I have heard conjecture that large loans are riskier since they have a larger payment… While this seems intuitive I thought I would try to show it visually.
To start with I ran a quick pivot with 2 different loan buckets. Loans with payments from $0-$499 and loans with payments $500+.
This graph proves the hypothesis… Is the difference priceable? Can you make up the difference with higher rates? Absolutely, but the lower the credit grade the higher the risk adjustment that needs to be made.
(Notice the holiday color scheme?)
Here is the sql I used to get the data:
select [key],creditgrade,
case
when (status like ‘%late%’) then ‘Late’
when (status like ‘%defaulted%’) then ‘Default’
Else ‘Current’ end as status,
(AmountBorrowed*BorrowerRate/12)/(1-power(1+(BorrowerRate/12),-36)) as payment,
round((AmountBorrowed*BorrowerRate/12)/(1-power(1+(BorrowerRate/12),-36)),-3) as paymentBucket
from loan
where borrowerrate!=0
order by payment desc
Categories:
Uncategorized
Related Articles Related Stores
7 comments ↓
Good work.
Interesting because payment amount is a function of two items. Loan amount and interest rate. I wonder how much of the increased risk is contributed to by each item.
Where might one find default rate by interest rate for each credit grade?
@Little Troll
The trick with such an analysis would be to not slice the data into too many buckets. As the amount of data in each bucket decreases teh less reliable the results.
Are you thinking something similar to this analysis only using interest rate buckets? What buckets would you use?
Here is what I have so far
Rate-> 6 7 8 9 10 11
A 0.0% 0.0% 0.0% 1.5% 3.4% 3.1%
AA 0.0% 0.0% 0.2% 1.2% 1.7% 2.2%
B 0.0% 0.0% 0.0% 0.0% 4.6% 4.5%
C 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
D 0.0% 0.0% 0.0% 15.4% 14.3%
E 33.3% 50.0% 0.0% 0.0% 0.0% 0.0%
HR 25.0% 23.1% 14.3% 20.0% 23.8% 30.3%
NC 0.0% 0.0% 50.0% 0.0%
Grand 6.7% 2.3% 0.4% 1.4% 3.7% 4.6%
This sql on ProProsper will give you the raw data you seek…
selectcreditgrade,
round(borrowerrate,2),
sum(case when ((status like '%late%') or (status like '%default%')) then 1.0 else 0.0 end)/count([key]) as LateOrWorsePercentage
from loan
group by creditgrade, round(borrowerrate,2)
order by creditgrade, round(borrowerrate,2)
I will write up some reuslts for apost next week.
Interesting. I watched your video on the Prosper website, nice stuff.
Cheers
http://p2plendingwithprosper.blogspot.com/
Leave a Comment