Mad Rush at Lending Club Loan Release Time: Part III - Delinquency Rate and FICO Score Change
Mad Rush at Lending Club Loan Release Time: Part V - Loan Term with Time to Fund
Mad Rush at Lending Club Loan Release Time: Part IV - Interest Rate with Time to Fund
This post is the fourth part of the Mad Rush at Lending Club Loan Release Time series analyzing loan attributes and performance with Time to Fund (TTF). You may want to consider reading the first three parts to get proper background:
- Mad Rush at Lending Club Loan Release Time: Part I
- Mad Rush at Loan Release Time: Part II - Loan Performance with Time to Fund
- Mad Rush at Lending Club Loan Release Time: Part III - Delinquency Rate and FICO Score Change
In the last three posts, we reviewed data collection and analysis methodology and analyzed data for listing and issued volume, charged off rate, bad to good loan ratio, delinquency rate, and FICO score change with Time to Fund and Credit Grades.
One criticism we often received to date was the possible influence of interest rate and resulting impact on delinquencies. Though we anticipated and tried to address using credit grades as interest rates and credit grades are closely linked, still it didn’t convince a few readers. In this post we address interest rate and delinquencies with time to fund.
Figure 1 below shows the Interest Rate and Delinquency Rate with Time to Fund for issued loans. The interest rate in percentage is plotted on left axis.
The delinquency rate for each time to fund group is shown as orange
+ sign. The scale for delinquency rate in percentage is on the right axis. The horizontal dashed orange line shows the delinquency rate for all loans in this dataset.
The interest rate for each time to fund group is plotted as box-and-whisker plot. The blue box shows the quartiles. The lines extending vertically from the box indicate the variability outside the upper (75%) and lower (25%) quartiles.The line within the box where blue color transitions from light to dark shade shows the median. The horizontal blue line shows the median interest rate and the horizontal gray lines show the lower quartile (25%) and upper quartile (75%) interest rate for all loans in this dataset.
Table 1 below shows the lower quartile, median, and upper quartile interest rate and delinquency rate for first 60 minutes of time to fund.
|Time to Fund||Interest Rate||Delinquency Rate|
|Upper Quartile||Median||Lower Quartile||Spread|
|0 - 1 min||23.10%||18.85%||14.30%||8.80%||4.14%|
|2 - 3 min||23.10%||18.70%||13.68%||9.42%||3.85%|
|4 - 10 min||23.40%||18.85%||13.68%||9.72%||3.44%|
|11 - 60 min||22.70%||18.25%||13.05%||9.65%||3.17%|
|All Issued Loans||22.90%||18.01%||12.67%||10.23%||2.99%|
The difference in interest rate between the issued loans that are funded the earliest, within one minute, and all issued loans is 0.20% for lower quartile, 0.84% for median and 1.63% for upper quartile. It is true that the interest rate for fastest funded loans is higher and spread smaller than that for loans in aggregate but same is true for loans funded between 2 and 3 minutes, between 4 and 10 minutes, and between 11 and 60 minutes.
Following observations can be drawn from Figure 1 and Table 1:
- There is no relationship between interest rate and delinquency rate for loans funded within first 60 minutes.
- The issued loans that were funded within first 60 minutes are similar when compared on the basis of interest rate.
- The data doesn’t support the impression of higher delinquency rate for loans funded within a minute is due to higher interest rate.
The above observations can also be verified by Figure 2 that plots listing volume with interest rate by time to fund. The listed loans that were funded within first 60 minutes carried higher interest rate profile compared to listed loans that took more than an hour to fully fund. But there is no significant difference in interest rate profiles of listed loans that were funded within a minute, between 2 and 3 minutes, between 4 and 10 minutes, and between 11 and 60 minutes respectively.
These observations lead us to the following conclusions:
- There is no benefit, from interest rate profile perspective, for lenders to rush to try to invest within a minute or two of loans being released.
- By waiting another 10 minutes after loans are released, lenders can potentially reduce the delinquency rate by about 25% while still maintaining the similar interest rate profile.
Figure 3 below shows the Interest Rate and Delinquency Rate with Time to Fund for Issued Loans with Grade A and B. Figure 4 below shows listing volume with interest rate by time to fund for Grade A and B loans.
The issued loans with Grade A and B that were fully funded within a minute of being listed has the highest interest rate spread (7.76% to 12.67%) among all Grade A and B loans issued in this study. The interest rate profile of Grade A and B loans that were fully funded within a minute is very similar to loans that took little longer (2 - 10 minutes). While the delinquency rate is lowest (1.04%) for Grade A and B loans funded within a minute among the loans that were fully funded within 10 minutes of being listed. This appears to agree with our prior finding that the speed pays when a lender is investing in Grade A and B loans.
Figure 5 below shows the Interest Rate and Delinquency Rate with Time to Fund for Issued Loans with Grade C, D, and E. Figure 6 below shows listing volume with interest rate by time to fund for Grade C, D, and E loans.
The case for Grade C, D, and E loans is very interesting. The median interest rate and spread is almost same for such loans funded within 5 hours of being listed. Even interest rate profile for such loans funded within 60 minutes is almost the same. The delinquency rate for such loans declines with longer time to fund up to 3 hours. What these observations indicate is that there is no benefit of speed with investing in C, D, and E loans because interest rate profile is same and only deciding criteria can be lower delinquency rate.
Figure 7 below shows the Interest Rate and Delinquency Rate with Time to Fund for Issued Loans with Grade F and G. Figure 8 below shows listing volume with interest rate by time to fund for Grade F and G loans.
Unlike Grade A through E loans, the median interest rate for Credit Grade F and G that were fully funded early is lower. The lower bound of interest rate spread is also lower than that for all Grade F and G loans. It appears while automated lenders were trying to quickly grab higher yielding Grade A and B loans, with Grade F and G loans they are trying to quickly grab lower yielding loans. Despite the earliest funded loans have lower median interest rate, the delinquencies are higher. Also, there is no identifying pattern in interest rate profile with the time to fund for Grade F and G loans. Once again, it appears that lender may benefit by waiting a few minutes after loans are released to select Grade F and G loans.
- There is no benefit for lenders to rush to invest within a minute or two of loans being released except for Grade A and B loans.
- By waiting for a few minutes, lenders can reduce delinquencies significantly while still maintaining somewhat similar interest rate profile as that for quickest funded loans.
In the next post, we will review loan term, 36 months versus 60 months, with time to fund.
Normally before publishing a post, I send out a draft for review and feedback to a few readers whose opinions I trust due to the quality of our past interactions. Below is excerpt from one of them on this blog post who I thought sums up the gist of the post nicely.
I find the reinforcement of your previously stated thesis of the value of speed for A and B loans significant. Also higher delinquency rate for loans is not necessarily due to higher rates. For non A and B loans speed of commitment is not that essential. Speed in picking loans for the shopping cart is essential–but I’m noticing more desirable – for me – loans available after 5 - 10 minutes. It generally takes me between 4 and 10 minutes to “pull the trigger”. I note that time period has an extremely low delinquency rate. Makes me feel good!
Regarding my portfolio–E, F and G – as you say–investors should take their time. Get the notes in the shopping cart and then carefully analyze. Anyone who “pulls the trigger” within a minute or two is making a mistake. It takes longer than that to analyze the notes and the analysis –with all the variables present–should only be done manually.
I have noticed an anomaly in G notes. There seems to be some reluctance to invest in them! I will look at a G note with its high interest rate and say: Now what is wrong with this? No delinquencies, no charge offs, 10 years employment, good salary, good D to I rating. 670–690 FICO, age 35–45, etc. The only negative is its G rating! – One other observation is the speed that 3 year loans are funded. I find loans with several blemishes being funded within seconds. Are they that more valuable than five year loans?
- JJ Hendricks | Wednesday July 9, 2014, 2:00 pm
- I've loved this series. Good to know that we can stay out of the rat-race to buy during the first minute and still get good loans. Do you ever do statistical test to verify there is no difference between averages? Like in table 1, you say "There is no relationship between interest rate and delinquency rate for loans funded within first 60 minutes." Did you do a t-test for difference between means?
- Anil Gupta | Wednesday July 9, 2014, 5:09 pm
- @JJ, thanks for the comment. Definitely, this analysis shows that being early doesn't payoff generally. Yes, statistical tests are frequently used to verify the results. I try to keep the mathematical and statistical information out of the posts to keep posts readable and of appropriate length.
- JJ Hendricks | Tuesday July 15, 2014, 3:20 pm
- @Anil - Thank you for the response. I can definitely understand the readability and length reasoning.