|
|
Post by perpetualtraveler on May 2, 2016 12:14:00 GMT
For those of you that use Beeplus to invest on Bondora, can you post your rules in this thread?
Mine are as per below, higher interest rates for lower ratings but lower bid amounts. Plus Estonia only. Let me know what you think.
1. A-C 25%+ Estonia
Rating AA-C
Bid: 25-35
Interest: 25%+
Verification: By phone, By other document, By bank statement
Country: Estonia
2. D-E 30%+ Estonia
Rating D-E
Bid: 20-25
Interest: 30%+
Verification: By phone, By other document, By bank statement
Country: Estonia
3. F 37%+ Estonia
Rating F
Bid: 15-20
Interest: 37%+
Verification: By phone, By other document, By bank statement
Country: Estonia
4. HR 40+ Estonia
Rating HR
Bid: 10-15
Interest: 40%+
Verification: By phone, By other document, By bank statement
Country: Estonia
|
|
carlos
I'm short Bondora and long p2p.
Posts: 104 
Likes: 21
|
Post by carlos on May 2, 2016 12:31:40 GMT
IMHO I think there are no loans AA-B with 25+% interest. When selecting for "grade" together with interest range, you should add one rule per loan grade.
|
|
|
|
Post by perpetualtraveler on May 2, 2016 13:15:19 GMT
Yeah I agree but in this case I didnt want AA-B loans under 25% so just added them in case...
|
|
|
|
Post by Estonia Invest on Jul 5, 2017 18:19:12 GMT
I know this was posted over a year ago but now that BeePlus has added Secondary Market, I was wondering if anyone has used the loan dataset to make statistical calculations to figure out the optimal settings for Secondary?
Mine are as follows but I don't know how to do advanced analysis, just using common sense...
Rating F: Estonia, Interest 37%+ XIRR 40%+ At least 10 months old, no lates, Income Verified+Income and Expenses Verified
Rating E: Estonia, Interest 33%+ XIRR 32%+ At least 10 months old, no lates, Income Verified+Income and Expenses Verified
Rating D: Estonia, Interest 28%+ XIRR 25%+ At least 10 months old, no lates, Income Verified+Income and Expenses Verified
Amount per individual loan is $23-140.
Just FYI I have Categories AA-C included in the "D" rule but it doesn't catch any loans.
Has anyone done any analysis of historical data to come up with better numbers? Can you share your settings similar to what I posted above?
Thank you.
|
|
miso
Posts: 25
Likes: 8
|
Post by miso on Jul 17, 2017 11:05:49 GMT
I've spent weeks learning statistical software and analysing/backtesting loan data last year before starting investing in Bondora and the result is that if you're buying older loans = as you say 10 payments done with perfect payment history there's no need to add additional conditions. The fact the borrower is paying on time beats any other variables you'll get from statistical analysis. Actually even 1st payment done will make big difference in future default probability compared with initial data set. Exception would be those high 100+ interest loans where actually any further payment term would make the loan more riskier - but that's just my gut feeling, in the time I was doing the analysis there were no such loans available and now I am lazy to analyze it again  |
|
|
|
Post by rahafoorum on Jul 17, 2017 11:56:11 GMT
I've spent weeks learning statistical software and analysing/backtesting loan data last year before starting investing in Bondora and the result is that if you're buying older loans = as you say 10 payments done with perfect payment history there's no need to add additional conditions. The fact the borrower is paying on time beats any other variables you'll get from statistical analysis. Actually even 1st payment done will make big difference in future default probability compared with initial data set. Exception would be those high 100+ interest loans where actually any further payment term would make the loan more riskier - but that's just my gut feeling, in the time I was doing the analysis there were no such loans available and now I am lazy to analyze it again Note that the improvement you see with 1-2 payments already is most likely somewhat skewed and misleading for today's loans. With Slovakia and both Spain and Finland, there were lots of very suspicious loans that defaulted with first payment initially when those markets were launched. I don't have the numbers here, but I'm pretty sure the situation has changed to better in this regard and thus the difference has decreased. However, it doesn't change your other conclusions though.
|
|
invest
P2P Blogger
Alive
Posts: 33
Likes: 17
|
Post by invest on Jul 30, 2017 17:11:59 GMT
|
|
