October loans i've again compared the A1000 loans for Estonia and Finland
For Estonia there were 151 loans, of which 1 has defaulted and 2 others are or have been 30 days late.
For Finland there were 22 loans, of which 1 has defaulted and 1 other is or has been 30 days late.
It's a small sample size but the Finnish loan performance doesn't appear to have improved between October and November and is very much poorer than the Estonian loan performance.
I'll be waiting until the differential reduces somewhat between Estonia and the new markets, but at least this dataset will give us an early 'heads up' as an when that occurs.
|
|
JamesFrance
Member of DD Central
Port Grimaud 1974
Posts: 1,323
Likes: 897
|
Post by JamesFrance on Feb 27, 2014 8:04:44 GMT
When I first started on isePankur in late August I bought many A1000 loans on resale to get my money invested quickly. It would have taken too long to analyse each borrower and resale does not show nationality. The result was that I have 1250€ of Finnish loans of which 300€ have defaulted. Because of that I will not be investing in Finland unless there is some evidence of recovery, although the ones I still have after selling a few late payers are up to date with payments. I have now 20 Spanish loans, none of which are late, but they have been carefully selected. As it is not possible to invest manually in good Estonian loans, I am worried about iP raising the DTI level based on statistics with such a short history. It is the likely rate of collection after default which concerns me.
|
|
|
|
Post by taavi on Feb 27, 2014 12:20:56 GMT
When I first started on isePankur in late August I bought many A1000 loans on resale to get my money invested quickly. It would have taken too long to analyse each borrower and resale does not show nationality. The result was that I have 1250€ of Finnish loans of which 300€ have defaulted. You can only look at a certain nationality loans on resale if you use the criteria in the search filter.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Feb 27, 2014 17:12:01 GMT
taavi, biggest problem buying on resale market is lack of a filter or sort for estimated return. That makes it very slow to find possibly interesting loans to buy.
On the selling side, I currently have about 2500 Euros in the resale market at an average interest rate of about 20.5%. Given market rates for initial purchases, all of those loans that are not late should have sold because I'm offering estimated returns well above those that are being accepted in the new loan market. Between the middle of January and now I have sold about 640 Euros of loans at an average buyer return of about 19.3%.
Something seems to be causing the resale market to not operate as effectively as it should and I think it is that estimated return filter not being present. It would probably also help to be able to filter loans by time remaining. This is because there is more lender competition for short loans than long loans and it would help to find loans of wanted remaining term in the resale list.
|
|
james
Posts: 2,205
Likes: 955
|
Post by james on Feb 27, 2014 21:47:15 GMT
I have had little success in trying to sell any loan where the rate for the buyer was below 17% but did sell one on 2014-01-20 with balance 38.56 Euros at about 15.6% after 38 days. The original loan interest rate was 15% and I sold at -2%. Next lowest rate for a sale of this type was over 17%. An update on this, I've recently sold some at returns between 15.7 and 16.8 for the buyer. These were loans with about 2.50 remaining and a little under 12 months of term remaining. All on time. They sold within a few days of being offered.
|
|
duck
Member of DD Central
Posts: 2,864
Likes: 6,895
|
Post by duck on Feb 28, 2014 6:46:11 GMT
On the subject of selling loans, what is the quickest way of seeing which loans you currently have up for sale? I want to cross check my spreadsheets and the way I have used in the past seems very clunky,,,,
|
|
|
|
Post by djia977 on Feb 28, 2014 11:11:28 GMT
Interest rate and defaults
Using the updated dataset which handily gives us a new column for interest rate I have investigated something which I had noticed from my own loanbook, namely that defaults rise very markedly for loans at or above 28%. I've now been able to get some numbers together to show just how big this effect is
Firstly, I've filtered for Estonian A1000 loans which have passed their 2nd payment date and are thus able to have defaulted, and i've used the loans from 2013 as my first test data.
Initially I grouped loans into rate bands such as below 20%, 20-23%, 24-27%, 28% and over. There is little difference between the sub 28% groupings so I feel there is only a need to show loans below 28% and those at or above 28%.
2013 shows 345 loans below 28%, of which 5 have defaulted giving a default rate of 1.45%
2013 shows 601 loans at or above 28%, of which 33 have defaulted giving a default rate of 5.49%
The average rate of interest for the 2 groups was 28.09% vs 18.11%
Bear in mind that the average age of these 'defaultable' 2013 loans will be about 10 months, and there is scope for the annual default rate to climb higher.
To check the validity of these findings I have run the same test on the 2012 data.
2012 shows 141 loans below 28%, of which 5 have defaulted giving a default rate of 3.54%
2012 shows 92 loans at or above 28%, of which 15 have defaulted giving a default rate of 16.30%
The average rate of interest for the 2 groups was 28.34% vs 16.85%
Summary
In 2012 the chance of default with 28% and above loans was 4.6x higher than with those with lower interest.
In 2013 the chance of default with 28% and above loans was 3.8x higher than with those with lower interest.
The higher average rate of interest with 28% loans does not, in my opinion, warrant the extremely higher risk of default.
|
|
Still tried to sell for at least 17% for the buyer.
