|
|
Post by reeknralf on Jul 24, 2014 7:50:54 GMT
As I've just started, I've been trying to pick up some secondary market loans, to accelerate entry.
Has anyone tried to estimate at what point secondary market loans are better value than primary market loans?
I made several simplifying assumptions, and did the sums for a 48 month loan, that has run 5 months. I reckon that the probability of repayment needs to have increased by ~4% to justify a 5% premium to BV. So if probability of default was 10% at the start, the fact that 5 payments have been made, needs to mean that probability of default is now 6%. I'm guessing this is not the case, so primary market loans are better value.
Does anyone know of any data or models that show probability of default against time, i.e. prob default in 1st month, 2nd month, 3rd month....?
|
|
|
|
Post by wiseclerk on Jul 24, 2014 8:15:00 GMT
I understand what you attempt to do, but for starters I'd just stick with shown "expected return" Defaults and recovery against time is in this curves: www.p2p-kredite.com/diskussion/grafiken-ausfallraten-2009-2012-t2028.htmlThey are orginally from the Bondora forum but right now I don't remember the source thread. Are you planning to automate your picking on the secondary market in any way. If not then I don't feel potential gains by your approach will offset the additional time spend for picking. P.S.: And if you are buying them at premium, you have to somehow estimate the effect of early repayments and figure that into. They don't happen that oftn but if they do and you bought at 5% premium (=6.5% cost including fee) you will take a loss if an early repayment happens in the next 2-4 months). |
|
|
|
Post by reeknralf on Jul 24, 2014 11:17:54 GMT
Thanks for the data. What are the units on the y-axis, number of loans?
You are right about the time spent and also that I have ignored early repayment. I can see that the time spent isn't worth it in the short-term (i.e. in terms of the improved returns I get on the loans I select, as compared with just buying at random), but it does allow me to familiarise myself with the system, and thus gain an idea of when to buy/sell loans in the future.
The data seem to confirm that 5 repayments do not decrease default probability by anywhere near enough to justify a 6.5% premium. I haven't done the sums, but am left with the impression that the secondary market is over-priced, and by extension that one should systematically sell any loan that can be sold at a 5% premium. This would of course be contingent on being able to pick up enough new loans to remain fully invested.
|
|
|
|
Post by coolrunning on Jul 28, 2014 19:56:18 GMT
... I haven't done the sums, but am left with the impression that the secondary market is over-priced, and by extension that one should systematically sell any loan that can be sold at a 5% premium. This would of course be contingent on being able to pick up enough new loans to remain fully invested. Flipping at 5% for B+ was being done quite successfully until recently, but now there are plenty on the primary market. Check out WiseClerk's own site, he has a nice report. Now I suspect little is sold on the secondary market above 2-3 %. Certainly I would never go higher.
|
|
