### Author Topic: Lending Club loan default prediction model question  (Read 3064 times)

#### rawraw

• Hero Member
• Posts: 2784
##### Re: Lending Club loan default prediction model question
« Reply #15 on: February 16, 2019, 07:45:38 AM »
Thanks for all of the replies.  I should have shared a little about myself and my methods.  I have experience building predictive credit models in financial institutions.  My primary tool of choice to build predictive models is R.  I'm very fond of the GLMNET package and my methods resemble Frank Harrells "Regression Modeling Strategies".

Very good. Thanks for the tip on the book.
Please share a bit more of your experience if you will. It would be so interesting so see how things are now.
Claim "secret sauce" where appropriate.

1) Is LC offering enough loans that meet your criteria for you to be able to stay fully invested? Would it be too much to ask that \$ amount?
2) Presumably you are using the API to access new loans at the four "feeding times". Is there still a race? Do you consider speed important?
3) What's the Term and Grade allocation of your portfolio 36(%A, %B, ...) and 60(%A, %B, ...) where %x is a percent of the total \$ principal invested?

TIA

My modeling method is using a Cox Prop Hazard multivariate survival model tuned with GLMNET.  Nothing really special.  I've never put a survival model in production and wanted to give it a go.  I've worked in auto finance for the last 7 years and have done applied math and data analysis for most of my career.  I've built credit scoring models for large lenders.  It is actually quite fun in my opinion.

1) to be honest I do it as a hobby.  I've only invested a few thousand in the last couple of years.  I'm doing it to keep my chops up and it interests me.
2) yes I'm using the API.  I don't invest enough frequency to see if speed is important
3) 36: B 10%, C 33%, D 17%, E 14%     60: B 2%, C 11%, D 4%, E 5%, F/G 2%
Are there any resources you'd recommend for someone starting in auto finance and predictive modeling?

#### larrydag

• Newbie
• Posts: 14
##### Re: Lending Club loan default prediction model question
« Reply #16 on: February 16, 2019, 08:33:29 AM »
There are a lot of opportunities to get started in auto finance if you have the right skillsets.  The typical skillsets that auto finance companies look for are STEM degrees and business degrees.  You can easily look up on a job aggregator to see the job descriptions.  Most auto finance companies are like every other company in they want to be able to make data driven decisions about loan applicants ability to repay on loans.  If you don't have previous financial or lending experience I believe you can still get in the door at an analyst or IT developer level and build your experience.  Even if you can't find an auto finance job you can find an analyst job at a bank and learn about credit and lending in that position.  The important things to know in auto finance is credit bureau data and loan portfolio management.

Getting started in predictive modeling is more broad.  There is a huge swath of companies and industries looking for that skill.  In fact even if you current job doesn't require you could probably take it on as as side project and show how your model would help your current organization.  Here is the secret untold story about predictive modeling that most academics do not tell you.  Predictive modeling is 80% data acquisition and management and 20% modeling.  So make sure you are a data skill hawk meaning that you can download, pull, connect, manipulate, slice, dice, warehouse, store, and distribute data.  That means having skills in SQL, Python, R or other data programming tool.   Trust me your bosses would like it even if you can just manage multiple data sources and provide meaningful data analysis.  Chances are the business decision makers in an organization doesn't know how and doesn't know the data exists.

Getting your chops up in the statistical and applied math of predictive modeling can be done on your own via online learning or in a more structured classroom setting.  Do it in baby steps if you have no applied math background.  Start with basic statistics 101 and move on to more advanced.

#### Rob L

• Hero Member
• Posts: 2096
##### Re: Lending Club loan default prediction model question
« Reply #17 on: February 16, 2019, 09:22:46 AM »
Thanks for all of the replies.  I should have shared a little about myself and my methods.  I have experience building predictive credit models in financial institutions.  My primary tool of choice to build predictive models is R.  I'm very fond of the GLMNET package and my methods resemble Frank Harrells "Regression Modeling Strategies".

Very good. Thanks for the tip on the book.
Please share a bit more of your experience if you will. It would be so interesting so see how things are now.
Claim "secret sauce" where appropriate.

1) Is LC offering enough loans that meet your criteria for you to be able to stay fully invested? Would it be too much to ask that \$ amount?
2) Presumably you are using the API to access new loans at the four "feeding times". Is there still a race? Do you consider speed important?
3) What's the Term and Grade allocation of your portfolio 36(%A, %B, ...) and 60(%A, %B, ...) where %x is a percent of the total \$ principal invested?

TIA

My modeling method is using a Cox Prop Hazard multivariate survival model tuned with GLMNET.  Nothing really special.  I've never put a survival model in production and wanted to give it a go.  I've worked in auto finance for the last 7 years and have done applied math and data analysis for most of my career.  I've built credit scoring models for large lenders.  It is actually quite fun in my opinion.

1) to be honest I do it as a hobby.  I've only invested a few thousand in the last couple of years.  I'm doing it to keep my chops up and it interests me.
2) yes I'm using the API.  I don't invest enough frequency to see if speed is important
3) 36: B 10%, C 33%, D 17%, E 14%     60: B 2%, C 11%, D 4%, E 5%, F/G 2%

Thanks for the reply. As I mentioned earlier it was a very nice hobby for me as well. If you had been lucky enough to stumble across LC back in 2013 you could have made some serious money with your skills. Plenty of good loans from which to select. From the previous post in this thread by mikedev10 there's just not very many loans offered any more.

#### rawraw

• Hero Member
• Posts: 2784
##### Re: Lending Club loan default prediction model question
« Reply #18 on: February 16, 2019, 08:50:58 PM »
There are a lot of opportunities to get started in auto finance if you have the right skillsets.  The typical skillsets that auto finance companies look for are STEM degrees and business degrees.  You can easily look up on a job aggregator to see the job descriptions.  Most auto finance companies are like every other company in they want to be able to make data driven decisions about loan applicants ability to repay on loans.  If you don't have previous financial or lending experience I believe you can still get in the door at an analyst or IT developer level and build your experience.  Even if you can't find an auto finance job you can find an analyst job at a bank and learn about credit and lending in that position.  The important things to know in auto finance is credit bureau data and loan portfolio management.

Getting started in predictive modeling is more broad.  There is a huge swath of companies and industries looking for that skill.  In fact even if you current job doesn't require you could probably take it on as as side project and show how your model would help your current organization.  Here is the secret untold story about predictive modeling that most academics do not tell you.  Predictive modeling is 80% data acquisition and management and 20% modeling.  So make sure you are a data skill hawk meaning that you can download, pull, connect, manipulate, slice, dice, warehouse, store, and distribute data.  That means having skills in SQL, Python, R or other data programming tool.   Trust me your bosses would like it even if you can just manage multiple data sources and provide meaningful data analysis.  Chances are the business decision makers in an organization doesn't know how and doesn't know the data exists.

Getting your chops up in the statistical and applied math of predictive modeling can be done on your own via online learning or in a more structured classroom setting.  Do it in baby steps if you have no applied math background.  Start with basic statistics 101 and move on to more advanced.
Thanks. I have broad experience in the lending and data analysis. But I've started to learn python and may be moving into a role that is for an auto lender.  Should be fun if it happens

#### Rob L

• Hero Member
• Posts: 2096
##### Re: Lending Club loan default prediction model question
« Reply #19 on: February 17, 2019, 09:12:27 AM »
There are a lot of opportunities to get started in auto finance if you have the right skillsets.  The typical skillsets that auto finance companies look for are STEM degrees and business degrees.  You can easily look up on a job aggregator to see the job descriptions.  Most auto finance companies are like every other company in they want to be able to make data driven decisions about loan applicants ability to repay on loans.  If you don't have previous financial or lending experience I believe you can still get in the door at an analyst or IT developer level and build your experience.  Even if you can't find an auto finance job you can find an analyst job at a bank and learn about credit and lending in that position.  The important things to know in auto finance is credit bureau data and loan portfolio management.

Getting started in predictive modeling is more broad.  There is a huge swath of companies and industries looking for that skill.  In fact even if you current job doesn't require you could probably take it on as as side project and show how your model would help your current organization.  Here is the secret untold story about predictive modeling that most academics do not tell you.  Predictive modeling is 80% data acquisition and management and 20% modeling.  So make sure you are a data skill hawk meaning that you can download, pull, connect, manipulate, slice, dice, warehouse, store, and distribute data.  That means having skills in SQL, Python, R or other data programming tool.   Trust me your bosses would like it even if you can just manage multiple data sources and provide meaningful data analysis.  Chances are the business decision makers in an organization doesn't know how and doesn't know the data exists.

Getting your chops up in the statistical and applied math of predictive modeling can be done on your own via online learning or in a more structured classroom setting.  Do it in baby steps if you have no applied math background.  Start with basic statistics 101 and move on to more advanced.
Thanks. I have broad experience in the lending and data analysis. But I've started to learn python and may be moving into a role that is for an auto lender.  Should be fun if it happens

Does sound like fun. Good luck!

#### larrydag

• Newbie
• Posts: 14
##### Re: Lending Club loan default prediction model question
« Reply #20 on: February 18, 2019, 05:53:00 PM »
Good luck to you rawraw.  You'll find auto finance to be a rewarding and challenging industry

#### rawraw

• Hero Member
• Posts: 2784
##### Re: Lending Club loan default prediction model question
« Reply #21 on: February 20, 2019, 08:08:50 PM »
Good luck to you rawraw.  You'll find auto finance to be a rewarding and challenging industry
It certainly seems that way. I've had exposure to it via reviewing the models and such, but never created them myself.    Thanks for  the luck:)