Default | nLoans |
0
Title: Effect of Previous Delinquencies on Probability of Default
Post by: Fred on December 16, 2015, 11:00:00 PM
I think a few people in this forum has done extensive "factor analysis"-- https://en.wikipedia.org/wiki/Factor_analysis -- on all the attributes in LC historical data. Not sure if delinq_2yrs really has a negative coefficient, and if it is statistically significant.
However, if you really think your analysis is correct, you should follow your numbers instead of your assumptions -- i.e., buy LC loans with delinq_2yrs > 0. Good luck.
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 16, 2015, 11:00:00 PM
I've run some basic logistic regressions, sampled and not, and with other factors included, the coeficient is significantly negative. As a modeler, anything that doesn't make economic sense, I would generally toss out. Trying to understand if there is a world where this makes sense.
One thing I have come up with is LC's underwriting possibly rejects borrowers with higher delinquencies with a higher probability to default. Those that make it through are better credit risks and have a lower probability to default.
Title: Effect of Previous Delinquencies on Probability of Default
Post by: jennrod12 on December 16, 2015, 11:00:00 PM
If I read that right, you are looking at loans that defaulted and what % of them were from borrowers with delinquincies?
Did you look at all loans from borrowers with delinquincies and what % of them defaulted? Jenn
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 16, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 16, 2015, 11:00:00 PM
Interesting info,
Is there a difference when looked at by the borrower's number of delinquencies or how many months since the last delinquency? Jenn
Title: Effect of Previous Delinquencies on Probability of Default
Post by: AnilG on December 16, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 20, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: lascott on December 20, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: hoggy1 on December 20, 2015, 11:00:00 PM
As I understand it, a Delinquency is 30 or more days late and I don't count one in the last two years against the borrower. By contrast a major derogatory is more than 120 days late and I don't buy notes with any.
Title: Effect of Previous Delinquencies on Probability of Default
Post by: Fred93 on December 20, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: brycemason on December 22, 2015, 11:00:00 PM
Being stumped by covariance is so 2011 on this forum, but you caught me in a good mood after a pleasant 16-hour day of building loan servicing logic. Data: LC 9/30/2015 extract, matured policy code 1 36-month loans between 3 and 5 years old (out of convenience); various other exclusions I'll cite later.
Define prior_delinq to be 1 just in case delinq_2yrs is greater than 0. Useful practice because of the few data points past 1 anyway. Avoids estimation issues. . tab delinq_2yrs delinq_2yrs | Freq. Percent Cum. ------------+----------------------------------- 0 | 39,617 89.42 89.42 1 | 3,493 7.88 97.31 2 | 779 1.76 99.07 3 | 243 0.55 99.62 4 | 73 0.16 99.78 5 | 47 0.11 99.89 6 | 23 0.05 99.94 7 | 12 0.03 99.97 8 | 3 0.01 99.97 9 | 4 0.01 99.98 10 | 3 0.01 99.99 11 | 4 0.01 100.00 18 | 1 0.00 100.00 ------------+----------------------------------- Total | 44,302 100.00 Observation (single-variable analysis): Having a prior delinquency increases the chance of a charge off event. Logistic regression Number of obs = 44302 LR chi2(1) = 27.03 Prob > chi2 = 0.0000 Log likelihood = -16924.115 Pseudo R2 = 0.0008 ------------------------------------------------------------------------------ chargeoff | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- prior_delinq | .2302795 .043396 5.31 0.000 .145225 .3153341 _cons | -1.94617 .015193 -128.10 0.000 -1.975947 -1.916392 ------------------------------------------------------------------------------ Observation (Multi-variate analysis): This relationship vanishes when controlling for the four horsemen of the consumer credit scoring apocalypse. Logistic regression Number of obs = 44302 LR chi2(5) = 1575.80 Prob > chi2 = 0.0000 Log likelihood = -16149.729 Pseudo R2 = 0.0465 -------------------------------------------------------------------------------- chargeoff | Coef. Std. Err. z P>|z| [95% Conf. Interval] ---------------+---------------------------------------------------------------- prior_delinq | -.0021604 .044922 -0.05 0.962 -.0902059 .0858851 fico_range_low | -.0153585 .0005158 -29.78 0.000 -.0163695 -.0143476 inq_last_6mths | .1900362 .013459 14.12 0.000 .163657 .2164155 emp_na_slf_ret | .4404685 .0522608 8.43 0.000 .3380393 .5428978 loan2inc | 2.286919 .1299917 17.59 0.000 2.03214 2.541698 _cons | 8.213554 .3604598 22.79 0.000 7.507065 8.920042 --------------------------------------------------------------------------------
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 22, 2015, 11:00:00 PM
Title: Effect of Previous Delinquencies on Probability of Default
Post by: TravelingPennies on December 22, 2015, 11:00:00 PM
You are super welcome! Don't take my snark personally .
Title: Effect of Previous Delinquencies on Probability of Default
Post by: Peter on December 22, 2015, 11:00:00 PM
*giggle* Welcome, Dompazz. Stick around, I think you're gonna fit in just fine!
|