Print

[112233]

(if this was discussed elsewhere, my apologies)

according to prosper's home page (http://www.prosper.com/),  you can become a lender and "Earn 9.49-12.81% Returns"

the criteria used for this range is as follows according to prosper

Rates of return shown are the average annual returns on Prosper loans originated between 7/1/06 and 10/30/07 to borrowers with AA - C credit grades who requested automatic funding, have 0 - 1 current delinquencies and 0 inquiries in the last 6 months on their credit record, as of 11/30/07.

I retrieved the average annual returns using this criteria on their performance page, but I dont get the same range. Since there is an observation date, I'm assuming I should get the same results for this observation date regardless of when I run the query. What am I doing wrong? The attached image is the query I used.

[LoanChimp]

yes, I retrieved the same image and figures (just today as a matter of fact). Part of the bad result is that Prosper recently changed the Roll Rates on the Performance Page (1/11/08), which appears to be affecting   most of the loan performances (from the loan performance page) in general.

I have found a couple of other "advertisements" that have relied on "old" (pre-revised roll rate figures) that no longer make the cut, according to the data ranges specified.

I've submitted a couple of bug reports lately. Depending on how these are handled, the old ads may start making new topics for my (not-so-shiny-anymore) new blog...

[ira01]

It is not the new roll rates (or anything else new) that is causing this discrepancy.  I posted months ago how these banner ads were not only misleading, but not even truthful using Prosper's own criteria.  I'll dig up a link to the thread.

ETA:  Here is the link.

[LoanChimp]

It is not the new roll rates (or anything else new) that is causing this discrepancy.  I posted months ago how these banner ads were not only misleading, but not even truthful using Prosper's own criteria.  I'll dig up a link to the thread.

The discrepancy is even worse now...

[ira01]

It is not the new roll rates (or anything else new) that is causing this discrepancy.  I posted months ago how these banner ads were not only misleading, but not even truthful using Prosper's own criteria.  I'll dig up a link to the thread.

The discrepancy is even worse now...

Wow.  I only recorded the projected ROI's, not the consituent components, so I can't say for sure what all is causing that.  But the 6.5% net default for AA sure seems quite high. 

[jmathree]

The closest I could get was 9.09% by weighting the return with respect to each credit grade (AA-C).  However, I still had to subtract out the adjustments and servicing fees to get this.

Total AA-C loans originated--  $2,518,056

(587,524/2,518,056)(.1202-.0644)+(540,580/2,518,056)(.148-.0654)+(752,571/2,518,056)(.1714-.0806)+(637,381/2,518,056)(.1975-.0667)= .090998

That's the closest I could come to generating the bottom end of their range, so I don't know how they came up with theirs.

[LoanChimp]

It is not the new roll rates (or anything else new) that is causing this discrepancy.  I posted months ago how these banner ads were not only misleading, but not even truthful using Prosper's own criteria.  I'll dig up a link to the thread.

ETA:  Here is the link.

The link that you provided shows that they were indeed close to their claims at one point. But when you run the data now, with the same origination dates and observation date, it's quite different.

It seems they have not updated their advertisements for quite some time. Considering that they clearly describe the performance page criteria on these ads (and they no longer make sense), they need to stay on top of this stuff...

[LoanChimp]

damn you, 111222334, you usurped a blog entry of mine (and saved me a lot of time...  )

[112233]

damn you, 111222334, you usurped a blog entry of mine (and saved me a lot of time...  )

I think this is still worthy of a blog entry if indeed the advertisement is incorrect. This is not just any old ad buried deep in the website or an email. THIS IS RIGHT IN THE MIDDLE OF THEIR HOMEPAGE next to a texas-sized orange button. Are you telling me that I noticed this before Prosper? It makes me more than a little worried if I did. I hope it's just a matter of me not doing this correctly because I cant believe something like this would not be noticed.

Also, there are some other questions I have like .. why is the observation date so close to the ending origination date?, why was automatic funding selected as a filter?, and why was this origination date range selected (1st 5 months left off, and last 12-15 months left off)?

[112233]

the query has changed. it now says

[1] Rates of return shown are the average annual return on Prosper loans originated between 6/1/06 and 10/26/07 to borrowers with AA credit grades who have 0 delinquencies and 0 to 2 inquiries on their credit record, as of 11/26/07. For more information

was there downtime last night?

[HollowOak]

thw query has changed. it now says

[1] Rates of return shown are the average annual return on Prosper loans originated between 6/1/06 and 10/26/07 to borrowers with AA credit grades who have 0 delinquencies and 0 to 2 inquiries on their credit record, as of 11/26/07. For more information

was there downtime last night?

No, we just have a loyal and dedicated readership at Prosper HQ.

[112233]

ok, I ran the average annual return query with the new criteria that magically appeared this morning. Among other items, they have restriced the credit grades to only AA. How can there be a range for a single credit grade given an observation date? Shouldnt there just be one number?

and BTW, the average annual return for AA doesnt even fall in the advertised range (but it's getting closer)

[LoanChimp]

I still see the other (first mentioned in this thread) criteria.

I think they rotate between these two...

[ira01]

ok, I ran the average annual return query with the new criteria that magically appeared this morning. Among other items, they have restriced the credit grades to only AA. How can there be a range for a single credit grade given an observation date? Shouldnt there just be one number?

and BTW, the average annual return for AA doesnt even fall in the advertised range (but it's getting closer)

Some things never change.  I posted on this "new" claim a month ago here, pointing out the absurdity of including a range for a single credit grade, as well as other problems:

Well, Prosper has now changed the footnote again, leaving the banner ad the same ("Earn 9.49-12.81% returns¹.")  The new footnote, which is now at least close to the bottom of the page (who says Prosper doesn't read this board?) reads:

Rates of return shown are the average annual return on Prosper loans originated between 6/1/06 and 10/26/07 to borrowers with AA credit grades who have 0 delinquencies and 0 to 2 inquiries on their credit record, as of 11/26/07. For more information click here.

Frankly, I don't even know where to start with this bullshit.  The first thing that is apparent, is that it makes zero sense for the footnote to now include a single credit grade (AA) while the banner ad continues to include a range of returns.  What is the range of returns supposed to signify -- whether you are lucky or not?  The return is what it is, and there is no range.

Second, Prosper's description of the cherry picked bucket of loans in the footnote is ambiguous -- what does "0 delinquencies" mean?  0 CDQs, 0DQ7s, or both?  I ran the numbers all three ways, and (to no one's surprise, I'm sure), Prosper's banner ad still misrepresents the return by substantially overstating it.  For 0CDQ's, the return for the described bucket is 9.26% (which is also the best return for any credit grade with those parameters).  For 0DQ7s, the return is 9.16% (which is also the best return for any credit grade with those parameters).  If you interpret the footnote to mean 0 CDQs and 0 DQ7s, the return for AA is 9.21% (although the return for B is 10.53).  Bottom line -- under any of these interpretations of the footnote, the banner ad's statement "Earn 9.49-12.81% returns" is a misrepresntation, if not simply fraudulent.

Third, as if to prove that Prosper is so incompetent they cannot do even the simplest things correctly, recall that the footnote ends: "For more information click here."  That is a link to the performance page -- however, the paramaters preloaded into the performance page when you click the link are completely different than what is specified in the footnote -- it uses DTI 0-40%, 0 CDQ's, 0-3 INQs, and (6/1/06-11/22/07, observed 12/22/07), which is apparently the current default settings for the performance tab.  That search yields a return of 9.09 for AA (and worse for all other credit grades).  Would it really be too much to expect that the footnote to the banner ad actually describes a cherry-picked subset of loans that actually have the same performance touted in the ad (and that the link in the footnote takes you to the performance page running those particular numbers)?  What is the point of Prosper's constant tinkering with the footnote, if they never fix the problem (and in fact, seem to make it worse each time)?

[112233]

thanks Ira. I dont jump on some of this stuff as fast other people so I figured Prosper or someone else would have noticed such a highly visible misstatement by now. Since I couldnt find any threads about the subject I thought I made a mistake 

on a side note, I noticed they have a spiffy flash animation on their homepage now