Author Topic: Worst Loan of the Day Thread, Grades A - C  (Read 6872 times)

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Worst Loan of the Day Thread, Grades A - C
« on: December 18, 2017, 06:11:06 PM »
This should be fun.

I mentioned in another thread that I still run my auto-invest program every feeding time and score all the new loans. I have it set to invest zero dollars since I'm cashing out but I still get scoring values. So, as time permits, I plan to post what I see as the worst grade A-C loan of the day. Many say the D and lower loans are all terrible (and I have experienced the pain) so I won't go there. Some daily loans will be much worse than others but all will be terrible (at least as I evaluate them). I encourage everyone to post loans they evaluate as even worse than mine:

So, let's start with:

Offered at 5pm EDT on 12/18/2017 a B4 at 10.91% APR:
https://www.lendingclub.com/browse/loanDetail.action?loan_id=125947093&previous=browse

jd

  • Jr. Member
  • **
  • Posts: 65
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #1 on: December 18, 2017, 08:08:56 PM »
I have this one ranked worse than yours:

https://www.lendingclub.com/browse/loanDetail.action?loan_id=126000465

I have added two other loans today (5.32 and 16.02)

Reginald

  • Newbie
  • *
  • Posts: 33
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #2 on: December 18, 2017, 10:07:40 PM »
Good work boys- lets expose this shite.

jd

  • Jr. Member
  • **
  • Posts: 65
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #3 on: December 18, 2017, 10:13:57 PM »
Quote
Good work boys- lets expose this shite

I'm not looking to expose them. I'm still buying loans. I'm just buying ones that I think are good.
« Last Edit: December 19, 2017, 11:31:14 AM by jd »

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #4 on: December 19, 2017, 10:45:24 AM »
I have this one ranked worse than yours:

https://www.lendingclub.com/browse/loanDetail.action?loan_id=126000465

I have added two other loans today (5.32 and 16.02)

Yeah, that one looks terrible. My program pre-screens (filters) a set of loan parameters and if the loan fails to pass the pre-screen then it doesn't even compute a score. This loan was one of those. Unfortunately it doesn't even log the rejection reason, it just tosses out the loan and moves on so it never even scored this loan. The code has some debug switches and I might be able to get some visibility into these rejected loans. Guess by definition many of the worst will be these. It never occurred to me that someday I'd be looking for the worst loans  ::)

BTW: what does "(5.32 and 16.02)" mean?

jd

  • Jr. Member
  • **
  • Posts: 65
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #5 on: December 19, 2017, 11:24:24 AM »
Quote
BTW: what does "(5.32 and 16.02)" mean?

Those are the interest rates of the two loans I got yesterday.


Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #6 on: December 19, 2017, 01:45:00 PM »
Quote
BTW: what does "(5.32 and 16.02)" mean?

Those are the interest rates of the two loans I got yesterday.

Guess it should have been obvious ...

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #7 on: December 19, 2017, 05:51:37 PM »
Okay, I opened up the flood gates so to speak and now get scores on just about all the A,B,C 36 month loans (I don't do 60's).
So many loans, so little time, ... The worst loan of the evening is (drum roll please):

https://www.lendingclub.com/browse/loanDetail.action?loan_id=125988314

A C3 at 14.08%, 9% funded now so there's plenty of time left to invest  :-\

« Last Edit: December 19, 2017, 05:54:30 PM by Rob L »

SLCPaladin

  • Full Member
  • ***
  • Posts: 202
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #8 on: December 19, 2017, 06:26:20 PM »
In addition to posting the duds of the day, I would love it if you also post why you think each loan is the biggest stinker. This would be good for me to understand how other people evaluate bad/risky loans.

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #9 on: December 20, 2017, 11:19:44 AM »
In addition to posting the duds of the day, I would love it if you also post why you think each loan is the biggest stinker. This would be good for me to understand how other people evaluate bad/risky loans.

The short answer is that I don't know. A logistic regression analysis using independent variables such as FICO, inquires last 6 months, employment status, DTI and many others from historic LoanStats data was performed. No results of LC's model such as Grade, Interest Rate, EXPECTED_DEFAULT_RATE, or INSTALLMENT were included as independent variables in the logit analysis. Only borrower data was and should be included. The dependent variable of the analysis was the variable charged off. The analysis result was a set of weighting coefficients that when applied to the same variables in a new loan to be evaluated yields the predicted probability that loan will be charged off or not. Given this probability of being charged off (probability of default or "risk") and the interest rate ("reward") an overall score for the loan is computed. So, there are many many factors that contribute to a loan's score and I have no way to unwind that score into its individual components.

Logistic regression is the classical way consumer loans have been scored for decades. If there is a magic bullet dominant independent variable I am unaware of it. Of course some are more important (more heavily weighted) than others. And some independent variables are of little use as they closely follow others and add little or no value (covariate).  For a good example see the following. It was an interesting and revealing post:

https://forum.lendacademy.com/index.php/topic,3570.msg31593.html#msg31593

For each loan scored I am able to see the probability of default computed by my program and compare it with the EXPECTED_DEFAULT_RATE provided by LC. Since LC began providing a single EXPECTED_DEFAULT_RATE per grade rather than per sub-grade the comparison is less precise. However I am amazed by the magnitude of the differences and there are very few loans per drop where LC's rate is higher than that computed by my program. When LC's rate is higher it isn't ever by much. These differences have grown over time and it even seems as if LC's newer model expects consumer behavior to be improving. On the other hand I spend very little time with LC anymore and the model I'm using is pretty old. Since I'm no longer investing in new notes I'm more of a casual observer so take this with a grain of salt.



« Last Edit: December 20, 2017, 12:07:22 PM by Rob L »

MarinBB

  • Jr. Member
  • **
  • Posts: 83
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #10 on: December 20, 2017, 12:52:17 PM »
In addition to posting the duds of the day, I would love it if you also post why you think each loan is the biggest stinker. This would be good for me to understand how other people evaluate bad/risky loans.

I also passed on https://www.lendingclub.com/browse/loanDetail.action?loan_id=125988314. Negative items for me included that: the payment is large relative to income, credit history is brief, credit utilization is very high, etc.

Fred93

  • Hero Member
  • *****
  • Posts: 2136
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #11 on: December 20, 2017, 01:59:19 PM »
In addition to posting the duds of the day, I would love it if you also post why you think each loan is the biggest stinker. This would be good for me to understand how other people evaluate bad/risky loans.

I also passed on https://www.lendingclub.com/browse/loanDetail.action?loan_id=125988314. Negative items for me included that: the payment is large relative to income, credit history is brief, credit utilization is very high, etc.

Other negatives: few credit lines.  goes along with short credit history to imply very little credit experience.  low income.  Renter.
Positive: The amount of loan nearly matches amount owed, so is likely an intentional credit card payoff.  The data shows that loans perform better when these two numbers closely match.

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #12 on: December 20, 2017, 07:09:11 PM »
Another winner from the 1pm EDT drop:
https://www.lendingclub.com/browse/loanDetail.action?loan_id=125917599
LC assigns this as a B4, 10.91% interest, 4.03% probability of default.
My wacky auto-invest program says it's a 31.8% probability of default; go figure. What can I say.

From the same drop there's also this gem:
https://www.lendingclub.com/browse/loanDetail.action?loan_id=126084792&previous=browse
LC assigns a C4, 15.05% interest, 7.00% probability of default.
My same wacky auto-invest program says it's a 32.2% probability of default, but it scores slightly higher than the first loan since the interest rate (reward) is a tad higher.

If anyone out there is using a third party invests for their account in any of the loans I post in this thread as terrible it would be fascinating to hear about it.
I'll take a sizable bet that will never happen. Bring it on ...

SLCPaladin

  • Full Member
  • ***
  • Posts: 202
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #13 on: December 20, 2017, 07:54:17 PM »
Quote
On the other hand I spend very little time with LC anymore and the model I'm using is pretty old. Since I'm no longer investing in new notes I'm more of a casual observer so take this with a grain of salt.

Thanks for explaining what you've done. Once upon a time I had my own special sauce. Most of what I looked for was gleaned by reading posts here and thinking about what others had suggested. I'm not sure if any of my filters helped improve my returns or not. From what I can tell, the most important factor that correlates with LC returns is not necessarily what notes you've selected, but which vintages (e.g. when) you've invested in.

I've been running off my portfolio for the past year. If charge-offs level off (or go down) and LC raises rates, I may dip my toe back in the water. But my modeling and my note analysis feels rusty from not having picked new notes for some time. I feel like I need to figure out what new factors I should be paying attention to were I to begin investing again.

Out of curiosity, could you briefly explain how you put together your logit analysis? Is this some statistical model you built yourself?
« Last Edit: December 20, 2017, 07:58:24 PM by SLCPaladin »

Rob L

  • Hero Member
  • *****
  • Posts: 2040
    • View Profile
Re: Worst Loan of the Day Thread, Grades A - C
« Reply #14 on: December 21, 2017, 10:27:52 AM »
Quote
On the other hand I spend very little time with LC anymore and the model I'm using is pretty old. Since I'm no longer investing in new notes I'm more of a casual observer so take this with a grain of salt.

Thanks for explaining what you've done. Once upon a time I had my own special sauce. Most of what I looked for was gleaned by reading posts here and thinking about what others had suggested. I'm not sure if any of my filters helped improve my returns or not. From what I can tell, the most important factor that correlates with LC returns is not necessarily what notes you've selected, but which vintages (e.g. when) you've invested in.

I've been running off my portfolio for the past year. If charge-offs level off (or go down) and LC raises rates, I may dip my toe back in the water. But my modeling and my note analysis feels rusty from not having picked new notes for some time. I feel like I need to figure out what new factors I should be paying attention to were I to begin investing again.

Out of curiosity, could you briefly explain how you put together your logit analysis? Is this some statistical model you built yourself?

I have a friend that's pretty good with stats that put the current logit analysis together. I built the real time auto-invest engine to make it go. I've performed a few logit analyses using LC historical data and the statistical package R (open source) and recommend R highly (terminal interface, not GUI). There is also a very good book "Credit Scoring, Response Modeling and Insurance Rating" by Steven Finlay that provides an excellent guide to the whole soup to nuts process of building a credit scoring model.