Whole loans

[ThinleyWangchuk]

When do they come online for individual investors (assuming institutional investors don't want them)?

[Fred]

After about 12-hr:

"a randomized subset of loans by grade will be available to purchase as a whole loan (i.e. not in $25 increments) only for a brief time period (12 hours), while all other loans will be immediately available for fractional purchase.  If the loans are not purchased as whole loans in the specified time period, they will become available for purchase in the standard, fractional manner."
(http://blog.lendingclub.com/tag/whole-loans/)

[AmCap]

Well wouldn't we want that disclosed...that the loan failed to sell as a whole loan?

[AmCap]

Just did a quick custom filter on IR...the results are counter intuitive.

[ThinleyWangchuk]

Just did a quick custom filter on IR...the results are counter intuitive.

What were your findings?
LC does disclose whole and fractional loans when the come on the platform.  One can filter by whole loan and fractional loan...

[AmCap]

No I meant disclose whether one kind has a worse track record. I filtered loans from Jan 2012 to today, grades D through G. Initial fractional loans had 6.2% loss. Initial whole loans with 10 or more lenders had 3.5% loss (I presume these to be whole loans that institutional folks didn't buy).  Initial whole loans with fewer than 10 lenders had 2.9% loss.  Baseline loss for all loans meeting those criteria was 5.7% 

[Randawl]

Keep in mind that those results are skewed due to the immaturity of the bulk of whole loans.  The whole loan program began late September/early October of 2012.

[AmCap]

I agree, I ran that analysis on my phone on the fly just to see. I'll narrow to sept 2012 when I'm home.

[Ran]

No I meant disclose whether one kind has a worse track record. I filtered loans from Jan 2012 to today, grades D through G. Initial fractional loans had 6.2% loss. Initial whole loans with 10 or more lenders had 3.5% loss (I presume these to be whole loans that institutional folks didn't buy).  Initial whole loans with fewer than 10 lenders had 2.9% loss.  Baseline loss for all loans meeting those criteria was 5.7%

My analysis also gives whole loan 2% advantage but I assume it means Institutional investors's model is 2% better than random pick

[brycemason]

I'll have a guest post on the whole loan program on the blog tomorrow.

[ThinleyWangchuk]

I'll have a guest post on the whole loan program on the blog tomorrow.

Could you share the link...

[AmCap]

Sorry team, to clean this up.  Running Sept-12 through today D-G notes, I see the following:

Fractional first: 3.9%
Whole (10 or more lenders): 3.5%
Whole (less than 10 lenders): 2.9%

Baseline loss: 3.8%

My point here maybe isn't to prove this to a mathematical certainty, but, if I understand the whole loan program correctly, it seems that we can show sum degree of funneling.  Put another way, I'm seeing an issue from a securities regulation point of view that they don't disclose that the poor quality Grade [X] notes are offered to the public as compared to the equally graded notes offered to preferred customers.

[Fred]

My point here maybe isn't to prove this to a mathematical certainty, but, if I understand the whole loan program correctly, it seems that we can show sum degree of funneling.  Put another way, I'm seeing an issue from a securities regulation point of view that they don't disclose that the poor quality Grade [X] notes are offered to the public as compared to the equally graded notes offered to preferred customers.

LC's phrase "a randomized subset of loans" could be in trouble with the regulators.

[brycemason]

Maybe the whole loans that actually got bought as whole loans had a higher default rate, such that the total whole loan pool was roughly equivalent to the fractional first pool. If institutions were targeting the highest yield stuff, and left the lower yield stuff to hit the fractional market, what you're seeing makes some sense. In any case tomorrow's article goes through a slew of borrower attributes and compares them across initial list status.

[AmCap]

Bryce I look forward to reviewing your work. I guess I am focusing more on this idea that within equal credit grades, retail investors are being offered notes with riskier traits.  There's no way I wouldn't disclose that in an S1 if I were LC. I am sure you will run a more rigorous statistical analysis.