Lend Academy Network Forum

Lending Club Discussion => Investors - LC => Topic started by: nscap on March 29, 2013, 05:17:17 PM

Title: Verified income
Post by: nscap on March 29, 2013, 05:17:17 PM
Hello all - lately (~2 months or so), I have noticed that the number of loans with verified income has greatly decreased. Given this is a required component of my investment process, this has reduced my total number of investments, and brought some doubt into my mind about scalability. As such, I have a few queries:

1) Has anyone else noticed the # of verified income loans has gone down? Or are they merely being snapped up so fast after being put up that I am missing them?
2) If in fact the # of verified income loans has gone down, what do you guys think is the reason?
3) Does verified income really matter that much? The reason I put so much faith in it is that is literally one of the few metrics I can trust to analyze... do people generally tend to put down their real income? What is stopping someone (who isn't verified) from lying about their monthly income? Is that common?

Many thanks!
Title: Re: Verified income
Post by: Lloigor on March 29, 2013, 05:57:51 PM
I think that just as many loans have verified income as before, but not during the funding process.  Things are funding so fast that you have to buy in before the information is verified.  You can buy as usual and just sell the ones that don't wind up verified for 1-2% markup.  I believe it's been pointed out that statistically, verified income actually do slightly worse.
Title: Re: Verified income
Post by: rawraw on March 29, 2013, 06:30:48 PM
Hello all - lately (~2 months or so), I have noticed that the number of loans with verified income has greatly decreased. Given this is a required component of my investment process, this has reduced my total number of investments, and brought some doubt into my mind about scalability. As such, I have a few queries:

1) Has anyone else noticed the # of verified income loans has gone down? Or are they merely being snapped up so fast after being put up that I am missing them?
2) If in fact the # of verified income loans has gone down, what do you guys think is the reason?
3) Does verified income really matter that much? The reason I put so much faith in it is that is literally one of the few metrics I can trust to analyze... do people generally tend to put down their real income? What is stopping someone (who isn't verified) from lying about their monthly income? Is that common?

Many thanks!
You shouldn't be worried if an individual borrowe liesr.  You should worry if you think lying would be common in the population AND LendingClub won't catch it (There are legal requirements for LC/Webbank to attempt to verify incomes. Not sure how it works in P2P context though).  We aren't dealing with individual exposures, but rather behavior of the entire pool.

Just FYI, verified income loans don't do any better (they do slightly worse) than unverified loans.
Title: Re: Verified income
Post by: nscap on March 29, 2013, 06:44:42 PM
Thank you.

Does LC make efforts to verify all loans? I.e. the ones that are listed as "verified" are just the ones that have already been verified, the other ones haven't been yet (meaning they haven't been left unverified for some sinister reason)? I appreciate the sentiment re: behaviour of the pool vs. an individual issuer, but given that I only have $1k invested, just one borrower lying has an impact on my overall performance (which is why I chose to go with verified only).

Can you point me in the right direction for where to find out more about verified vs. unverified income loans in terms of defaults, performance etc?
Title: Re: Verified income
Post by: Lloigor on March 29, 2013, 07:31:06 PM
They use internal criteria to decide if they should get verification.  Some classes of borrowers (I'd guess the more creditworthy ones) have less tolerance for hassle and will just abandon LC if it gets too difficult so that population gets fewer verifications, while others are required to verify for approval.
Title: Re: Verified income
Post by: viking on March 29, 2013, 07:43:39 PM

Just FYI, verified income loans don't do any better (they do slightly worse) than unverified loans.
What is that based on?
I did a quick test using Interest Radar:
Loan Grades D-G
All loans (verified and not)=> Av IRR= 9.4%
Inc Verified => Av IRR= 10.5%
Title: Re: Verified income
Post by: brycemason on March 29, 2013, 08:25:47 PM
I've also independently checked out whether verified incomes do better or worse. There is no impact according to my research, as I expected. A suprising proportion of loans get verified.
Title: Re: Verified income
Post by: viking on March 29, 2013, 09:59:59 PM
I've also independently checked out whether verified incomes do better or worse. There is no impact according to my research, as I expected. A suprising proportion of loans get verified.
Well, at least for all D-G loans, verified income produced a higher return Av IRR= 10.5% vs 9.4% (not sure about significance level though....)
Title: Re: Verified income
Post by: brycemason on March 29, 2013, 11:19:43 PM
Univariate analyses bore me.
Title: Re: Verified income
Post by: rawraw on March 30, 2013, 08:31:34 AM
Univariate analyses bore me.
This made me LOL
Title: Re: Verified income
Post by: Fred on March 30, 2013, 09:24:08 AM
If you remember our financial crisis experience from 2007, one of the factors for the problem was that banks were lending money to borrowers without verifying their income.

Perhaps people at that time tend to lie more on their loan application  than today's borrowers.

Or, perhaps size matters: income verification does not affect small loans (< $35 K) as significantly as it does mortgages?

Title: Re: Verified income
Post by: rawraw on March 30, 2013, 06:09:11 PM
If you remember our financial crisis experience from 2007, one of the factors for the problem was that banks were lending money to borrowers without verifying their income.

Perhaps people at that time tend to lie more on their loan application  than today's borrowers.

Or, perhaps size matters: income verification does not affect small loans (< $35 K) as significantly as it does mortgages?
Fred, I like the way you are thinking.  I wish more people on this board thought critically.  But I think you are going down the wrong path.  The MBS mortgage didn't blow up solely because of verification of income (The Credit Card industry is still chugging along, without verifying most people's income).  The MBS market blew up because loans were being made with the source of repayment being collateral and not cash flow.  If you aren't involved in finance/banking, basically what this means is that everyone forecasted the collateral values would at least remain constant (most forecasted they'd increase forever).  They weren't worried about incomes because if the borrower didn't pay, the bank could foreclose on the asset and sell it.  Since they were increasing Y/Y, not much risk to the bank (we do have selling costs and such, but those are helped by the gain in market value).  But what happened?  The loans were made to bubble-level prices and the floor came out from under them.  Borrowers don't want the houses anymore (they are underwater and mortgages have lots of various laws (state and federal) which limits the recourse.  These precipitous drops in collateral values were accompanied with spikes in unemployment as well.  Basically, what I'm saying is income wasn't the sole purpose of the problem.  It did play a part, but there were a lot of moving parts.
Title: Re: Verified income
Post by: Fred on March 31, 2013, 12:20:01 AM
Fred, I like the way you are thinking.  I wish more people on this board thought critically.  But I think you are going down the wrong path.  The MBS mortgage didn't blow up solely because of verification of income (The Credit Card industry is still chugging along, without verifying most people's income).  The MBS market blew up because loans were being made with the source of repayment being collateral and not cash flow.  If you aren't involved in finance/banking, basically what this means is that everyone forecasted the collateral values would at least remain constant (most forecasted they'd increase forever).  They weren't worried about incomes because if the borrower didn't pay, the bank could foreclose on the asset and sell it.  Since they were increasing Y/Y, not much risk to the bank (we do have selling costs and such, but those are helped by the gain in market value).  But what happened?  The loans were made to bubble-level prices and the floor came out from under them.  Borrowers don't want the houses anymore (they are underwater and mortgages have lots of various laws (state and federal) which limits the recourse.  These precipitous drops in collateral values were accompanied with spikes in unemployment as well.  Basically, what I'm saying is income wasn't the sole purpose of the problem.  It did play a part, but there were a lot of moving parts.

Well put, rawraw.  I agree that there were a lot of moving parts.

Full disclosure: I was a "quant" in 2007-2008 working on RMBS (Residential MBS) pricing engine for a Wall Street bank.  RMBS, CDO, CDS are hot potatos now.  However, I do believe some lessons learned on credit risks are still applicable to P2P consumer loans.

There are almost 100 attributes in downloadable LC loan data, somebody can analyze which are (statistically) significant factors for default. 

I do give statistical preference in my model to loans whose "is_inc_v" is TRUE.  Note, this is just a statistical preference, not a binary yes-no decision.



 
Title: Re: Verified income
Post by: rawraw on March 31, 2013, 08:14:10 PM

Well put, rawraw.  I agree that there were a lot of moving parts.

Full disclosure: I was a "quant" in 2007-2008 working on RMBS (Residential MBS) pricing engine for a Wall Street bank.  RMBS, CDO, CDS are hot potatos now.  However, I do believe some lessons learned on credit risks are still applicable to P2P consumer loans.

There are almost 100 attributes in downloadable LC loan data, somebody can analyze which are (statistically) significant factors for default. 

I do give statistical preference in my model to loans whose "is_inc_v" is TRUE.  Note, this is just a statistical preference, not a binary yes-no decision.

I do think there are lessons as well.  Recently Lending Club took the interest rate determining formula from publicly displayed to a proprietary formula.  If LC gets greedy (as humans and institutions inevitability do), I could see similar incentives that helped the MBS spiral out of control since its much harder to see what underwriting criteria they are using.  Everyone thought MBS were risk free instruments as well since they were geographically diversified - this sentiment is held by many on this forum now about P2P.   These are one of the main reasons I'm skeptical to let my portfolio go on an "auto pilot" filter investing strategy, as others on this forum are now doing.  Not looking at stuff critically is how the market for Alt-A and subprime MBS became such a large part of people's portfolio.  One of my buddies in college saw the junk piling up and shorted some entity with a large exposure.  He made an average American's yearly income off of that trade lol.
Title: Re: Verified income
Post by: Fred on March 31, 2013, 09:18:48 PM
Recently Lending Club took the interest rate determining formula from publicly displayed to a proprietary formula.

I think LC still posts their "rates table" -- along with the various modifiers --  here:
https://www.lendingclub.com/public/how-we-set-interest-rates.action


Not looking at stuff critically is how the market for Alt-A and subprime MBS became such a large part of people's portfolio.  One of my buddies in college saw the junk piling up and shorted some entity with a large exposure.  He made an average American's yearly income off of that trade lol.

Ah Alt-A ....  It's been a few years since I last dealt with this term.  The years of living dangerously.  ;)

Once LC becomes a public company, investors of LC Notes can always hedge their LC Notes holding by shorting the LC stocks.




Title: Re: Verified income
Post by: rawraw on April 01, 2013, 07:23:31 AM
I don't know if shorting the stock would be a good hedge -- seems like a risky proposition. 

Quote
If an applicant passes the initial credit screening criteria, the applicant is assessed by Lending Club’s proprietary scoring models which can either decline or approve the applicant. The initial scoring model provides the applicant with a score, which when combined with the applicant’s FICO score and certain other credit attributes arrives at a Model Rank. The Model Rank is based upon an internally developed algorithm which analyzes the performance of Borrower Members and takes into account the applicant’s FICO score, credit attributes, and other application data. The Model Rank is between 1 and 25 and corresponds to a Base Risk Sub-Grade as follows:
This is what I was talking about.  It wasn't like this before.
Title: Re: Verified income
Post by: nscap on April 02, 2013, 06:09:19 PM
Thanks for the enlightening discussion... Rawraw to your point, I guess I am just a little wary of trying to equate to RMBS as in this case, there is no collateral behind the LC loans at all. So, sans that, I feel like the only real form of analysis one can do is income. I guess the flip side is the borrower can lose their job and the income could disappear which is perhaps somewhat analogous to house prices dropping in 08 blowing up the collateral on the loans, but I guess that is just an inherent risk with this kind of investing?

That may have made more sense in my head than typing it out...
Title: Re: Re: Verified income
Post by: rawraw on April 02, 2013, 06:49:15 PM
Thanks for the enlightening discussion... Rawraw to your point, I guess I am just a little wary of trying to equate to RMBS as in this case, there is no collateral behind the LC loans at all. So, sans that, I feel like the only real form of analysis one can do is income. I guess the flip side is the borrower can lose their job and the income could disappear which is perhaps somewhat analogous to house prices dropping in 08 blowing up the collateral on the loans, but I guess that is just an inherent risk with this kind of investing?

That may have made more sense in my head than typing it out...

The problem is for prime borrowers, it annoys them to ask for too much information. They'll quit and get financing elsewhere
Title: Re: Verified income
Post by: Iris on May 31, 2013, 12:59:05 PM
My analysis of the data has shown a marked difference in default rates between
verified and unverified loans in the  same class--this is especially true for
the lower classes. Verified loans do better.
 
LC's prospectus says that they verify 60% of loan applications, and in an NYT
article from a couple of years ago, Lending Club’s co-founder, Renaud Laplanche
explains LC's verification policy, saying: "The main reason why we do not
perform income verification on 100% of the loans is to avoid adverse selection:
the borrowers who have perfect credit history and do not exhibit any particular
risk factors (who fall into the 40% we do not verify) are also those who have
the least tolerance for a cumbersome income verification process, and are most
likely to abandon that process and seek funding elsewhere. They are, however,
the exact kind of borrowers we want to retain."
 
Unfortunately, Laplanche's explanation does not make much sense to me given
other LC data. First, while the group of preferred borrowers is estimated at 40%
of the loan applications, LC accepts only about 10-15% of its loan applications.
If we make the plausible assumption that LC would want to accept loans from
among their better applicants, then the funded loans should only be taken from
among the unverified group--contrary to what the data shows. Any ideas as to how
to resolve the apparent inconsistency?
 
Irit
Title: Re: Verified income
Post by: core on May 31, 2013, 01:14:27 PM
Since nowadays loads are fully funded WAY before verification is complete, I don't see why any of this matters now.  If you stick with verified loans only then you'll get a great return on $0 invested.

Might be an interesting academic discussion but I'm more interested in making money.
Title: Re: Verified income
Post by: Laslow on May 31, 2013, 02:21:22 PM
Although we're no longer able to use income verification as a criteria for filtering, I feel it still has it's uses. How many loans are declined because they failed income verification? We don't have access to this info, but I bet that verification has saved us from many defaulted notes.
Title: Re: Verified income
Post by: core on May 31, 2013, 03:07:59 PM
How many loans are declined because they failed income verification? We don't have access to this info

Sure we do.  It's in the prospectus.  40% of all loans selected for verification do not pass.  As I have written before elsewhere, that either says something good about LC's selection algorithm, or something very bad about the honesty of these borrowers.