Wednesday, October 31, 2007

Prosper Begins Slow Legal Road To Secondary Loan Market

As RateLadder has already spotted, Prosper has filed a Form S-1 with the SEC. In addition, a press release from Prosper also hit the wires indicating that this is, indeed, part of the effort for a secondary market. From the press release:

Following effectiveness of the registration statement, Prosper intends to establish and maintain a secondary trading market online auction platform, or Resale Platform, pursuant to which lenders may seek to transfer borrower notes to other Prosper registered lenders.

The registration statement relates to the Prosper Marketplace Lender Participant Rights, which represent Prospers obligations to seek to establish and maintain the Resale Platform and to provide various administrative, clerical, recordkeeping, identity verification, payment processing, collection and other services, associated with $500,000,000 in aggregate principal amount of Prosper Marketplace Borrower Non-Recourse Notes originated and to be originated on the Platform from time to time after the registration statement is declared effective by the SEC.

Tuesday, October 30, 2007

Lend2's Take On Prosper's Collections Handling

Lend2 runs a fairly large group (aptly named Lend2's group) with over 200 loans to the group's name. He wrote a summary of his borrower's feedback regarding the collections process:

  • The Penncro collectors are really "mean." Several people independently used this same word to describe them. Some have asked if there's someone else they can talk to about their loans. I've re-directed them to Prosper customer service.
  • Many people didn't know that they could make a partial payment on their loan (smaller than the monthly payment amount). There seem to be a lot of people trying to negotiate better terms on their loans. As we know, Prosper's loan notes don't allow this. Many borrowers are misinterpreting this to mean that they can't pay anything less than what is due. In fact, Prosper allows borrowers to pay as little as $25 against their loan.
  • Many people didn't realize that they could make a payment directly through the Prosper website instead of through Penncro. For many, this is easier or less threatening than having to call the collection agency.
  • Several thought that their loans had been sold to the collection agency, thinking that their payments would now go to the collectors instead of the private Prosper lenders who had originally loaned the money. These people brightened up when they realized that their payments would still flow through to us.
  • Most have received rude, obnoxious or threatening e-mails from Prosper lenders. In many cases, it sounds like this has negatively colored their view of Prosper lenders, making them less inclined to give their Prosper payments a higher priority.

His full post is worth the read, so have at it.

Big Prosper Upgrade Goes Live

Prosper's downtime last night brought several upgrades. This includes the unspeakable upgrades like portfolios and estimated default rates. I'll walk through the upgrades later in more detail, but for the moment, here's the highlights:

  • Estimated default rate is shown and a ROI calculator estimates the return for a given interest rate.
  • Bidding portfolios are now active
  • Tweaks with the standing order system (I'm ambivalent)
  • Borrowers can have second loans as long as their total balance is below $25k
  • Max interest rate is now 36%, state restrictions allowing
  • AA servicing fees are now 0%, A servicing fees are 1%
  • Expired listings can only be viewed if you bid on them (as lenders).
  • Collection agency information is broken out by credit grade
  • The API has a way to access extended credit information
And finally, some eye candy:

Update: Comments from other bloggers. It'll be updated as I see them.

Monday, October 29, 2007

Another Way To Monitor Prosper Collections

In a previous post, I looked at Prosper's pro-offered collections statistics. There is another way to look at their collections, however, that has the potential to identify quickly changing trends in Prosper's collections department. This will be useful if the new VP of Operations walks the walk from his talking the talk.

To do this, lets look at the life-cycle for a late loan. When a borrower misses a payment, they're flagged. Loans proceed to 1 month late, then 2 months late, 3 months late, and so on. If a borrower catches up, their loan drops out of the lateness pipeline. We can spot collections happening by observing how 1 month late loans in January become 2 month late loans in February and 3 month late loans in March. As some loans repair, they'll drop out of the aging.

Month Initially Late
Initially Late
(1 month late)
1 Month Later
(2 months late)
2 Months Later
(3 months late)
Repair Rate
January 07
February 07
March 07
April 07
May 07
June 07
July 07
August 07
Sept 07
October 07
If you read across the top row, on January 1st, there were 95 loans late. 1 month later (Feb 1), there were 83 loans 2 months late. This implies that Prosper's collections had repaired 12 of those loans. 2 months later (March 1), there were 82 loans 3 months late, suggesting that 1 more loan got repaired in the process.

Since we're working with raw results, we'll see emerging trends before they show up in Prosper's 12 month averages. I'll be watching the results for October very closely. If Prosper is making progress on improving their collections efforts, we should see signs of improvements in the September and October families of late loans. Nov 1 is just a few days away...

Friday, October 26, 2007

Fair Debt Collection Practices Act (FDCPA)

At times it's a bit confusing why Prosper does some of the things it does. Laws & regulations (national or state-wide) are one of the things that ties the hands of any company in the banking world. Five Cent Nickel has done a review of the Fair Debt Collection Practices Act (FDCPA), one of the dominant pieces of legislation that limits what lenders can do to collect on loans. Some of the topics are very reasonable (can't go over with a baseball bat and threaten their pets, for example) while others may be more controversial in the Prosper space (can't publish a list of debtors).

The highlights reel on restrictions comes from a Bankrate article:

Collection agents may not:

  • Call before 8 a.m. or after 9 p.m.
  • Talk to anyone but you (or your attorney, if you have one) about the debt.
  • Threaten to garnish wages or seize property unless they actually intend to do so. Garnishment is illegal in some states, and in others requires a court order. In many cases, property seizure is not permitted. Check with your state attorney general's office or state consumer protection office to find out what is legal in your state.
  • Threaten to sue unless they are actually taking legal action. In some states, third-party collection agencies may not sue.
  • Threaten you with arrest or jail.
  • Use obscene language.
  • Annoy or harass you with repeated calls.
  • Call at work if you have asked them to stop.
  • Falsely claim to be an attorney, a representative from a credit bureau or a member of law enforcement.

Also in there is a requirement to not publish a list of deadbeats or otherwise indicate that someone is in default unless you're telling a credit agency.

If you're wondering why Prosper isn't doing something related to collections, remember the laws and restrictions preventing them from doing something.

Update: Here's the FDCPA's full text if you need some help falling asleep at night. Also noted the anti-disclosure requirement.

End Of The Dryspell

I know what you're thinking. Hey Mike, make with the posts. I mean, you were cranking out at least one a day for a while and then the dry spell hit. What gives?

I'm just mentioning that I havn't abandoned Prosperous Land. My job occasionally requires me to think hard on math problems (nonlinear audio signal processing and distortion mechanisms, if you're curious) and that I'm not very good about juggling multiple things while doing math. Those are the breaks, I guess.

However, the hard part's about to end and the mind will be free to ponder Doug Fuller's ability to alter collections performance, the impact of Prosper's portfolio scheme on the lending market, and the latest outrage(s) perpetuated on the forums run by a company (whom shall remain nameless) that isn't used to herding cats.

As always, stay tuned.

Tuesday, October 23, 2007

Prosper's Poor Collections

It has been an axiom that Prosper's collections need some, shall we say, improvement. I think it's important to back up this axiom with some hard data. Fortunately, Prosper provides some useful statistics for analyzing their default rate.

The easy answer is to look at their collection agency statistics. I'm only going to look at Penncro's collection information as they handle a large majority of the collections for Prosper (~97%).

CurrentPast 12 monthsLifetime
Amount sent to agency: $1,015,661.09 $1,877,692.45 $1,881,095.05
Amount collected:
Net after agency fees



Accounts sent to agency: 1229 2265 2276
Accounts brought current:
In 1st mo. at agency
In mo. 2
In mo. 3
In mos. 4+
n/a 13.25%

The number of interest is that Penncro has brought about 13.25% of referred accounts current in the last 12 months. That's rather depressing. This implies that if a loan goes 1 month late and makes it to the collection agency, the odds are exceedingly high (86%) that it's gone.

The new VP of Operations was brought in to help clean up collections. Hopefully he'll have an affect.

Distress Rate Widens For High Dollar Loans

While doing a regular update on my personal lending tables (interest rates to charge versus various credit criteria), I noticed that the distress rate penalty for high dollar loans (>= $10k) has increased dramatically in the last two months. Distress rates refer to loans that are late in any possible way, either 15 days, 1-3 months, 4+ months, or in default.

In August, I calculated the following distress rates for loans under $10k and over $10k.

Loans Under $10k
Loans Over $10k

This has changed significantly in October.

Loans Under $10k
Loans Over $10k

Sunday, October 21, 2007

Can Borrowers Eat On $100 A Month?

There's been a running discussion on whether borrowers who claim that they spend almost nothing on month on food (see here, here, and here for recent random examples) are full of it. The credibility threshold seems to be around $100 / person. One of the examples claims $200 for a family of four. Is this doable?

The first thing to consider is that other personal finance bloggers seem to believe it's possible, though not very fun.

So, yes it is possible to eat without spending a fortune. Again, my food budget was radical by necessity, but the principles would still work today. I think $15/wk might not be enough now, but I think $20/wk would work, and I know that $30/wk would be fairly easy for a single person. For reference: $15/wk per person = $65/month for one and $260/month for a family of four. $30/wk per person = $130/month for one and $520/month for a family of four (which is about what my family spends on food now, and we don’t eat anywhere near the way I did back in the ’90s).

I would also add that a few politicians tried living on $3 a day for a week to help get in touch with the foodstamp experience. They didn't find it very much fun either.

What can you eat for $3 a day? Mostly carbohydrates. Oakland Democratic Rep. Barbara Lee's diet consisted primarily of crackers, a loaf of whole-wheat bread, tortillas, and brown rice. Assemblyman Mark Leno, D-San Francisco, filled up on 19-cent banana-and-peanut butter sandwiches. Rep. James McGovern, D-Mass., said he would've killed for a candy bar or a cup of coffee. "I've had enough lentils for three years. For us, this is an exercise that ends Tuesday. For millions of people, this is their life," he said.

The budgets used by the Congress-critters and the blogger above work out to $80 - $100 / month. This implies that if someone claims $100 / person or less, scrutinize their listing if you're otherwise interested or just move on. In these cases, I usually question the borrower as to how they've achieved their food budget.

I've only found one good answer in all my questioning, however. Consider this listing. Employees (or owners) of restaurants do get huge breaks on food and may often eat free at their establishments. That's not a free pass, but a mitigating circumstance that makes it plausible to only spend $100 a month.

Some other good links for your own evaluation:

Friday, October 19, 2007

Here A Link, There A Link, Everywhere A Link-Link

Old McLinker had some links, E-I-E-I-O...

Visible Banking did a video interview with Chris Larsen right after his Finovate presentation. Was he careful not to talk about features that may or may not be appearing in the future? You be the judge.

Peer-to-Peer Lending Offers Solution for Strapped Consumers:

A YEAR AGO, Nicole Newberry was in a financial hole so deep that she had trouble making the minimum payments on her credit cards. Worse, the then 22-year-old had gotten tangled up in the predatory cycle of payday loans. Every two weeks, when she repaid the two loans she owed, she borrowed the money right back to pay for groceries and diapers for her two toddlers. Her predicament was cruelly simple: "Each month, I was making all these minimum payments and getting nowhere," she says.

Then a co-worker told her about, a peer-to-peer lending web site that facilitates loans between strangers. Consumers seeking a loan list the details of how much they need and why, while those with cash to spare scour the listings and make loans to the ones they choose. Generally, borrowers get lower interest rates than they would with a bank or credit card, while lenders can earn better returns than they would in a money market or savings account

CircleLending is now Virgin Money USA (also here, and here):

Ever since Virgin bought CircleLending for $50 million earlier this year (previous coverage here), I've been looking forward to its launch. We hoped they might launch at our FINOVATE conference two weeks ago, but we lost out to the Mortgage Banker's Association's 94th Annual Conference in Boston, where Sir Richard Branson delivered the opening keynote a few hours ago.

Fred93 ponders on Prosper's late borrower collection rates:

I've been charting Prosper's collections statistics ever since Prosper first made them avaiable. Prosper sends loans to a collection agency when they are 1 month late, and the outcome after the collection agency's efforts are charted below.

The big picture is that Prosper's collection results continue to be scandalously bad.

Prosper and Zopa: Looking into the World of Online Consumer-to-Consumer Lending:

When most people are short on cash and need a few hundred to a few thousand dollars, they walk down to the local bank and try to persuade the loan officer to give them a loan. If the consumer doesn’t meet the bank’s cookie-cutter standards for who can get a loan, they’re rejected and the person’s just out of luck. Those who do get loans are often stuck with unfavorable terms and the only person that wins in the situation is the bank. Now two companies are hoping to empower consumers by offering services which will allow consumers to provide loans to each other online.

Internet cuts middleman in borrowing, giving:

What if you needed to borrow a couple of thousand bucks to remodel your bathroom but didn't want to pay 26percent interest on a credit card?

What if you had $2,000 but wanted to earn more than 2 percent on your money that you would on a passbook savings account at a bank?

Such is the idea behind, a Web site designed to anonymously link borrowers and lenders. It is a new trend called "peer to peer lending." (Like the old Napster but with cash!) Of course, the best idea is not to borrow money at all (see also: Dave Ramsey). But if you have to, this is one interesting concept.

Is Prosper Going Flash?

Prosper is searching for some freelance web designers in Silicon Valley. No surprises there, other than their interest in Flash and ActionScript development experience. Prosper doesn't appear to use Flash right now (not that it's a bad thing). This may be an indicator towards a website refresh. Their site's design style is getting a bit long in the tooth, so I wouldn't be surprised if it was coming. It may also be that some of those fancy Finovate features will utilize Flash to make them snappy. Either way, it's a great opportunity for rumor mongering.

Position: Full-time freelance (on-site only: Downtown San Francisco)

Specific responsibilities:
  • Work closely with marketing team to create original web pages/web sites/HTML emails and emulate existing web pages, utilizing standards-compliant X/HTML and CSS
  • Ensure development standards, best practices and site technical quality are met and maintained
  • Bachelor’s Degree
  • 3+ years demonstrated development experience with X/HTML, CSS and JavaScript and cross-browser/platform web development techniques
  • Knowledge of development tradeoffs between speed, design, flexibility and the underlying page architectures
  • Solid experience in Adobe Photoshop and image optimization techniques
  • Demonstrable experience managing multiple projects on tight schedules
  • Working knowledge of Offermatica
  • Flash/ActionScript experience is a nice-to-have

Thursday, October 18, 2007

Circle Lending Puts Up Volume

I have never paid much attention to CircleLending because they facilitate loans between parties who have already agreed, in principle, to make the loan. In other words, they handle the bookkeeping. Virgin (of Richard Branson fame) purchased CircleLending a while back, rebranded it (Virgin Money USA), and has started a marketing push (see here, here). I have to admit that I didn't make it past the first few lines of the press releases. However, I developed an attention span long enough to read all the way through one article and there was something interesting buried in the bottom:

CircleLending administers about $200 million in loans and has 30 employees in Waltham. Advani said it will have about 60 employees within a year as it adds new services.

It charges $99 to arrange a personal loan and $199 to arrange a business loan.

CircleLending has over $200M in loans? You have my attention now.

Wednesday, October 17, 2007

How's That Income Doing

Back in August I observed that the borrower's income seems to weigh heavily on distress rates (late or worse). With two more months of statistics under Prosper's belt, I felt it was time to revisit this topic and see how the loans were fairing.

Income > $50k
Income >$25k
Unstated or <$25k 1.65%

This is very similar to the trends reported in August. I also picked up those with incomes >$50k, and those were surprising numbers. More income doesn't appear to be helping the C and E borrowers.

There's also something funny with the ROI numbers. Prosper now has enough loan information to extrapolate estimated returns for these searches and the low or no income loans are fairing better than the income producers.

Prosper's ROI
Income > $50k
Income >$25k
Unstated or <$25k 10.29%

Well heck, that's just gone and messed up my plans. This is going to take more analysis. At least I was right that income reduces default rates.

Tuesday, October 16, 2007

More Slowing State Economies

I'm starting to feel like Chicken Little. It looks like reports of slowing sales tax receipts from California and Georgia were the tip of the ice berg. The San Diego Union Tribune is reporting that half of the states (unlisted) are showing slowing collections.

About half of all U.S. states are collecting less from their sales taxes than expected, which could signal a recession lies ahead as the home market fades.

The receding housing boom could then reveal the underlying economic weakness it had camouflaged, according to Philippa Dunne, a co-editor with the New York-based Liscio Report, published by an economic research firm.

The big gotcha from this report is that it appears to use August sales tax numbers. States like Georgia and California didn't jump off the cliff until September. Some states cited in the article include Florida and Michigan.

While this looks like a nation-wide trend, this could just be a financial hiccup in response to all the mortgage craziness. So, what's a person to do? I'll keep monitoring economic indicators to see how things are going. Until things swing back, I'm tightening my lending criteria to avoid the marginal borrower and those who may be tempted to drop the loan. My tightened criteria:
This is going to lead to some lean bidding, but I'd prefer to have fewer loans than get caught holding loans I didn't really want.

Monday, October 15, 2007

Prosper Got Adding Bank Accounts Right

I have to admit that it's a whole lot easier to write about things that Prosper isn't doing right. I did have one of those, "Wow, it just worked" moments last week, though. I wanted to add a new bank account for transferring around funds since we'd recently changed our primary bank (yes, you can infer that I'm still transferring money into Prosper). They were quick and I found the quick most surprising.

If you're not familiar with the procedure, you print a page from the Prosper, attach a voided check, and fax it in to them. The interesting thing is that their fax page has a 2-d barcode on it that appears to contain enough information to automate things. As to quick, I had an e-mail confirming the fax receipt before I could get back to the computer from sending the fax. Within the hour, my new checking account was active for transfers.

Color me impressed.

Georgia Sales Slowing Down Too - Tighten Those Lending Standards

First California sales tax receipts were dropping, and now Georgia has reported a similar problem. From the Atlanta Journal-Constitution (hat tip to Calculated Risk):

Collections in Georgia were down $2.3 million in September, compared to the same period in 2006. Sales taxes dropped 10.6 percent. For the first quarter of the fiscal year, which began July 1, overall collections are up 5 percent, or $199 million. The state needs a growth rate of 5 percent or more for the rest of fiscal 2008 to meet its $20.2 billion budget.

This is another indication that state economics are slowing down. Tighten up those lending standards and look for the best folks in all the credit grades that you lend in.

Friday, October 12, 2007

A Linkfest By Any Other Name Is Just As Lazy

And into the links!

zcommodore, 36% Interest Rates?

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.

From the Chicago Tribune, Financing tool follows social networking:

An emerging Internet-based financing tool that brings borrowers and lenders together the way eBay links buyers and sellers is gaining momentum, turning more Main Street investors into Wall Street wannabes.

Called peer-to-peer or social lending, the concept has gotten a big boost in recent months, thanks in part to a credit crunch that has tightened traditional lending channels at the same time that online social networks like Facebook have spread the word about how such financing works.

Hindu Times, Net gain: meet money lenders online:

Zap those banks. Zap your friends and family. There are enough strangers out there on the Internet now willing to lend you money at affordable interest rates and without all those extra hidden costs that big banks build into their loans.

Social lending might be well known to most of us — we keep tapping friends and relatives for money all the time — but in the cyber world it is still a new concept. It is gaining currency as people eager to make and save money race to websites that bring lenders and borrowers together.

There's at least one happy Prosper borrower out there, and some reasonably happy lenders as well (the loan was paid back):

I have been aggressively paying my Prosper loan since July of this year. What originally was at $3,500 in June of 2006 was whittled down to $769. With the balance transfer offer that I received, I have initiated the payment that will pay off my Prosper Loan in full. I know for sure that two of my Prosper lenders stop by from time to time. I’m not sure if anyone else does. In any case, if you are reading and were one of the lenders on my loan…thank you.

California Sliding Into Recession?

I had previous considered California as high-risker risk lending terrain due to housing conditions there. Calculated Risk has highlighted another reason to be careful:

Based on sales tax revenue, it now appears that the California economy is in recession.

The graph shows the September retail sales tax collections (and fiscal year through Sept) since 1998. The large decline in 2001 was related to the '01 recession, the tech bust, and 9/11.
September sales tax revenue was off 7% compared with last year.

I've seen a couple things that imply an inverse relationship between economic activity and default rates (another post for another time). In other words, when economics slow down, default rates go up. Most of the papers were correlating to the US GDP, but I'm expecting some affect will be seen on the state level. Even if California isn't slipping into a recession, this is a definite economic slowdown for the state.

I could be wrong, but it's taking a big risk to bet on California default rates not increasing. Therefore, I'm curtailing my lending in California. When I do lend, I'll be adding 1% - 2% to my minimum bid to cover the increased default risk. It may take longer to get loans funded, but I like the alternatives less.

Update: I saw on Yahoo financial news that, nation wide, retail sales were up in September. For California to see a 7% decline in sales tax revenue while the nation was up in overall retails sales is not a good indicator for California. From the article:

Retail sales posted a stronger-than-expected gain and prices at the wholesale level jumped up significantly in September.

The Commerce Department reported Friday that retail sales increased 0.6 percent last month, compared to August, as a big increase in auto sales helped offset weak demand for clothing. The increase was double the gain that economists had been expecting and was also in contrast to reports Thursday of sluggish demand from the nation's leading retail chains

Outside of autos and gasoline, areas of strength included sales at grocery stores and electronics and appliance stores. Areas of weakness last month included department stores, down 0.5 percent; clothing stores, down 0.4 percent and furniture stores, down 0.6 percent. Furniture store sales have been hurt by the steep slump in home sales.

Update 2: California is also the loan leader by a significant margin, so this has implications for the Prosper marketplace in general.

Thursday, October 11, 2007

Notes From Prosper's Hawaii Meet & Greet

Prosper's CEO Chris Larsen participated in a meet & greet organized by Prosper member islandmele this last weekend. At the Honolulu, Hawaii gathering, Chris (noted as CL) did comment on Prosper and answer a few questions from folks. islandmele was kind enough to post her notes from the weekend. Here's the highlight reel as interpreted by me. Go read the original to get it straight from the islander's keyboard:

  1. CL doesn't see profitability for several years. It'll take a 4x - 5x in volume to get there. If (and it's a big if) my estimates of Prosper's cash flow is accurate, that puts their burn rate in the $400 - $500k / month region. This is fair for a startup with a the size Prosper appears to have. The missing detail is how Prosper's expenses scale with loan volume, which I'm sure CL will never reveal.
  2. Some of the features suggested at the Finovate conference should be rolling out shortly. My interpretation is that the new interface showing Prosper data derived default rates should be active in a few weeks. The Prosper data will be refreshed every 30 days (ish). The portfolio's will be some time afterwards.
  3. They're trying hard to get the secondary market in place, but those darn government regulators keep getting in the way (don't they realize that Prosper works at the speed of the internet?!)
  4. Prosper will be showing statistics related to whether friends and endorsements (with and without bids) are statistically significant.
  5. Prosper is looking at raising the max interest rate to 36%. I suggest reading zcommodre's rebuttal if you're a big fan of this idea.
  6. They brought in Doug Fuller to handle ID theft and fraud (on top of improving collections), and can't wait to lock up the bad guys.

Wednesday, October 10, 2007

PII Humor

While there continues to be resentment about the shifting and sometimes ambiguous policies regarding personally identifiable information (PII), it's also become a humor generator for those who enjoy sarcasm and satire. Unfortunately for some, I'm one of those people. This listing and associated forum thread are worth a look, and the borrower Q&A is equally entertaining.

Buying Inventory -- A with no DQs, no PRs, no group, no PII

My name is [Personally Identifiable Information deleted], I live in [Place Identifying Information deleted], and I would like to be your borrower.

Purpose of Loan: I am in the business of purchasing [Product Identifying Information deleted] and reselling them to retail and wholesale customers. Feel free to check out my website at [Possibly Informative Information deleted] for more information. I am seeking a loan to purchase additional inventory.

My Financial Situation: As you can see, my ScoreX credit score is A, and I have no delinquencies or public records. Prosper says that ScoreX is “the only credit grade relevant to lenders placing bids on Prosper,” so you probably stopped reading this listing three paragraphs ago.

I am also a lender on Prosper, with [Portfolio Identifying Information deleted] invested. According to, my ROI is [Pathetically Insignificant Interest deleted], but says that I am making [Practically Impossible Income deleted]. Since I’m a glass-is-half-[Partial Imbibing Information deleted] sort of person, I tend to believe

Budget: My monthly income is [Particularly Important Information deleted], and my monthly expenses are [Particularly Important Information deleted]. As you can see, the difference between income and expenses has enabled me to pay my bills on time and maintain a low utilization ratio.

Since I am not a member of a group, I have taken the liberty of submitting vetting documents to myself. Since Prosper did not disclose any Personally Identifiable Information to me, I can only speculate as to whether the documents that I submitted to myself are in fact my own. If they are, they present a strong case for funding this listing and bidding the interest rate down to [Percentage I’d-Like Information deleted].

During a lengthy conversation with myself, I acknowledged that [Potentially Incriminating Information deleted] but explained that [Perfectly Innocent Information deleted]. Further, it has been more than two years since the last time [Public Intoxication Information deleted], and more than five years since [Paranormal Implantation Information deleted].

Oh, yes, one last thing . . . I am serious about borrowing this money, and I am serious about paying it back. Please be forewarned that I am likely to repay it early (over six to nine months), though, if the final interest rate is low enough, I might keep it longer.

Thanks for bidding.

Prosper Poofs Potentially Powerful Features From Forum

I have a love / hate view of the forums. For every insightful Q & A with Doug Fuller there's an equally frustrating flame war over personally identifiable information.

In the latest forum power struggle between the members and Prosper, Prosper has poofed their description of some upcoming features.

Because the post about our announcements at Finovate included some features that won't be released until possibly far in the future, I was asked to remove them. Sorry about that. More to come as the features are released.

Best regards,

Several forum members quickly pulled the posts and images from Google cache. Between the cached pages, several blogs, and new posts on forums, all the information is in tact. I'm not exactly sure what Prosper hoped to accomplish other than testing the Streisand effect, but none of the possibilities I've come to leave a favorable impression.

And if you're looking to recreate the information, here's where to start:

Tuesday, October 9, 2007

Collections Follow-On Q&A With Doug Fuller

I'm almost starting to like the guy. Doug Fuller has responded via the forums to another 20 questions (ok, a few were softballs) asked in response to his original Q&A session. It's a good read and I've picked up a few of the best questions. I encourage you to follow the whole forum thread for reactions as some are equally insightful.

1. Would you be willing to list what you have seen wrong with Prosper collections at this point and if said weak areas will be corrected?

The two biggest weaknesses in our collections process have been “Agency Management” and the lack of legal (court) collection actions. The problem with Agency Management stems from the realities of a start-up. People are wearing lots of different hats to get the job done. Prior to my arrival, the person who ran the agency relationship is really a “credit and underwriting” expert with little collections experience.

As to the lack of a legal option, there are a number of novel challenges in implementing a legal option for Prosper, but it is clear that we need one and so I’m working through those issues as quickly as I can. I’m going to talk about the legal option in a later section of this Q & A.

2. When you talk about transparency, how transparent are you willing to go? When will the collections process be more transparent? How will we know when the collections process is more transparent? (This is not a trick question, or a joke. While your stated goal is to make collections more transparent, what I've seen in my nine months as a lender has been a move away from transparency. Will you or Prosper post changes in collection procedures, and what they mean in terms of lenders' bottom lines?)

As the first person-to-person credit marketplace, Prosper has to set up a system that is collection-aggressive, but is in compliance with the Fair Debt Collection Practices Act, which protects delinquent borrowers from things like publishing “deadbeat lists.” Collection transparency is more appropriate using aggregate information as opposed to loan level info.

When I talk about transparency, I mean that I am going to be very upfront with the lender community about the steps I am taking to improve collections, the reasons for taking those steps, when the change will be implemented, and when I expect to see results. I also intend to create additional reporting to help people understand what’s happening with collections – the information currently on the site is all “snapshot” data. I also want to create monthly summaries so it is possible to look back and see how many accounts were in collections at the end of August vs. the end of September, etc.

Here are the things I am working on right now:

A. Increased agency oversight – I have already visited both locations of PennCro and have implemented a bi-monthly strategy call. I’m expecting an increase in collections simply from the increased focus on our current agency.
B. I plan to implement a “pilot” legal program during the month of October (meaning first law suits should be filed in November).
C. I am looking to augment current phone channel collection efforts with some off-line collection letters.
D. I have been interviewing agencies as potential replacements to our current agencies, focusing on smaller agencies that I believe will give us more attention.
E. I am evaluating the pros and cons of bringing the collections activities in house. I expect to reach a decision on this by the end of November.

By the end of the year, I would hope to see a significant increase in the number of delinquent dollars collected.

6. Will Prosper change its TOS to include suits against delinquent borrowers as an option? Right now the TOS says junk debt sales at 4+ months.

In order to change our current agreements to allow for lawsuits, a number of challenges need to be met. I want Prosper account holders to know that we will sue them and if we sue, we will win. There are some challenges in this. These include:

A. It is not cheap to sue people. While they vary by state, filing and service fees average something over $200 per law suit across the country.
B. Additionally, given our size and novelty of our asset, no law firm is going to take our business on a contingency basis. At least initially you can assume that legal fees on going to be in the neighborhood of $1500 for a NONCONTESTED suit. If the defendant files an answer, that number goes up. A really nasty case could be $15K to $20K in legal fees.
C. There are so many new aspects to how Prosper works that there will have to be a “custom development” to create the pleadings for a Prosper lawsuit.

What I’m doing at this point is putting together a pilot legal program. I have identified a group of loans which have already defaulted or are on the verge of default. These loans will be included in our next debt sale. In order to gain the legal standing needed, I’m proposing that Prosper buy these loans for the same amount that the debt buyer would and use these as an initial test.

My thought is to do this with a group of loans from borrowers who are all in California, so we only have to deal with one state’s court system. I have a meeting scheduled for Wednesday with the managing partner of what I consider to be the best collections law firm in California. My hope is that we can formally place these loans with the law firm this month and have the first suits filed in November.

10. Do you have an especially dark corner in your heart for delinquent borrowers who are also active Prosper lenders?

Yes. I know we are now putting borrower-lender accounts on hold on a monthly basis and I hear we owe the community a thank you for bringing this problem forward.

Prosper's September Lending Market Summary

Prosper released their September market summary. Check it out for all the details.

Their noteworthy section included some interesting statistics. Apparently self-reported computer programmers are the least likely to default, followed by government workers, the nebulous analyst, and then mechanical and electrical engineers. This is a fun statistic (since I'm either a computer programmer or electrical engineer, depending on the day of the week), but don't adjust your bidding based on it. This is self-reported information, and the cynic in me expects the number of computer programmers applying for loans to magically jump.

Also noted was that the lowest default rates by state were Minnesota, Ohio, New Jersey, Colorado, and New York. This information conflicts with Lending Stat's analysis of state diliquency rates. I suspect that this is because Lending Stat looks at loans that are late as opposed to loans that have gone into default and have been sold off.

And, from Chris Larsen's market commentary:

When the Fed cuts interest rates people often expect mortgage rates to drop. However, this is rarely the case given that mortgage markets typically anticipate rather than react to moves by the Fed. On the flip side of the coin, the variable credit card and savings rate markets react sometime after the Fed moves. In fact, some variable credit cards have a 90-day window to make adjustments reflecting the rate cut. So the question is: did the Prosper marketplace anticipate or react to the Fed rate cut?

Many might assume that the Prosper marketplace would act less like the mortgage markets and more like the credit card and savings rate markets given that the latter compete with Prosper. Nevertheless, the month over month drop in average borrower rates indicates that the Prosper marketplace may have anticipated the Fed cut.

For example, in September the average borrower rates for all prime and near prime loans funded in the Prosper marketplace were 12.29% and 18.22%, respectively; down 0.37% and 0.28%, respectively, from August.

What is interesting about these percentage drops is how close they are to what was widely anticipated to be a quarter-point instead of a half-point reduction by the Fed. However, what remains to be seen is whether the market will continue to push rates down further in line with the Feds surprise move.

I want to highlight the decrease in Prosper loan interest rates. Chris Larsen has attributed this to the decrease in the Fed interest rates. Because these two items were correlated, does not imply causation. I suspect a much more pedestrian effect was causing the decrease, and it's less press release friendly. As zcommodre has noted, there was a significant decrease in funded loans in September compared to August and Eric's CC shows a similar drop in listings (Marketing 101 says don't put that in a press release). However, there did not appear to be a commiserate drop in lenders. The same dollars chasing fewer borrowers looks like an Economics 101 supply and demand problem. I'd expect the average interest rate to drop under those conditions regardless of what the Fed does.

Monday, October 8, 2007

More Details On Upcoming Changes

More details emerged over the weekend on the new Prosper features hinted at in a Yahoo financial news article and confirmed by Prosper. In addition to his Q & A with Prosperous Land, Prosper Andrew posted details to the forums and answered a few questions. The thread adds color to previously known information on the two key features: default rate estimation and a portfolio tool. In addition, some other nuggets slipped out. Prosper will be changing the loan servicing fees on AA and A grade loans:

1) I see that in Slice 2 of your Balanced Plan example, the servicing fee is listed as 1% -- Is Prosper planning on raising the servicing fee (since it is currently 0.5% on AA and A loans)?

Yes, we are planning to change the servicing fee on AA to 0% (yes, 0%), and on A to 1%.

And, in a different thread, he reiterated that they're working on a secondary market to allow lenders to resell loans. He emphasized that the limitations on the roll-out were regulatory.

Hi guys,

We are working on building a secondary market. And I know we have been saying that for several months, but the primary hurdle is regulatory, which, unfortunately, is not something that we have direct control over.

Anyway, it is still one of our top priorities. Just thought you should know that we haven't forgotten about it.

And, of course, the screenshots.

And then for the portfolio tool:

Update: The images were briefly lost due to operator error. It's been fixed now.

Friday, October 5, 2007

Prosper Confirms Upcoming Lender Features

Earlier today I pointed to a Yahoo financial article that referred to several features that weren't currently available on the Prosper site. It turns out that these features were previewed by Prosper at the Finovate 2007 conference in NYC this last week and the Yahoo article was based on Prosper's presentation at the conference.

Prosper was kind enough to confirm the upcoming features were presented at the Finovate conference. The first major feature will show lenders the estimated default rate based on Prosper's default information (not Experian's) when lenders bid on on listings. The second major feature will allow lenders to assemble a loan portfolio to help control risk in their lending.

Prosper Andrew was kind enough to answer a few questions about the upcoming features.

Mike @ Prosperous Land: Will the estimated default rates be shown when lenders click on the "Bid" button in something similar to the current warning on E/HR loans?

Prosper Andrew: Yes, it will be shown on the bidding page, and will actually be shown on every single listing (unlike the current E/HR warning). So good listings (AA, 0 DQs, etc.) will have a very low estimated default rate, and very poor listings (HR, multiple DQs, etc.) will have a very high estimated default rate. As always, it will be up to the lender to decide whether the level of risk is appropriate to his or her lending strategy.

Mike @ Prosperous Land: Have you settled on which credit criteria will be used to bin the loans for analysis? I've noticed that the criteria can only be sliced so thin before there's not enough loans to do a good extrapolation.

Prosper Andrew: You’re on the right track with this, and yes, we’ve settled on the criteria. Basically, using public data, for each credit grade, we’ve determined which credit factors contribute most to risk, and used those to slice up the borrower population. There are about 50 slices across all 7 credit grades. Some grades have fewer slices (AA has 3, for example), and some have more.

Mike @ Prosperous Land: The Yahoo article suggested that you could base the criteria on whether they're a recent college graduate. Loan categories like "recent college grad" sound like self-reported information. Isn't that a bid dodgy?

Prosper Andrew: Yeah, this is probably just a misunderstanding on the Yahoo author’s part. We are only using quantitative data like credit grade, delinquencies, inquiries, and stuff like that.

Mike @ Prosperous Land: Will Prosper do automatic bidding based on the portfolio or will it provide guidance as to how a listing fits into the portfolio?

Prosper Andrew: It will be similar to the way standing orders work, where listings that meet the portfolio plan’s criteria get bids.

Mike @ Prosperous Land: The portfolio planner sounds similar to features provided by Lending Club's portfolio system. Are there features that differentiate Prosper's implementation?

Prosper Andrew: It’s fairly different, actually. LendingClub’s tool just lets you take your money, choose a risk level, and actually places your money at the very moment in what’s available now. Our tool will offer 4 pre-defined plans that offer different levels of risk and return, and will basically set you up with a standing order that places bids on listings that meet the plan over time. So your money could be bid immediately (like LC’s tool), or it could be placed over time as more listings that match the plan are created.

Mike @ Prosperous Land: When can we expect these features to roll out?

Prosper Andrew: it will be over the next few months. And we’ll roll these features out in phases, so you might see some functionality now, then more later.

Thanks to Prosper Andrew for taking time to clarify the new features.

Prosper Sets Date For Prosper Days 2008

Looks like the 2008 incarnation for Prosper Days will be Feb 25 - 26 in San Francisco. They're gearing it mostly toward lenders, but of course, all our welcome. Since I have no scruples, I've shameless listed the lecture schedule from their site (it's posted below). Doug Fuller, new VP of Operations, will be covering Collections and Fraud Prevention. That should be an interesting follow-up to his online Q & A.

Monday 2/25:

7:00 AM Registration & Breakfast: Collect your badge, meet other Prosper users, and plan your day Prosper 101: "Introduction to Prosper" session for newcomers and guests (8:00 AM)
9:00 AM Keynote address: Prosper CEO Chris Larsen
10:45 AM Lending 101: Comprehensive lender education Blogger Panel: Moderated by Jim Breyer, Director, Accel Partners
12:00 PM Best Practices Lunch: Bring your best Prosper idea and share it during lunch
1:30 PM Collections with Doug Fuller, Prosper's Vice President of Operations Borrower Panel
3:15 PM Forum Sandbox with Andrew Martinez-Fonts, Prosper’s Product Manager Managing Large Portfolios with Kirk Inglis, Prosper's Chief Financial Officer
5:30 PM Networking event and VIP Lender Cocktail Hour with Prosper Executives
6:30 PM Dinner: Group dinner at a fun San Francisco locale

Tuesday 2/26:
8:00 AM Breakfast: Network with other Prosper members
9:00 AM Keynote Speaker TBD
10:30 AM High-ROI Lender Panel: Learn portfolio secrets from solid Prosper earners Fraud Prevention with Doug Fuller, Prosper's Vice President of Operations
12:00 PM Town Hall Meeting: Lunch and Town Hall Meeting with Prosper's cofounders, Chris Larsen and John Witchel, moderated by Bob Kagle, Director, Benchmark Capital

Upcoming Prosper Features Hinted In Yahoo Article

A recent Yahoo article on online financial services may have accidentally provided hints on some upcoming Prosper features:

Prosper mined its own data to give lenders a better view of their risk. A new feature tells borrowers the historical risk of default on a specific kind of loan, such as loans to recent college grads. In addition, a new tool allows lenders to construct a portfolio of Prosper loans with a predictable rate of return and a specific risk level. Users can create and publish their own portfolio plans, which other lenders can grab and personalize. "We're expanding the power of community and collective wisdom to get great lending decisions," says Chris Larsen, Prosper CEO and cofounder.

I'm sure it's entirely coincidental, but there's some scheduled Prosper downtime coming up this weekend for "upgrades". Wow, what are the odds of a press article talking about new features around the same time an upgrade is happening?

Prosper's September Revenue

Following up on last month's post documenting how Prosper makes money, I plan to keep looking at their monthly income monthly to see how they're progressing as a business. Into this month's totals. First, origination income:

Credit Grade
Loans Originated
Borrower Fee

And now loan servicing income:
Credit Grade
Outstanding Loans
Lender Fee

The sum total is $107.6k for September, a 8.8% decrease over $118.0k in August. Comparing September to August shows a significant decline in new loans originated, with all credit grades showing a significant dollar declines except for B which was nominally flat (+.1%). September did only have 30 days instead of 31, which should account for about 3% of the loan origination decline, but overall this was not an impressive month.

Origination Income
Servicing Income

Thursday, October 4, 2007

Prosper Confirms New VP's Collections Experience

Prosper has published a Q&A with the new VP of Operations Doug Fuller (Prosper Forum, Rate Ladder, Prosper Lending Review). The introduction to the Q&A confirms him as the Doug Fuller with a hefty collections and efficiency oriented background. This is the same one whose resume surfaced in the forums a little bit ago.

From the Q & A, he's confirmed that he's quite aggressive on debtors who don't pay up:

Q: Well why don’t we just sue everybody?
A: The phrase “blood from a turnip” comes to mind. One of the ways that you can go broke in a big way is by suing people that will never be able to pay you at all. Simple math, it costs a lot of money to sue people.

Q: Okay, so you need to decide who to sue, then what?
A: Put quite simply, my philosophy is this – if you won’t pay, but can (or will in the future) be able to pay, I’m going to sue you. If I sue you I’m going to win.

Q: That sounds kind of arrogant, can you back it up?
A: Courts in seven states have recognized me as an expert at consumer debt litigation. At Credigy, if a case got really nasty, I would go testify live. I refuse to lose.

Q: Really? What’s your win/lose record?
A: In my last 18 months at Credigy, I testified live at 42 trials. My record was 41-1. By the way, I fired the law firm where we lost.

Monday, October 1, 2007

Loans From Groups Declining

zcommodore has been monitoring the number of group loans funded vs non-group loans funded in response to Prosper's change in group rewards. A few weeks ago he noted that the ratio was swinging to non-groups:

Back in May/June when I first started checking, the number of "no group" listings at any point in time was around 60%*. After the removal of match rewards in early June along with a few changes relating to the removal of directions on Prosper encouraging borrowers to join groups, I noticed the number was hovering right around 70-72%. This morning, a little over a week after Prosper's change to remove group leader fees, I see the number of no-group listings has risen to 77%.

His latest calculation has things swinging further:

The number of listings, percentagewise, that are now not in a group has increased significantly since then. When the change was first implemented, roughly 70% of all listings were "no group". This morning, I checked again and it had risen to 82%.

Anecdotally, I've been finding the listings to be of lower quality. Correlation or causation - you be the judge.