Print

[Cushie]

http://www.miamiherald.com/business/small_business/story/380991.html

While 13 percent is higher than Kond might get at his local bank, there are no penalties for paying off the Prosper loan early -- and the process was far more pleasant, he said. In business since 1994, Kond said when he goes to banks, he still feels on the defensive: ``I feel like they are looking for holes in my character or credit report [to find excuses] to not lend me money.''

At Prosper, on the other hand, bidders were actually competing to give him cash.

If making unsecured loans to absolute strangers sounds risky, Prosper claims it's not.

Since launching in 2006, the company has facilitated $110 million in loans from 550,000 users and has an overall default rate of 1.37 percent.   

[Cushie]

Some more interesting numbers from the article.  Maybe someone knows if this is true or not..

When Prosper launched in 2006, about 75 percent of all borrowers had AA credit.

Now, almost 95 percent are in that category and just 5 percent are considered sub-prime borrowers, he said.  (Larsen)

The default rate on its highest rated, AA-loans is 0.27 percent -- in line with brick and mortar banks.

The article invites comments.

[ira01]

Some more interesting numbers from the article.  Maybe someone knows if this is true or not..

When Prosper launched in 2006, about 75 percent of all borrowers had AA credit.

Now, almost 95 percent are in that category and just 5 percent are considered sub-prime borrowers, he said.  (Larsen)

More crapola from Larsen (assuming the reporter accurately reported what he said). 

According to Lendingstats, there have been 17,941 loans originated on Prosper.  1,908 were AA.  That is 10.6%.
Looking at the last 3 months (10/16/07-1/16/08), there were 2,678 loans, of which 359 were AA.  That is 13.4%.
Looking at the last month (12/16/07-1/16/08), there were 927 loans, of which 125 were AA.  That is 13.5%.

The article invites comments.

I provided pretty much my standard comment, although the comment software butchered it up.

[sb92075]

Code: [Select]

mysql> select count(*) from Loan where CreditGrade = 'AA';
+----------+
| count(*) |
+----------+
|     1914 |
+----------+
1 row in set (0.04 sec)

mysql> select * from dq;
+-------------------------+
| Status                  |
+-------------------------+
| Defaulted (Delinquency) |
| Late                    |
| 1 month late            |
| 4+ months late          |
| 2 months late           |
| Repurchased             |
| Defaulted (Bankruptcy)  |
| 3 months late           |
+-------------------------+
8 rows in set (0.02 sec)

mysql> select count(*) from Loan where CreditGrade = 'AA' and Status in (select Status from dq);
+----------+
| count(*) |
+----------+
|       61 |
+----------+
1 row in set (0.04 sec)


The default rate on its highest rated, AA-loans is 0.27 percent -- in line with brick and mortar banks.

61/1914 = 3.19%
I guess it depends upon what is classified as a "default".

Code: [Select]

mysql> select Status, count(*) from Loan where CreditGrade = 'AA' group by Status;
+-------------------------+----------+
| Status                  | count(*) |
+-------------------------+----------+
| 1 month late            |       12 |
| 2 months late           |        6 |
| 3 months late           |        6 |
| 4+ months late          |        9 |
| Current                 |     1359 |
| Defaulted (Delinquency) |       11 |
| Late                    |        4 |
| Paid                    |      482 |
| Payoff in progress      |       12 |
| Repurchased             |       13 |
+-------------------------+----------+
10 rows in set (0.04 sec)


[ira01]

61/1914 = 3.19%
I guess it depends upon what is classified as a "default".

In fairness to Prosper, I don't think you can count some of those categories (especially "late" and "repurchased") as a default.  I think you should count the two types of "defaults," 4+ lates, and 3 month lates as defaults.  Even that is pretty generous, because the vast majority of the 2 month lates will also default, but even so, that seems fair. 

Out of curiosity, what if you only count true "defaults" (both by loan numbers and by dollar amounts)?

[sb92075]

Here are the raw numbers.
You can perform what ever calculations you deem appropriate & post the results back here.

Code: [Select]

    -> ;
+-------------------------+----------+---------------------+
| Status                  | count(*) | sum(AmountBorrowed) |
+-------------------------+----------+---------------------+
| 1 month late            |       12 |              194050 |
| 2 months late           |        6 |               65495 |
| 3 months late           |        6 |               72800 |
| 4+ months late          |        9 |              117250 |
| Current                 |     1359 |            13162099 |
| Defaulted (Delinquency) |       11 |              150500 |
| Late                    |        4 |               53001 |
| Paid                    |      482 |             2744315 |
| Payoff in progress      |       12 |               87941 |
| Repurchased             |       13 |              131400 |
+-------------------------+----------+---------------------+
10 rows in set (0.04 sec)

mysql> select  count(*), sum(AmountBorrowed) from Loan where CreditGrade = 'AA';
+----------+---------------------+
| count(*) | sum(AmountBorrowed) |
+----------+---------------------+
|     1914 |            16778851 |
+----------+---------------------+
1 row in set (0.04 sec)
11/1914 = 0.5747%
150500/16778851 = 0.89696%
"measurably" greater than those stated by PMI

[ira01]

Here are the raw numbers.
11/1914 = 0.5747%
150500/16778851 = 0.89696%
"measurably" greater than those stated by PMI

Thank you.  I will try to crunch these in various ways later.

But certainly any way you slice it, the numbers stated by Prosper are (once again) BS.  I do wonder, however, if perhaps the interview for that article was before the last debt sale, so the number and amount of defaults would be much lower than now.