Friday, September 28, 2007

Friday Linkfest Junket

I've been having a linkfest every Friday. Why stop now?
And on we go...

TechDumpster has a rundown on the (hypothetical) Zopa versus Prosper match-up:

Although Zopa is the only service of its kind in the United Kingdom,
when they do finally make it to the United States, they are going to
have to go head to head with Prosper. Here’s a quick breakdown of
the main differences between Zopa and Prosper

Stephen Oaks Of Juita Group interviews Prosper CEO Chris Larson with a Q&A session:

Following the conversation, I thought a bit about my first impression of Chris. I can tell you first hand, the guy is a straight shooter who knows the lending business inside and out. What separates his company from the typical establishment is his love for the individual and community.

“Peer-to-peer borrowing/lending is the purest form of capitalism,” said Larsen.

Wednesday, September 26, 2007

The Next Forum Revolt Brewing Over Fraud

It's looking like the next big forum uproar will be over Prosper's actions in the forums to stamp down the revelation of personally identifiable information (PII). Where does this information mostly come up? Why, when lenders go hunting for fraud. This has coalesed around a posting by Prosper Shira reinforcing Prosper's current PII position:

Members have done a terrific job helping Prosper identify and prosecute fraud and we appreciate your desire to alert fellow members on suspected fraud. After reviewing the numerous discussions about fraud and personally identifiable information (PII) occurring on forums over the past few weeks, I’d like to take an opportunity to communicate directly with you about this important subject.

The best way to alert Prosper to suspected fraud is to click the “report this listing” link located on every Prosper listing. Your reports go into a priority queue that is monitored daily by Prosper’s trust and safety team and every report is taken very seriously.

Prosper has the right to cancel any listing at any time, and if we find the reported listing is fraudulent we will exercise that right. Our investigations may take several business days to complete, and an open listing under investigation could become fully funded on the site by the time it is cancelled. When a listing is cancelled, we send a cancellation email to the bidding lenders informing them of the cancellation. Sometimes reported listings are cancelled for reasons other than fraud. If a listing is cancelled due to fraud, the borrower will be banned from Prosper.

If you choose to post your suspicions about a specific listing or member on the forums, please do not include a member’s personally identifiable information (PII) or links to any website that reveal the member’s PII. Even when a listing or member appears to be fraudulent, Prosper has an obligation to protect our members’ personally identifiable information on our forums. If you have PII information linking a member to fraud, please include it in your report to Prosper, and not on the forums. Creating posts revealing PII on the forum will result in a seven-day suspension for the poster for first time offenses. Members with repeated offenses will be banned from the forums and could have their Prosper registration terminated.

To further protect a member’s identity, Prosper cannot share specifics back to those who report fraud or on the forum talking about what action, if any, was taken. However, you have given us great suggestions about how to improve our communications both back to the original reporters and forum participants. In the next few site releases, we will update our onsite and outbound messaging as well as forum policies to reflect your suggestions.

There's some very good posts from Ira, 112233, and others, but xraider seemed to sum it all up:

Shira, why is Prosper seeking to take all-important fraud detection away from those who do it best? Prosper already relies on lenders to report suspicious listings, and it looks like Prosper is trying to defang lenders' ability to (1) detect fraud (as Ira said, it is a collaborative effort); and (2) having detected fraud, warn others about it.

Prosper, more and more, appears to be a safe harbor for scamming borrowers. Is Prosper going to do anything to give lenders any assurances that with watered-down fraud protection we can have some certainty that we're lending to people who really will pay us back?

Also, what is the justification for preventing lenders from sharing publicly-available information? Prosper has eliminated the use of names, email addresses and telephone numbers in listings, taking that tool from lenders. Now Prosper is saying that if, somehow, lenders find fraud through other means (such as FaceBook, or MySpace, or eBay) that information should not be shared? Why not? If it's publicly available, there is no privacy to protect.

I would love to know Prosper's reasoning for banning the sharing of publicly available information.

Implied Call Option In Prosper Loans

Tanta over at Calculated Risk has gone off an obscure topic for mortgages, but it has relevance for Prosper lenders and borrowers. She had previously touched on the idea of an embedded option in a previous post, but really rounds it out here. From her A Clockwork Mortgage post:

... so suffice it to say that the main issue is the imbedded options in mortgage loans. In options theory terms, a mortgage gives the borrower a put (the right to default or "send jingle mail") and call (the right to prepay the mortgage with proceeds of a refinance or any other funds). Although both options have costs--especially if you have a prepayment penalty on your loan--those costs can end up being much lower than the cost of keeping your current mortgage. Because the "strike price" of these options is so heavily dependent on local and national economic conditions, interest rate levels, and home price changes, it is notoriously difficult to predict for any given borrower over any given stated loan maturity.

The lender, on the other hand, is "long a bond, short an option." Mortgages cannot be called or accelerated by the lender, except in case of default. No lender can make you refinance ... On the other hand, if market rates drop, you may exercise your right to prepay, which means that the lender loses your old higher-rate mortgage and must reinvest its funds at new, lower market rates.

Tanta compares the borrower's ability to prepay a loan to there being a call option built into the loan. The borrower can exercise the option to pay down their loan ahead of schedule any time they want. In typical options markets, the seller of the option (the lender) is usually compensated by the option buyer (the borrower) for the option.

I've brought this all up because Prosper allows prepayments and it's an interesting mathematical way to view prepayment risk. There is an intrinsic value to the prepayment option, and it's value should be captured. Because of this, I'll be adding a prepayment risk factor to my Prosper loan value modeling.

Saturday, September 22, 2007

Prosper VP To Toughen Up On Late Payments?

I completely missed this. Fortunately, Lending Stats didn't. From their September e-mail only newsletter:

Prosper Tackles Collections

Prosper is pleased to introduce our new VP of Operations, Doug Fuller. Doug brings 25 years of experience to the Prosper team and will be instrumental in helping our team crack down on delinquencies and ramp up our collections efforts. We are excited to have Doug on board; he will help us to continually make Prosper a safer and more reliable marketplace for all users. Doug is a huge asset to the Prosper team and we look forward to updating you on his systemic improvements in Prosper operations.

Some enterprising Googlers on the forums found a possible candidate and a possible resume. IF (and that is a big if at the moment) this is the right guy, he has some interesting default handling credentials. I'm cautiously optimistic.

First Select Corporation/Providian Financial, 1999 to 2002
Senior Vice President FSC – Corporate Analytics & Process Management
  • Overall responsibility for the development and refinement of First Select's collections strategy as well as the processes and infrastructure necessary to support that strategy.
  • Development, implementation and monitoring of decision models.
  • Development team for Recovery Data Warehouse.
  • Development and implementation of new Recovery Management System.
  • Led joint development process with Ontario Systems to create “next generation” collections and recovery system. Included automated workflow, embedded decision rules, integrated process test and evaluation capabilities, and self-documentation of account history. All committed dates met or exceeded. 1.3 Million accounts transferred with no loss of production time. First week productivity improvement of 37% for outbound phone reps.
  • Design and implementation of Phone Channel Call Prioritization and scheduling system. Results in a 3x increase in “Promises to Pay” per outbound dialer hour. At the same time, resulted in a 30% reduction in outbound calls placed.
  • Development of Portfolio Pricing model and methodology as well as “Process Yield Model” to forecast and monitor account movement and revenue production through process.
  • Design and implementation of Recovery Data Warehouse. Received assignment after previous attempt costing more than $2MM and 24 months effort was declared a failure. Prototype live in less than 10 weeks. Production Model live three months later. All committed dates and budgets met or improved.

Friday, September 21, 2007


It's Friday and I've had a long week at the day job (hence, few posts), so now I'm just lazy. Let the linkfest begin...

Lending Club comes out:

This is it. The great moment we have been working towards: expanding Lending Club beyond Facebook and making it available to anyone out there looking for better rates. We continue to enhance and support our Lending Club application in Facebook, but now, users don’t have to be in Facebook to use our site.

As part of our public launch, we are unveiling new features that will benefit the whole p2p lending community, whether you are a Facebook user or not.

Prosper Lending Review spots a Newsweek article on Prosper:

I find it interesting that peer to peer loans are compared against credit cards instead of banks or home equity loans. Matt wrote a good article back in July about when it makes sense to borrow from Prosper - Why would a borrower use Prosper instead of a traditional bank? He did not consider credit cards in his analysis but my gut tells me that credit cards are only going to beat banks or Prosper if you have a promotional rate. Generally credit cards are not a good place to carry a large balance. In fact, my personal recommendation is to never carry a balance on credit cards.

Press Piece - Borrowers ask cyber-friends to fund debts, dreams:

Never lend money to a friend, the old saying goes, but what about lending to a stranger a few keystrokes away?

The emergence of online communities has prompted some people needing funds to pay off debt or pursue their dreams to turn to cyber-friends to help them, rekindling people-to-people lending and investment opportunities that sidestep banks.

zcommodore looks at the percentage of funded listings in groups:

Periodically, I check how many listings there are on Prosper (2220 last check a few minutes ago) and how many are not in a group. Then I get out my calculator to find out the percent of "no group" listings relative to all available listings.

Wednesday, September 19, 2007

Studying The Lending Market Summary

I've had a few days to ruminate over the Prosper's August lending market summary, and I've found a few items that are noteworthy. Specifically, their "Prosper Select" criteria and the top Prosper borrower states.

The Prosper Select criteria were listed in the fine print at the bottom of the release:

The Prosper Select Index return is the estimated average annual return on invested principal, based on actual delinquency performance to date. The Prosper Select Index includes AA - E credit grade loans for borrowers whose credit reports at the time of application indicated zero current delinquencies, three or fewer credit inquiries, and a debt-to-income ratio of 40 percent or less. The annual return period reflects loans originated in the twelve month period ending one month prior to the observation date of August 31, 2007. Prime Select includes AA and A credit grade loans (credit scores of 720+). Near Prime Select includes B, C, D credit grade loans (credit scores between 600 and 719). Sub Prime Select includes E credit grade loans (credit scores between 560 and 599).

The Prosper Select criteria was DTI less than 40%, 0 current defaults, and 0 to 3 credit inquiries. This was of mild interest until I noticed that the sub-prime category was returning 14.95%. Then I became really interested. The table below compares the Prosper Select ROI versus Prosper's overall ROI.

Prosper Select
Prosper In General
That is a huge swing in ROI for the lower credit grades, implying an equally huge delta in default rates. There may be something to this whole Prosper Select thing.

As I had previously noted, Prosper doesn't have equal access to the lending markets in all states due to usury laws. I had ranked the 20 most populous states based on a quick estimation of Prosper's ability to service the market. They confirmed the jist of my post when they listed their top states:

Top Five Prosper Borrower States in August 2007

  1. California
  2. Georgia
  3. Illinois
  4. Ohio
  5. Florida

I ranked California, Georgia, and Illinois as excellent loan opportunities with their maximum loan rate of 30%. Ohio was also ranked excellent but had a maximum loan rate of 25%, which appeared to be sufficient to slow it down. Florida, despite having twice the population of Georgia, was 5th on Prosper's list. I suspect that this is because of their maximum loan rate of 18% and only achieving a "good" rating. Any guesses why the "marginal" rated Texas (2.5x Georgia's population) and New York (2.1x Georgia's population) failed to make the top 5 list?

Saturday, September 15, 2007

New And Shiny Instant Transfer

It was like magic. Just after I commented that Prosper should give lenders a hint about whether a bank transfer qualifies for the rapid transfer, it magically appears in an update two days later. It's not called rapid transfer anymore, though. It's called "Instant Transfer", but all the other good stuff I was looking for showed up.

The new interface for transferring funds calls out the dollar range that qualifies for the instant transfer and indicates if the current transfer will qualify with some cute icons down at the bottom of the page. Points to Prosper for taking this feature out of "beta" (it was starting to feel like Google around here) and grabbing the quick and easy interface upgrades.

Thursday, September 13, 2007

Prosper Growth And Usuary Laws

As can be observed from my analysis of Prosper's August income, they need to do some growing if they intend to survive long term. Prosper's ability to expand it's lending has been limited by the various state usury laws that limit the interest rates that lenders can fetch. This has been highlighted by Prosper Lending Review's commentary on the Pennsylvania rate cap and one borrower's attempt to get a loan despite the state's cap. The question, though, is how much can Prosper grow it's available market by navigating the various state usury laws.

To answer this question, lets look the populations for all the US states (circa 2005). The top 20 states represent about 75% of US population, and improving the borrowing and lending options in any one of these would be a tremendous boost to Prosper's available market. I've ranked each state based on Prosper's current loan limits with ratings of Excellent (loans above 20%, AA-HR likely to fund), Good (loans 20% - 17.5%, AA-C likely to fund), Marginal (loans 17.5% - 12.5%, AA-B likely to fund), or Poor (loans 12.5% - 10%, only AA's likely to fund), Very Poor (loans under 10% - good luck getting anything). The ratings were roughly derived from Prosper's 30-day interest rate average.

Max Rate
(Personal) 10%
(Business) 18%
New York
Very Poor
New Jersey
North Carolina

It's important to remember that a majority of Prosper's loans (dollar weighted) come from the B - D credit ratings, and, from the analysis, it's obvious that Prosper has a fair number of pick-up opportunities if they can get their legal ducks in a row for the states that require additional licensing to loan at higher rates (like Pennsylvania). After all, there's almost 77 million people in those marginal states.

Another Prosper Upgrade

Prosper's performed their September upgrade last night. As someone named Andrew (Prosper Andrew perhaps?) commented in a previous post, the rapid transfer is now functional and will be called "Instant Transfer". Prosper's also implemented their much talked about changes to group rewards as well as some additional changes. From the announcement forum post:

Instant Transfers for Lenders

After several months of beta testing transfers with little to no delay (I think it was called "rapid transfer" at the time), and after your copious amounts of feedback, we've built what is now known as "Instant transfer" into a proper feature, which indicates when you're eligible for an instant transfer, and makes the money available instantly (no more 15-minute wait).

Here's how it works - when a new transfer to Prosper meets the requirements below, funds are made available in a lender's Prosper account instantly, with no delay:
  • Transfer amount must be between $500 and $10,000
  • Sum of pending transfers must be <20%>
For example, if you had an active loan value of $10,000, you could transfer between $500 and $2,000 to your Prosper account instantly. If you attempted to transfer $3,000 to Prosper, it would not qualify for instant transfer because the amount is more than 20% of your active loan value. Your $3,000 transfer would still continue, but would be subject to the standard 4-day waiting period.

Group changes

As we announced a few weeks ago, we are making some changes to the group model to more closely resemble our original vision for the concept. As of this release:
  • Group leader payment rewards: These will no longer be available to group leaders. Listings created on Sep 13, 2007 and after will not be eligible for group leader rewards.
  • Search results changes: Group names and ratings have been removed from search results. When doing research with new lenders, we found that introducing groups in the search results introduced confusion and blocked lenders from proceeding to find listings easily. You can still search on specific groups using the advanced search feature.
  • Group categories: We've discontinued the use of group categories. We will eventually introduce a more intuitive set of loan categories (debt consolidation, business financing, student loan, etc.) for listings shortly.
  • Personally identifiable information: In the interest of group member privacy, we've started removing personally identifiable information from group membership requests.

Prosper Andrew also notes that instead of one huge upgrade, they'll be dribbling the upgrades out slowly over time.

Wednesday, September 12, 2007

Prosper Monthly Market Summary

Prosper has published their first monthly Peer to Peer Lending Market summary. The market survey, surprisingly, covers just their own lending market. There's some interesting statistics that we'll digest in the next few days, but I want to highlight the market commentary provided by Chris Larson, CEO:

The market turmoil stemming from the ongoing credit crunch, subprime mortgage meltdown and housing value slump naturally begs questions about what impact this market environment is having on the Prosper marketplace. In a nutshell, we would categorize the impact as broadly constructive for Prosper lenders and prime and near prime borrowers.

As consumers are being hit with the evaporation of introductory credit card rate offers and home equity loan options, Prosper is becoming an even more attractive financing alternative, particularly for more creditworthy borrowers.

At the same time, lenders on Prosper are exhibiting rational behavior by steering their bids toward borrowers in the higher credit categories and being far more cautious about chasing higher rates offered by subprime borrowers. Evidence of this flight to safety is seen in Prosper's mix of funded borrowers. For example, the subprime category accounted for only 9 percent of loans funded in August 2007, a marked decrease from August 2006 and the 2007 year-to-date average of 25 percent and 14 percent, respectively.

That's an interesting spin on the sub-prime melt down. Beyond that, the statistics are interesting for observing already existing trends, but do not appear provide insight on where a lender should be bidding. It's an interesting read for a stats junkie, and Prosper provides a way to sign-up to their mailing list (I'm a sucker - I did it).

On a public relations front, however, this looks like a good move for Prosper, so I'll give it a cautious thumbs up until a few more issues come out.

Update: If you're wondering why it hit all the blogs about the same time, Prosper's Communications Director e-mailed the release out to us. That being said, here's a few more commentaries on the Market summary (updated as I spot them):
  • Prosper Lending Review: "All things considered, I think that this monthly data report from Prosper will be valuable. It's good to hear Chris Larsen's analysis and it gives a regular public way for Prosper to report their growth and success."
  • Prosper Forum
Update 2: Prosper has finally posted the press release to their website (and formatted the tables nicely).

YALF - Yet Another Linkfest

RateLadder spotted that Prosper Andrew requested a loan:

This is a $10,000 Loan that started at 15%. It is currently @ 7%. While I have every faith that Andrew will repay the loan. I have a hard time justifying the current rate. The highest yeilding 3 year CD is currently an APY 5.41% from Eloan. I will try to snipe $50 from my standing orders and bid $50 at 5.5%.

Prosper Lending Review talks about the interest rate cap in Pennsylvania

Pennsylvania, for example, is the sixth most populous state with over 12 million people. Based on their state lending laws, Prosper only allows loans to borrowers residing in Pennsylvania at 6% and below. Rather than helping borrowers get a good interest rate the practical effect of this legislation has been to prevent borrowers from obtaining a loan. According to LendingStat's loan breakdown by state, only 21 loans have been made to Pennsylvania borrowers ranking them a distant #42 despite their large population. Almost all of these 21 loans are for the minimum loan amount - $1,000.

ProProsper has been updated to use the newly released Listing-Loan pairing:

A nugget that I missed from the Mid August Prosper Upgrade is that the loan and listing objects in the API are now tied together with a Prosper provided listing key on the loan object. While I had achieved 99.8% loan to listing matching, I was still missing a few.
One borrower gets creative to work around his state's rate cap and get a loan (hat tip RateLadder):

Let’s make this interesting. I have a loan listing on Prosper. It is capped at 11% because of a state cap in Virginia. 11% is not competitive, so there is a high probability that the loan will NOT be funded. How can we get this funded? How about a double interest incentive:

Watch out for falling home sales in California:
DataQuick will report SoCal home sales in about a week (NAR reports U.S. August existing home sales on Sept 25th), but the following story in the LA Business Journal suggests sales fell off a cliff: August Home Sales Take a Major Plunge

Tuesday, September 11, 2007

Rapid Transfer Behavior Confirmed

A while back, I disclosed the rapid funds transfer criteria after a Prosper lending services rep was kind enough to tell me (that'll teach him). I can now confirmed that it works roughly as outlined. Some time after I passed $2500 in outstanding loan balances, I transfered $500 into my Prosper account. I got a notice that the money would be available in the usual 4 business days. Later that afternoon, the money was available for bidding.

This all happened automatically without any additional action from me. I did not test the maximum transfer limit (20% of loan balance) - sorry. The one frustration with the experience is that Prosper's website does not provide any guidance or information. I cannot find any reporting that shows my account having rapid funds transfer enabled. On the money transfer page, they do not hint that the feature is available or what the minimum and maximum qualifying transfer amounts would be. Small details, but not very hard ones either.

Monday, September 10, 2007

How Prosper Gets Paid

In addition to all the intricacies of being a lending on Prosper, I also think it's helpful to understand how Prosper makes it's money. After all, Prosper requires cash to pay it's employees and stay in business, and it's in the lender's best interest for Prosper to keep it's doors open. And, to make things more fun, they give us all the statistics necessary to determine their corporate income.

Prosper lists their fees on the website. They classify them as borrower fees (closing costs) and lender fees (servicing fees). The borrower fees are very similar to the origination fee on a mortgage, where the borrower pays a percentage of the loan amount to the lending company to cover the cost of establishing the loan. Borrower fees are 1% for AA-D loans and 2% for E-HR loans. The servicing fees pay for the ongoing cost of collecting the loan payment every month, dividing up the payment to all the lenders, and actions (if any) Prosper takes to collect on late loans. It is collected as a percentage, 0.5% annually for AA-A and 1% annually for B-HR, of the outstanding loan balance and comes out of the lender's share of the earned interest.

It is possible to calculate Prosper's monthly income based on the readily available loan statistics. To get the borrower fees, sum up all the loans that originated during the month and multiply by the borrower fee rate. The lender fees are more complicated to get perfectly accurate, but a close approximation can be reached by looking at the outstanding loan balances at the beginning of a month and multiplying by the lender fee rate. The fees collected for August are shown below:

Credit Grade
Loans Originated
Borrower Fee

Credit Grade
Outstanding Loans
Lender Fee

That's right, Prosper made around $118k in August.

Sunday, September 9, 2007

Timeless Money Tips

Money magazine highlighted 20 top money tips (hat tip to Free Money Finance), and there's good wisdom in here for Prosper lenders. It's all worth a read, but here are a few that are most relevant:

#1 Be Humble
Money cites Confucius, "When you do not know a thing, to allow that you do not know it--this is knowledge.", as a reminder that you should accept that you don't know everything and to keep questioning when things are unknown. One proven way to find out what you don't know about Prosper is to post on the Prosper forums or the forums.

#2 Take Calculated Risks
To potentially get higher returns, you have to take on more risk. That is not advocating chasing the highest risk item you can find. Make sure that you understand the risks involved. The Prosper lenders who jumped into HR loans without understanding the risks got burned the hardest.

#4 Mix It Up
As in diversify. Don't put all your eggs in one basket, or, in the Prosper case, don't put all your cash in one loan. Even AA loans go bad, so to insulate your loan portfolio from the inevitabilities of owning loans, make more loans with smaller amounts to prevent one bad loan from parting you and your cash.

#7 Practice Patience
It's very tempting to want to bid even if the listings aren't up to your usual standards. Resist that temptation. Be patient and wait for listings that meet your standards before bidding.

Friday, September 7, 2007

Forum Whomping Over Group Rewards

After writing about Prosper's change in group leader rewards, I chanced a venture onto the forums to throw my thinking into a no holds barred environment. It'd be charitable to say that I got knocked around a bit. It starts with my post:

The good groups help guide borrowers with help for their listing descriptions. I don't have hours to play back and forth with borrowers trying to sooth out the information that's in the average listing from a No Bulls borrower. That's a value-add for me as a lender. I'm expecting to see the listing quality drop as the good group leaders exit and borrowers flounder around.

Shenendoah replies:

If a borrower doesn't know how to write a decent description, I want to know that. I don't want a GL to make it look pretty.
If a borrower does not know how to do that, I don't go back and forth with them at all, I simply pass on their listing and find another.

Groups are a good thing for borrowers, but not for lenders. The so-called "good" groups have gotten a lot of loans funded that never should have been funded.
It goes down hill for my logical thinking from there. I do recommend the forum thread as good read on the topic. I was most impressed with Penelope, who provided a solid reply that characterized the group problem:

If a less than attractive loan is put up, you have the option of questioning the borrower.

If that same unattractive listing is spit shined by a GL, you feel more confident in it.

We have witnessed GLs encouraging borrowers to state all types of false information to make the listing look good. Massaging DTI, couching on % to get SOs to hit, etc.

The fact remains that there is just another layer of potential BS you have to wade through.

She also provided a link to a forum thread on the issues with group rewards and community payments. I suspect that I'm going to have a change of heart coming soon.

Thursday, September 6, 2007

Group Rewards And Unintended Consequences

In an e-mail to group leaders (see RateLadder and Prosper Lending Review), Prosper announced their intention to discontinue all group leader rewards and move to a referral-only reward structure. The prevailing attitude on the lender side can be summed up with this Prosper forum post:

No group has added any value whatsoever.

Repeat it with me: No group has added any value...oh, well at least to lenders. And we are in the lender forum, right? You are a borrower.

I don't mean to cut on you but it gets crazy to keep allowing these untrue opinions go unchallenged. And I'm sure half a dozen folks will pounce on me (as a poster) and defend you (as a person), what? I'm still right. No group has added any value to lenders. And I don't care if other borrowers want to have threads called "Group Groups" or whatever. Go ahead. But that talk is for borrowers -- in here we are discussing the lender side of things.

My first reaction was "Oh %@#$". I understand the frustration, but I believe Prosper lending system will soon learn the law of unintended consequences. There were some groups that pumped and dumped their loans and provided no value. As a counter example, I always found listings from some groups (No Bulls, Have Money Will Bid, Bald Eagle Loans as examples) to be of much higher quality than average. These group leaders were using the group rewards to justify their time spent aiding borrowers, and that some borrowers do take a fair amount of time when group leaders verify utility and pay stubs. It's a value-add that I'm happy to bid on.

My expectation is that the strong, energetic group leaders will exit, reducing the listing quality in the process. Bald Eagle posted in the forums:

So, thank you Prosper. It has been a good run...

I will contact each of my current group members and give them my direct contact info. I hope to keep contact with them outside of Prosper and I believe in each and every one of them.

To the rest of the Prosper faithful, Adios!

I don't think this is what Prosper intended, but it will be a consequence of the change in policy and, should it go through, lenders will need to adjust to it. Be patient, and be ready to ask borrowers lots and lots of questions.

Update: Some comments from zcommodore, one of the blogging group leaders:

As a group leader who charges fees, I must say I don't really mind. I had originally intended to take no fees all along but a few people convinced me to charge for my services since I was spending a lot of time on it and they suggested that it would be reasonable to charge something so I did. I still waived fees whenever I had a borrower who was rate-capped but it never seemed to make a lot of difference since they never got funded anyway.

Update 2: I'll post other responses here as they show up.
  • Lending Stats Blog: "I think this is a huge step in the right direction..."
  • Prosper's Group Leader Forum thread
  • Wise Clerk: "While some group leaders did a good job screening and vetting borrower applications and the group leader could be seen as a compensation for time invested; the majority of lenders seems to see the removal of group fees as a step in the right direction."

Wednesday, September 5, 2007

Borrower Loan Status Reporting

One of the things that came out of Prosper's latest upgrade (see this linkfest for details) was increased information about a borrower's previous loans. I stumbled on a listing for a multi-loaner and found the interface very useful. I got a screenshot of this random listing (for an unreferenced borrower with all the credit info erased to protect the innocent).

This was a good example. Several useful bits of information are shown. The borrower took out a loan in Jan, 07 (denoted by the small dollar sign icon under the credit history section) and has improved their credit grade by one step (E->D) as (noted by the much larger icon and green area under the Credit History). The payment history for the previous loan is shown. What's missing is an indicator that the first loan was paid off. This information is available elsewhere, but it would be time saving for lenders to see it directly instead of hunting for it.

Overall, a welcome upgrade.

Tuesday, September 4, 2007

Higher Incomes And Lower Defaults

In the Prosper Forums, Risk_Reward wrote on incomes versus default rates and that post contained some interesting nuggets.

It has been widely known in credit circles that higher income decreases credit risk. One reference: Sexton D. E. (1977).

On the other hand, credit default risk decreases when the income
and age increase.

According to Carling K., Jacobson T., and Roszbach K’s study,
probability of credit delinquency increases when the income decreases.
So, we expect to find a negative relationship between income and
clients’ payback performance, assuming there is no inflation.

This is another statistics point backing my observation that increasing incomes decrease Prosper loan default rates.

Brain Dump And Back At It

I ended up taking the entire Labor Day weekend off from Prosper and blogging to spend some time with the wife and lil' one. As part of getting going again, here's a bit of brain clearing.

I've got several different threads in my head for research topics. Through the series of posts on accessing Prosper's internal default calculations (part 1, part 2, part 3), I got access to some pretty cool statistical information. It's time for me to get working on describing techniques to model default rates (like I said I would). There is also a meme running around (most recently pondered by WiseClerk and Jutia Group) that it may be possible to analytically discern a loan's expected default rate based on the listing description. I have some theories on how to evaluate this, and it'll take some off-blog testing to see if there's value here. Either way, I'll post results when I have them.

On the blog front, a few changes. I finally added a contact e-mail address ( if you have questions or comments that you want to keep off a public forum. I'm planning on at least addressing the non-Prosper P2P sites from time to time, and incorporating more information from personal finance blogs when it relates to lenders (investing and risk mitigation strategies) and borrowers (handling debt properly), even if it's not directly related to Prosper. The site layout may get messed with - it'll settle down eventually.