Author Topic: LC delinquencies went DOWN in Feb & March 2016  (Read 4980 times)

Fred93

  • Hero Member
  • *****
  • Posts: 2221
    • View Profile
LC delinquencies went DOWN in Feb & March 2016
« on: May 26, 2016, 05:18:22 AM »
The charts at insikt tell the story.  They've now been updated with the Q1 data dump from LC.  All-portfolio delinquency rates went down during the 1st quarter.  Sorry this isn't one of those ever-popular doomsday movie plot stories.  Just the facts.

If you go to insikt.com and choose the right options to get that chart to appear, you can mouseover to see the data values.  The last data point on the right is March 2016.  In this presentation, each color is one credit grade.  A at the bottom, G at the top.

« Last Edit: May 26, 2016, 05:21:59 AM by Fred93 »

Fred

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #1 on: May 26, 2016, 06:07:32 AM »
Thanks for posting this.

Ever since the big LC shake-up recently, many people in this forum worried about BRV.

However, my biggest worry has been the delinquency of existing loans.  I have been trying to get the latest delinquency statistics, but could not find time to do it.   

So, thank you again for posting the good news. Hopefully, no surprises in the next few weeks.

LonghornSF

  • Full Member
  • ***
  • Posts: 208
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #2 on: May 26, 2016, 09:25:37 AM »
Looks to me like a multiyear upward trend. How much did delinquencies decline in the 1Q, a few basis points? I can't even tell if they declined in that chart.

rawraw

  • Hero Member
  • *****
  • Posts: 2790
    • View Profile
LC delinquencies went DOWN in Feb & March 2016
« Reply #3 on: May 26, 2016, 09:52:51 AM »
These sorts of graphs are not very helpful due to growth. The vintage graphs in LC filings are much more useful

Sent from my SAMSUNG-SM-G935A using Tapatalk


Emmanuel

  • Full Member
  • ***
  • Posts: 157
    • View Profile
    • LendingRobot
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #4 on: May 26, 2016, 10:47:46 AM »
These sorts of graphs are not very helpful due to growth. The vintage graphs in LC filings are much more useful

Good point.
It's possible to compensate for that, see http://blog.lendingrobot.com/industry/a-review-of-lending-club-defaults/

Fred93

  • Hero Member
  • *****
  • Posts: 2221
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #5 on: May 26, 2016, 04:18:53 PM »
Looks to me like a multiyear upward trend. How much did delinquencies decline in the 1Q, a few basis points? I can't even tell if they declined in that chart.

Go to the insikt web site and mouseover the chart there.  The numbers appear.

Fred93

  • Hero Member
  • *****
  • Posts: 2221
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #6 on: May 26, 2016, 04:22:36 PM »
These sorts of graphs are not very helpful due to growth. The vintage graphs in LC filings are much more useful

Different ways of charting the data are useful for different purposes.  Vintage curves are useful.  They show you whether one vintage has a different performance than another, which could be due to things like LC's borrower customer base changing or underwriting standards changing over time.  However, when you want to see the effects of something like changes in the economy, which affect all vintages at the same time, it is best to use a chart that puts all the affected payments in one place.  On a vintage chart, "this month" is in a different place on every vintage curve, so you can't see such things very well at all.

By the way, if you want to break out by vintage, there's a button for that on insikt also.

rawraw

  • Hero Member
  • *****
  • Posts: 2790
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #7 on: May 26, 2016, 04:40:27 PM »
These sorts of graphs are not very helpful due to growth. The vintage graphs in LC filings are much more useful

Different ways of charting the data are useful for different purposes.  Vintage curves are useful.  They show you whether one vintage has a different performance than another, which could be due to things like LC's borrower customer base changing or underwriting standards changing over time.  However, when you want to see the effects of something like changes in the economy, which affect all vintages at the same time, it is best to use a chart that puts all the affected payments in one place.  On a vintage chart, "this month" is in a different place on every vintage curve, so you can't see such things very well at all.
No, this form of charting isn't useful at all.  It will hide risk very well and then blow up in your face.  It's the oldest trick in the book that growth hides problems.

In terms of the above, I'd have to disagree.  It impacts them at the same time sure.  But that isn't the same thing as at the same place in the graph.  At a firm with LC's growth profile, you will always need the vintage to see whats going on.  I can't think of any benefits of the above graph for fast growing firms.  But I look at these sort of graphs for a living, so if I'm missing something feel free to share.  In financials, high growth firms are always a challenge because they don't provide data sufficient for vintages, like LC.

mchu168

  • Full Member
  • ***
  • Posts: 147
  • “Integrity is the ability to stand by an idea.”
    • View Profile
    • Email
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #8 on: May 26, 2016, 04:49:51 PM »
No, this form of charting isn't useful at all.  It will hide risk very well and then blow up in your face.  It's the oldest trick in the book that growth hides problems.

Disregarding this data is simply confirmation bias for the bears. Like most people, I think we all seek confirmation, and this data set (as flawed as it may be) doesn't conform to your prediction that note performance will continue to deteriorate. You reasons are probably somewhat valid, but I think it's useful to identify yourself as a bear, as I and a few others openly admit that we're still bullish on the asset class.  AnilG is another bear who has been vocally dismissive of future returns.  Who knows what the future holds, but I think we all need to keep open minds.


Fred93

  • Hero Member
  • *****
  • Posts: 2221
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #9 on: May 26, 2016, 05:46:42 PM »
In terms of the above, I'd have to disagree.  It impacts them at the same time sure.  But that isn't the same thing as at the same place in the graph.

Perhaps you don't understand what is being graphed.  The horizontal axis is payment date.  If something happens in the economy that affects a lot of borrowers, all at the same time, then yes, this appears all at precisely the same place on the graph.

BTW, the fed publishes this very same graph for credit card delinquencies, consumer loan delinquencies, etc.  It isn't something somebody made up to hide something.  There is also value in comparing the delinquency curves for LC to the delinquency curves for consumer loans, credit cards, and other loan types, so you can see how they are same or different.

Don't worry, I'm not taking vintage curves away from you.

PhilGD

  • Full Member
  • ***
  • Posts: 150
    • View Profile
    • Email
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #10 on: May 26, 2016, 06:05:52 PM »
In terms of the above, I'd have to disagree.  It impacts them at the same time sure.  But that isn't the same thing as at the same place in the graph.

Perhaps you don't understand what is being graphed.  The horizontal axis is payment date.  If something happens in the economy that affects a lot of borrowers, all at the same time, then yes, this appears all at precisely the same place on the graph.

BTW, the fed publishes this very same graph for credit card delinquencies, consumer loan delinquencies, etc.  It isn't something somebody made up to hide something.  There is also value in comparing the delinquency curves for LC to the delinquency curves for consumer loans, credit cards, and other loan types, so you can see how they are same or different.

Don't worry, I'm not taking vintage curves away from you.

Fred I think you're missing rawraw's point about growth. Since LC has been growing exponentially, the average seasoning of its borrower population will be biased downward on a month to month basis. Due to the non linear shape of the delinquency curves this attenuation/bias in the average seasoning of the population will make it hard to draw conclusions from the graph you posted above.
« Last Edit: May 26, 2016, 06:13:38 PM by PhilGD »

Fred93

  • Hero Member
  • *****
  • Posts: 2221
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #11 on: May 26, 2016, 06:16:34 PM »
Fred I think you're missing rawraw's point about growth. Since LC has been growing exponentially, the weighted average seasoning of its borrower population will be biased downward on a month to month basis.

I do understand that point.  The more recent loans are weighted heavily in this simple average.

This doesn't change anything in my response to rawraw. 


Quote
Due to the shape of the delinquency curves this attenuation/bias in the weighted average seasoning of the population will make it hard to draw conclusions from the graph you posted above.

Actually, under exponential growth, the average seasoning does not change much.  It is always weighted heavily toward the most recent vintages.  Just as it was not hard to observe, in this presentation, the degradation in performance of the high risk loans over the past few months, it is not hard to observe that they improved during Q1. 

If you or rawraw don't like the weighting by volume, with its recent-vintage-bias, you are free to weight the data a different way.

No one simple chart is most excellent for all purposes. 

Meanwhile, the overall portfolio delinquency rate came down in Q1.  That's a fact.


P2PFact

  • Jr. Member
  • **
  • Posts: 91
    • View Profile
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #12 on: May 26, 2016, 08:24:13 PM »
No, this form of charting isn't useful at all.  It will hide risk very well and then blow up in your face.  It's the oldest trick in the book that growth hides problems.

In terms of the above, I'd have to disagree.  It impacts them at the same time sure.  But that isn't the same thing as at the same place in the graph.  At a firm with LC's growth profile, you will always need the vintage to see whats going on.  I can't think of any benefits of the above graph for fast growing firms.  But I look at these sort of graphs for a living, so if I'm missing something feel free to share.  In financials, high growth firms are always a challenge because they don't provide data sufficient for vintages, like LC.

Well said! Too bad people didn't get it.

mchu168

  • Full Member
  • ***
  • Posts: 147
  • “Integrity is the ability to stand by an idea.”
    • View Profile
    • Email
Re: LC delinquencies went DOWN in Feb & March 2016
« Reply #13 on: May 27, 2016, 01:26:19 PM »

Well said! Too bad people didn't get it.

We do get it. But watch, now with volumes declining if delinquencies continue to fall it will be because of negative growth.

People look for a reason to discount new data to match their own beliefs.  This is a bias shared by all.