Do you think loan quality (as measured by say NAR by vintage) since the poor vintage period you cited has improved?
Perhaps improvements in lower risk categories but not higher ones, or in 60 but not 36 month term, etc.? I dunno.
Seems CircleT009 might think that the poor performance was/is a calendar period period problem, not a vintage one. Did I get that right?
Well, NAR is particularly ill-suited for this sort of comparison, because it means a different thing at different points in a loan's life.
LC publishes delinquency & default data broken out by vintage and age, and in that data you can see that different vintages have performed differently.
I believe there is some component of BOTH vintage and calendar period in the results. I just think the vintage differentiation is stronger.
I can't readily tell from
https://www.lendingclub.com/info/demand-and-credit-profile.action.[/quote]
Agreed. Not enough breakout. Try this alternative...
https://www.insikt.com/#/invest/mycro/vintage/cumLoss?po=Partner&or=LendingClub&fr=Quarterly&tY=2017&tM=10&fY=2016&fM=1&fi=creditRating&sGO=B<=36,60This web site lets you break out the data various ways.
You have to play with it awhile to get familiar with the UI. I think you can see that in 2015 and much of 2016, newer vintages were worse than the one before them, and this has stopped. Depends of course on which grades and terms you break out, which measure you look at (late 30, late 60, default). It is very easy to get so many curves that you can't see the more recent ones, so you have to set it for a large date range to get context, and then set it for a narrow date range to see the relative position of recent quarters.
You can also download the delinquency and chargeoff by vintage spreadsheets from LC. The links are quietly hidden at the bottom of one of the statistics pages. The spreadsheet lets you choose grades, and displays a matrix with vintage columns and month (payment) number rows. I copy & paste this to another spreadsheet where I draw charts from it. I haven't updated in a long time. Another thing you can do is simply look at a row. For example, pick the 12 month row, and just look across vintages to see if newer vintages are worse or better at 12 months, etc.