zcommodore

It's been more than a month since Prosper decided to eliminate group leader fees and it looks like Prosper is doing just fine.  One group leader started a thread presumably to document the slow death spiral that Prosper was entering into by removing these fees.  His methodology seemed a bit off since he was checking daily how many listings were fully funded compared to how many were available and how many groups there were.  At any rate, this would completely miss all autofund loans that filled and many listings that only filled in the last few hours.

I decided to do some checking of my own by seeing how many loans originated each day and used the performance tab to go back a few months to get some historical data as well.  Here is what I came up with:

The graph was a bit noisy looking at each day so I added the 10-day average line to smooth things out a bit.  Note that the average is over 10 business days which usually corresponds to two weeks except when a holiday occurs.

The change happened on the night of September 11 so all loans from listings that started on September 12 or later were without group leader fees if the borrower was in a group (unless the group leader had a listing under review that was approved and funded from prior to the changeover.)

Looking at the graph, it looks like there was a temporary low right after the change but since that time, the number of loans approved has been on a slight upward trend.  Obviously it's still a relatively short period of time but in my opinion it is anything but a death spiral for Prosper.

There are rumors flying about new information provided by Chris Larsen at the Hawaii Meet & Greet that I mentioned in my last post. One of them is the possibility that Prosper will raise their rate caps to 36% from 30%.

Personally, I think this is a bad idea. If you look at my previous blog post, you will see where I mention how default rates appeared to drop significantly in the months where Prosper's rate cap was a lot lower (24%?). Additionally, when I look at the 3rd party sites at the early loans made at 30+% rates, very few of them are still current. A few paid but many (most?) are defaulted or headed that way.

High rates attract newbie lenders who don't understand risk. A defaulted loan affects ROI a lot more than lower rates. More bad loans will likely be made if/when this happens and it's going to hurt Prosper due to the higher default rates I believe.

I suppose Prosper may be trying to make HR borrowers more attractive again to lenders since by every measure I've seen, HRs are a losing proposition even at 29%. It doesn't really matter what rate is offered if the borrower can't pay. 36% of 0 is still 0. Even if a borrower can pay for a few months at those rates, it can take quite a few months before a lender even breaks even and few HR borrowers can hold out that long.

As I said before, 36% rates are a bad idea.

The last couple of months, I've posted the following graph showing how lenders have changed their bidding patterns over the past few months, especially since the new extended data came out in February.

As you can see, HR borrowers fell to their lowest percentage of all loans funded in the month of September at approximately 7% of all loans funded for the month. This is really a very low number considering the fact that September had the fewest number of loans in quite a few months.

After looking at this graph, I decided it was too busy and needed to be condensed to be more meaningful. After reading a few of Prosper's press releases I noticed they grouped the credit grades more as follows: AA-A, B-D, E-HR. I decided to do the same thing and came up with the following:

Note that in this graph, all months up to and including Feb-07 include NC grade borrowers in the E-HR category. Prosper eliminated NC in February so it is not included in later months and I chose not to include it in the key names.

When I first posted the previous graph, I got a note from Islandmele in Hawaii saying she would like to use an updated version in a presentation at her Meet & Greet she was planning. Last week, I forwarded these two graphs to her and I believe she used at least one of them in her presentation earlier today. She asked me what I would say if I were presenting these graphs so I wrote the following few paragraphs, mostly referring to the second graph since it is easier to read.

A knowlege of Prosper's history is useful when looking at these graphs. There is a dip in the high-risk categories in April/May '06 and I believe it coincided with a significant reduction in the top borrower rates available on Prosper.

Later, a lot of publicity brought in a lot of new members to Prosper later in '06. In addition, groups started getting rated in October. Once there was a rating system, it added incentive for group leaders to get lots of loans funded so they could reach the top of the list. This lead to "pump & dump" and lots of high risk borrowers were funded due to newbie lenders chasing high rates and group leaders taking advantage of this fact.

By February, when the lates started rolling in and Prosper added the new extended credit data, lenders figured out that high rates don't always mean high returns. The highest risk borrowers stopped getting funded since lenders saw the risk just wasn't worth it and focused more on the better mid-credit grades where the perceived risk was lower and the returns were more consistent.

It will be interesting to see how things change going forward, especially since there is very little incentive for groups to "pump & dump" now that there are no longer group leader fees associated with loans in their group. Ultimately, I think it will be good for Prosper even though the volume of loans has continued to drop. The ones getting funded now, for the most part, are better quality overall so that should help default rates even with new lenders arriving who don't understand the risk.

One other thing to note that I've hinted at before and that is November-January is when the groups Financial Assistance Network, Two Millionaires, and to a lesser extent, Fairplay and Life Changing Loans, were operating at their peak. These groups seemed to feed off of the highest risk borrowers and encouraged a lot of inexperienced lenders by their pumping and dumping to bid on these high risk borrowers. At least one large lender, reguyincali, also was bidding on a lot of these high risk borrowers and large lenders can affect the market quite a bit since the marketplace is still relatively small.

I am glad to see the better borrower grades getting a larger piece of the Prosper pie and I hope that will help default rates to go down and encourage more lenders to participate and help Prosper grow.

A number of people have mentioned on the Prosper forums and maybe some blogs that the total number of loans dropped in September by quite a bit. Here is a graph illustrating that point:

I suspect there are a number of reasons for this drop and it may be partly due to the removal of group rewards. I don't think that is a bad thing though since it may be better for default rates in the end in my opinion.

Anyway, one thing I noticed is that the total number of business days in September was relatively low since the month started and ended on a weekend. I decided to check my theory by counting the number of business days (excluding weekends and holidays) for each month and plotting the average dollar amount of loans made per business day.

As you can see, the number peaked in April, similar to the previous graph, but was very close to flat between August and September. Honestly, I don't know if it means a whole lot but it was an interesting way to look at things. I am curious to see October's numbers since that will be a full month without group leader rewards.

One more graph I made that isn't quite as flattering though was the total number of loans made per business day.

As you can see, the number of loans made per day did drop even when considering there were fewer business days in September.

On a lark, I created one more graph showing average size of all loans each month.

I honestly am not sure what this graph really means other than the fact that bigger loans were approved on average in September than they were in August.  The overall uptrend in the graph does seem to be good for Prosper since it makes more money from bigger loans. Now they just need to find a way to get more loans funded from month to month instead of the downtrend we've seen over the last few months.

There are literally thousands of groups on Prosper, most of which are 0-0-0 (0 members, 0 listings, 0 loans).  In a post by LoanChimp a short while ago, he indicated he got a notice that Prosper is closing his inactive group.  He started it to bring a couple of friends to Prosper but they didn't need to go that route so they never showed up.

I guess this is just the next step for Prosper to make as they move away from the group model.