Over the past week, Dec 22 - 29 I've noticed that there has been a large increase in the number of loan listings where the applicant has a public record on file. Anecdotally prior to this week I rarely saw listings with public records but now it seems like the majority of them have one. Has something changed with the way information is being reported or what gives?
I normally screen out public records anyway, but I just looked at total loans available a few minutes ago & it stood at 1093. If you screen out only those with public records you're left with 591. I don't have anything to compare this with the average historically, but yes, 46% of all loans available having a public record seems dramatically elevated right now. Though I wouldn't want to suggest any reasons or draw any conclusions just yet.
Here's a table by vintage quarter with the percentage of loans with a public record. Now, these are issued loans, so perhaps a TON of the ones on the platform won't issue, but 46% is so far out of line with the past, it makes me think something's wrong. That said, a public record isn't the end of the world.
. table iq, c(mean publicrecord)
--------------------------
iq | mean(public~d)
----------+---------------
2007-2 | 0
2007-3 | .0740741
2007-4 | .0653951
2008-1 | .0856553
2008-2 | .0979167
2008-3 | .0734266
2008-4 | .0698925
2009-1 | .0552359
2009-2 | .0541825
2009-3 | .0361538
2009-4 | .0465632
2010-1 | .0584795
2010-2 | .0680249
2010-3 | .0613372
2010-4 | .0560286
2011-1 | .0562956
2011-2 | .0614628
2011-3 | .0634016
2011-4 | .0386435
2012-1 | .0314555
2012-2 | .0304205
2012-3 | .0296941
2012-4 | .0096432
--------------------------
Can someone re-check that my 46% is ballpark correct ................because that doesn't even sound right to me. The more I think about it, the more ridiculously high it sounds.
My script says 48% of 3-year loans in the BrowseNotes file have a public record. Something's got to be wrong on their end.
Yeah, that can't be accurate.
One point of view might be that there are 46% of the currently available notes with public records because the notes with public records stay available longer, and the notes without public records get funded faster.
So to your point Dan the only way to really compare this is to look at how the browsenotes.csv file changes over time.
Here's data I have from some older downloads:
Date/Time Total Notes Notes with PR % with PR
12/25/12 09:14 AM 764 109 14.3% browseNotes(4)
11/25/12 08:52 AM 721 19 2.6% browseNotes(3)
11/18/12 10:06 AM 863 18 2.0% browseNotes(2)
09/25/12 12:00 PM 1352 19 1.4% browseNotes(1) (okay that's just spooky)
Wow substantial change. the other interesting thing was the notes from Nov & Sept didn't have any public records greater than 2. The stuff from December has notes with public records as high as 5.
Still doesn't really answer the question of if people are changing what notes they are buying up. Did the change happen because of who LC is approving or because of which loans are leaving the platform faster? Is it because of the discussion on this forum about public records being bad for ROI?
I couldn't attach the files (file size too big). Is there a good way to post them somewhere for folks?
edit: okay got some drop box links:
these are all my browseNotes files that I have.
(1)
https://www.dropbox.com/s/e4tva3b7fqxy15p/browseNotes%20%281%29.csv 9/25/12 11:00 AM
(2)
https://www.dropbox.com/s/c5qd83q3ssplaum/browseNotes%20%282%29.csv 11/18/12 10:06 AM
(3)
https://www.dropbox.com/s/r5tawrqyvrwfp0v/browseNotes%20%283%29.csv 11/24/12 8:52 AM
(4)
https://www.dropbox.com/s/26qllni2ojfnwza/browseNotes%20%284%29.csv 12/25/12 9:14 AM
(5)
https://www.dropbox.com/s/1tlfs5yrln0i754/browseNotes%20%285%29.csv 12/28/12 7:14 AM
(6)
https://www.dropbox.com/s/652v307v89lemsl/browseNotes%20%286%29.csv 12/29/12 8:53 AM
I believe it. I typically invest in B-4 or worse rated loans. Up until this past week I didn't even bother to have public records in my filter since there were so few I would just skip the few that had them as I went though. There are so many now in the C and D grades esp. I started to filter them out to avoid mistakes.
Here is a break down of loans with public records by credit grade from today's browse Notes. I'm getting an over all Public Record percentage of 46.7%
Grade, Pub Records (Total Loans), % w/ Record
A | 26 (241) | 10.8%
B | 233 (398) | 58.5%
C | 173 (307) | 56.4%
D | 58 (100) | 58.0%
E | 31 (69) | 44.9%
F | 21 (39) | 53.8%
G | 2 (11) | 18.2%
I think it has to be a mistake somehow in the data. Things don't change that fast without some explanation. Perhaps the explanation is they bought the National Association of People with Public Records mailing list! Or maybe the database is pulling the wrong records somehow.
Does anyone know what percentage of the general adult population have a public record? It's under 5%, correct?
I would estimate that 4% of credit files would have a bankruptcy on record which is looks like most of the records in the current listings are. Here is how I made my estimate...
There are some bankruptcy statistics on the US Courts website;
http://www.uscourts.gov/Statistics/BankruptcyStatistics.aspxI can't find any number about people with it on their record but if you look at those stats 1.3 million people declared bankruptcy during the year and 72% of those where first time filers so 1300000 * 72% = 936000 first time filers. A bankruptcy will be on your record at most 10 years in some cases less per experian's FAQ
http://www.experian.com/ask-experian/20110413-public-records-that-can-appear-in-your-credit-report.htmlso 936000 * 10 = 9.36 million people with a bankruptcy record at any given time
According to Census data 76.3% of the population is over the age of 18 and there are 311 million people so 237 million adults who could have a credit record
http://quickfacts.census.gov/qfd/states/00000.html9.36 / 237 = 3.9%
So a high estimate might be about 4% of credit files would have a public record from a bankruptcy.
looking at the browse notes there is actually a public record, and a public record from bankruptcy field and most of the records are in fact bankruptcy related. Even is 4% is a total WAG clearly 46+% is way out of line with what you would expect.
Okay my files are on drop box. Please tell me if you see something that I didn't in the files. I also added the ones that I have from the last week. Does anyone else keep these that they are willing to share? It looks like there's an obvious change with my data set, but maybe someone else's data would show something else?
I don't think we need to do any more research. We've established a major change in the data from week to week. It's time to call LC.
I figured they'd be on skeleton crew.
A gentleman named Tim promised to give me a return call once he verifies the status.
My hypothesis was fundamentally correct, if hyperbolic!
LC adjusts the interest rate according to their risk model, which probably incorporates the presence of public records either directly or through the credit score. I'm glad to have confirmed the data is correct.
So has this changed how people are selecting notes?
I haven't modified my filters at all, but I'm mostly just reinvesting at this point so having fewer notes to invest in isn't that bad of a thing.
Nope, as long as the data is correct, I have no reason to believe the underlying relationship between a public record and a loss is any different. Same script.