A first look at the "unlucky" question

To address the questions raised in my last post, I need to look at some Lending Club data. Fortunately, Lending Club offers a large amount of data and in a downloadable format (one of the reasons I chose Lending Club over Prosper several years ago). Unfortunately, there is so much data that it is hard to handle in simple spreadsheets. For instance, when I download the data on my notes, I get 168 columns of data, mostly uninteresting. As for all notes data, there are over a million records.

I wrote a program to condense the records by selecting only those columns of interest. These can then be manipulated in a spreadsheet. I was also able to place on flag each note’s record as an indicator if it was a note that I owned. I used this data to begin answering the first question.

Have I just been unlucky?

Quantifying this turns out to be more complicated than it should. First, the notes have been issued over different times and for different terms (36 or 60 months). Next, while each has a transactional history, that is too much and too complex to deal with. Finally, what we do have is a snapshot of each note in mid-2017. Each note has a status reflecting where it stands at that moment. The initial and most prevalent status is “Current”, meaning the borrower is making their scheduled payments. If the borrower fails to make payments on time, their status moves through varying stages of lateness. I classify these as “Other”. Finally notes reach a final status of “Fully Paid” or “Charged Off”.

I’ve grouped notes issued in the last two years by the quarter in which they were issued. Following is a table depicting the percentage by status of all notes.

ALL NOTES	Count	Fully Paid	Current	Other	Charged Off
15_Q1	84,277	37.2%	46.8%	2.9%	13.1%
15_Q2	95,825	31.9%	52.9%	3.2%	11.9%
15_Q3	110,489	27.5%	59.1%	3.3%	10.1%
15_Q4	130,503	23.5%	64.8%	3.6%	8.1%
16_Q1	133,887	18.8%	71.0%	3.9%	6.3%
16_Q2	97,854	15.3%	76.1%	4.0%	4.5%
16_Q3	99,120	13.2%	80.2%	4.1%	2.5%
16_Q4	103,546	9.3%	87.0%	3.0%	0.7%
Total	855,501	22.1%	67.2%	3.5%	7.2%

Next is the same chart but only with notes I own.

MY NOTES	Count	Fully Paid	Current	Other	Charged Off
15_Q1	365	33.2%	41.9%	4.1%	20.8%
15_Q2	38	28.9%	50.0%	5.3%	15.8%
15_Q3	1,664	24.9%	53.4%	4.0%	17.7%
15_Q4	308	18.2%	58.1%	5.2%	18.5%
16_Q1	300	14.7%	77.0%	3.7%	4.7%
16_Q2	287	12.5%	74.2%	6.3%	7.0%
16_Q3	289	14.9%	80.3%	2.1%	2.8%
16_Q4	235	6.8%	90.2%	3.0%	0.0%
Total	3,486	19.3%	65.6%	4.2%	10.9%

It’s clear that my notes have performed more poorly when compared to notes of comparable age. But it’s still too early to draw a definitive conclusion. It’s possible that the difference might be due to the difference in loan grade distribution. I’ll address this in my next post. Follow me on Twitter, @billlanke, for notice as to when that occurs.