Author Topic: Question about deliquencies  (Read 6832 times)

ktrdsl23

  • Jr. Member
  • **
  • Posts: 50
    • View Profile
Question about deliquencies
« on: April 27, 2015, 10:30:59 AM »
I've read a lot regarding filtering and a few of the standard ones people like to look at.  While I've seen many discuss limiting the number of inquiries in the last six months I don't see as much discussion regarding delinquencies in the last two years and current accounts now delinquent.  I've been trying to plug some of this into Nickel Steamroller to evaluate without much success. 

So in general do people use that filter a lot and is it telling or do most people assume that many of the target audience (particularly those using this for Debt Consolidation) are doing so because of debt issues so delinquent accounts of the last year or so aren't very important.

Thanks.

kya

  • Jr. Member
  • **
  • Posts: 84
    • View Profile
Re: Question about deliquencies
« Reply #1 on: April 27, 2015, 10:45:47 AM »
i try to avoid both current delinq and delinq last two years.... even with both those being zero in a listing it still doesnt mean there hasnt been some late payments. one handicap for all of us is we cannot see the full credit report.... im sure i would "wince" a bit on some of my notes if i was able to see it....but this game is all about "batting averages".... i do get somewhat concerned that both lc and prosper get used by people on the "way out" ...meaning note listing, fico, job time, income etc all look good but its the customers last grab at $ before they spiral down and out

RazzleDazzle

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
    • reamerge.com
Re: Question about deliquencies
« Reply #2 on: April 27, 2015, 01:58:30 PM »
Interesting and a very valid point kya

While I haven't personally met with people who do this last ditch money grab before defaulting on a somewhat different but relevant note, I know of folks who "improve" their FICO score pretty easily, get a loan then go back to sucky FICO. It is interesting that they are able to make payments but their FICO hangs in 650 ish not 700+

ktrdsl23

  • Jr. Member
  • **
  • Posts: 50
    • View Profile
Re: Question about deliquencies
« Reply #3 on: April 27, 2015, 02:12:36 PM »
What I find strange is that if I load my saved filters in Nickel Steamroller (many of the standards) I see 11.43% historical returns (this is without using either of the delinq categories).  If I change only delinquencies in the last 2 years to 0-0 it goes down to 11.39%.  If I put in 1 for the minimum delinquencies it then goes up to 11.55%.  If I put in 1 for the minimum for months since last delinq it goes up further to 11.60%.

So surprisingly adding in a filter specifying delinqs actually raised the historical returns.  In my mind it would be silly to specifically search for people with delinquencies on their record however this data does make me think there is no purpose to intentionally discarding those individuals.

Thoughts?

kya

  • Jr. Member
  • **
  • Posts: 84
    • View Profile
Re: Question about deliquencies
« Reply #4 on: April 27, 2015, 03:59:02 PM »
filter without eliminating delinq opens your filter up to higher yielding notes...again batting averages

AnilG

  • Hero Member
  • *****
  • Posts: 1063
    • View Profile
    • PeerCube
Re: Question about deliquencies
« Reply #5 on: April 27, 2015, 04:37:10 PM »
Your results may be biased because most loans to borrowers with 7+ delinquencies in last 2 years were first introduced in 2010. Majority of loans to borrowers with increasing delinquencies in last 2 years carry higher interest rate, poor credit grade, longer loan term, and smaller loan volume. All factors contribute to appearance of higher "phantom" returns.

Did you perform your analysis only on matured loans? Also analyze the loss rate in addition to returns. Most amateur investors focus on return without considering the loss rate. You have to find the right balance between risk and return. Always try to understand and explore the data.

This chart from https://www.peercube.com/histperf/perfbyattr/delinq_2yrs might be of interest to you.




So surprisingly adding in a filter specifying delinqs actually raised the historical returns.  In my mind it would be silly to specifically search for people with delinquencies on their record however this data does make me think there is no purpose to intentionally discarding those individuals.

Thoughts?
« Last Edit: April 27, 2015, 04:41:41 PM by AnilG »
---
Anil Gupta
PeerCube Thoughts blog https://www.peercube.com/blog
PeerCube https://www.peercube.com

LoanShark

  • Newbie
  • *
  • Posts: 18
    • View Profile
Re: Question about deliquencies
« Reply #6 on: April 27, 2015, 10:57:48 PM »
As Anil said, you can look at seasoned loans to get a better idea of loss rates.  I also look at the average age.  If one filter produces 7% losses with 16 months average age, and another gives 7.5% losses with 21 months average age, I'd like the second one more.

LC's credit model seems to control for FICO/deliquincies pretty well.  Lower ficos with more delinquencies are charged higher interest rates.  Inq 6 month is one of the few gaping loopholes. 

Smkj79

  • Newbie
  • *
  • Posts: 23
    • View Profile
Re: Question about deliquencies
« Reply #7 on: April 29, 2015, 08:27:19 AM »
What time frame is a note/portfolio considered seasoned?

lascott

  • Hero Member
  • *****
  • Posts: 1360
    • View Profile
    • Appreciate my post and want to try LendingRobot? URL below
Re: Question about deliquencies
« Reply #8 on: April 29, 2015, 10:33:59 AM »
What time frame is a note/portfolio considered seasoned?
Prosper uses 10 months.
Seems to depend a little on the grade if you eyeball this LendingRobot chart.  10-18 months is common referred to as seasoned for LC notes as I recall.

Via: https://www.lendingrobot.com/#/resources/charts
Tools I use: (main) BlueVestment: https://www.bluevestment.com/app/pricing + https://www.interestradar.com/ , (others) Lending Robot referral link: https://www.lendingrobot.com/ref/scott473/  & Peercube referral code: DFVA9Y

AnilG

  • Hero Member
  • *****
  • Posts: 1063
    • View Profile
    • PeerCube
Re: Question about deliquencies
« Reply #9 on: April 29, 2015, 01:14:31 PM »
I consider "seasoned" loans to be when most of delinquency risk has been accounted for. In Lending Club case, that would be 13 to 16 months old. I don't consider "seasoned" loans to be same as "matured" loans. The "matured loans" are those for whom at least one vintage has gone through the complete cycle, from cradle to grave, or very close to completing this cycle. IMO, at this point, any return/loss results for loans issued after Q2, 2012 should not be construed as representative long term performance of loans. With amortized nature of loans, the returns are overstated and losses are understated. Of course for a portfolio that continually re-invests and never liquidates, the seasoned loans offer a decent performance benchmark.



What time frame is a note/portfolio considered seasoned?
---
Anil Gupta
PeerCube Thoughts blog https://www.peercube.com/blog
PeerCube https://www.peercube.com

rawraw

  • Hero Member
  • *****
  • Posts: 2744
    • View Profile
Re: Question about deliquencies
« Reply #10 on: April 29, 2015, 02:44:33 PM »
The returns will move up or down with LC's pricing.  The probability of default should be more stable and is what people should pay attention to IMO.  Then adjust current yields downward by that assumed POD