Author Topic: a view of LC's deteriorating investor returns  (Read 29280 times)


AnilG

  • Hero Member
  • *****
  • Posts: 1122
    • View Profile
    • PeerCube
Re: a view of LC's deteriorating investor returns
« Reply #76 on: March 21, 2017, 08:16:06 PM »
I don't make any assertion of which side of credit cycle we are in, only that we most likely have passed the strongest performance vintages for P2P loans during this economic cycle. Anyone else's guess is as good as mine about where we are and what to expect.

I am glad to see that you brought up prepayment rate. It is often an overlooked statistics compared to default, loss and delinquency rate. When prepayment rate goes up, it indicates that we are in credit expansion phase, i.e. the new credit is available to borrowers for refinancing existing debt most probably at better terms. Once prepayment rate stagnates or decline, credit expansion may have stopped. In the table below, that vintage seems to be 2014 and later. At that point onward delinquency, default and loss rate should start to go up. If you quint at the table, you might see somewhat inverse relationship between prepayment rate and delinquency/loss rate.

I don't believe this trend is unique to Lending Club considering most high yield unsecured consumer lending platforms have come out with bad news in last year or so - Lending Club, Prosper, Avant, Marlette, SoFi, Upstart just a few I remember. Similar trends are also being seen in collateralized market, specifically auto, motorcycle, boat type small amount high yield loans.

36 month loans at 12 month
VintagePrepayment RateDelinquency RateLoss Rate
20166.38%0.00%0.66%
201512.18%2.01%2.31%
201412.62%1.77%1.99%
201310.85%1.61%1.87%
20129.57%2.09%2.44%
201110.40%1.57%1.70%
201010.35%2.62%2.37%
20099.78%3.82%3.89%
20087.35%6.79%2.68%

Anil, it sounds like you believe we are on the weak side of a credit cycle. How would this (or how is it presenting itself)? You mentioned competition and loosening. It seems that LC is affected by both. When I look at performance on existing loans, I see that ALL of them are taking a hit. Even 60 month loans in the 2014 vintage have come way off in performance. It is not just the 2015 and 2016 vintages that everybody complains about. That seems to tell me that there is a lot of competition for LC customers these days. Maybe LC competitors have loosened up to the point that LC customers pile on more debt until it becomes default time. We hear fears that borrowers are stacking. We also hear from Lending Club that they see issues with people that have a penchant for taking on too much debt. We also see prepayments rising. In my mind, prepayments are a relatively small hit. These two things are just indicators of the real problem. The real hits seems to be additional borrowing by people that I have already loaned money to.
---
Anil Gupta
PeerCube Thoughts blog https://www.peercube.com/blog
PeerCube https://www.peercube.com

AnilG

  • Hero Member
  • *****
  • Posts: 1122
    • View Profile
    • PeerCube
Re: a view of LC's deteriorating investor returns
« Reply #77 on: March 22, 2017, 01:17:50 AM »
I came across this 30 minutes YouTube video on Reddit. I thought it is relevant to discussion in this thread. Check it out to understand short and long term debt cycles.

https://youtu.be/PHe0bXAIuk0
How The Economic Machine Works by Ray Dalio
---
Anil Gupta
PeerCube Thoughts blog https://www.peercube.com/blog
PeerCube https://www.peercube.com

rawraw

  • Hero Member
  • *****
  • Posts: 2794
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #78 on: March 22, 2017, 07:47:15 PM »
I came across this 30 minutes YouTube video on Reddit. I thought it is relevant to discussion in this thread. Check it out to understand short and long term debt cycles.

https://youtu.be/PHe0bXAIuk0
How The Economic Machine Works by Ray Dalio
Ha that has to be the most famous video in finance. If you don't know anything about Bridgewater, read a little about it.  What a strange place

SLCPaladin

  • Full Member
  • ***
  • Posts: 202
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #79 on: March 22, 2017, 11:36:15 PM »
I came across this 30 minutes YouTube video on Reddit. I thought it is relevant to discussion in this thread. Check it out to understand short and long term debt cycles.

https://youtu.be/PHe0bXAIuk0
How The Economic Machine Works by Ray Dalio
Ha that has to be the most famous video in finance. If you don't know anything about Bridgewater, read a little about it.  What a strange place

I always enjoy Matt Levine's cheeky comments on Bridgewater and their "radically transparent" culture:

Bridgewater's founder Ray Dalio spoke at a New York Times conference, criticized the story, and explained that Bridgewater's culture really isn't that weird when you think about it:

Speaking at a conference sponsored by The New York Times in Half Moon Bay, Calif., the 67-year-old billionaire likened that culture — which he calls “radical transparency” — to going to a nudist camp for the first time.

“You first walk into a nudist camp and it’s very awkward,” he said at the New Work Summit conference. For people to get beyond that, he said, “there has to be getting over the emotional reaction.”

Levine: Yes wait what? Bridgewater is not that unusual, I suppose, if your context is that it's like a nudist camp; but that's an unusual context to have. Most people are not going to work at hedge funds because they want to expose their most private parts to strangers' critical judgment. But that is exactly what Bridgewater wants. The difficulty is that Bridgewater is run by people who are very smart and rigorously logical, so everything that they do and say makes perfect sense in their context. It's just that they start from a slightly different place from the rest of us.

Fred93

  • Hero Member
  • *****
  • Posts: 2243
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #80 on: April 27, 2017, 06:50:48 AM »
Well, here's another way to look at the situation.  We usually look at returns.  Many of us calculate returns different ways, but all of them involve interest rates, chargeoffs, fees, and sometimes other factors.  Tonite I'm going to look at just one factor.  Its the factor that has the little problem ... chargeoffs.

I've calculated a chargeoff rate for two portfolios.  These happen to be two portfolios that I have access to.  The first is my own Fred93 portfolio.  The second is Lending Club's Broad-Based Consumer Credit Fund (which I refer to as BBQ). 

For my own account, each month I take the net chargeoff $ (which means chargeoffs minus recoveries) and divide by the account balance at the beginning of the month.  This produces a monthly chargeoff rate.  I then multiply it by 12 to get an annual rate.  Nothing fancy.  No freaky math like IRR or ANAR.  Everybody can follow along.

Since the beginning of 2016, the BBQ fund has published their monthly net chargeoff rate.  I've multipled this by 12 and plotted it on the same chart as the Fred93 chargeoff rate.



I'm very happy that my chargeoff rate is lower than LC's, but that's not the main point of this note. 

You can see that both of these curves used to run along sorta flat, and then in mid 2016, they both started climbing.  That's the cause of the degradation of loan performance that I've often talked about. 

The first thing I notice is that the net chargeoff rate for both these portfolios are moving up together.  That gives me some confidence that it isn't my bad loan selection that made the bend in the red curve.  I don't have the data for all of your portfolios, but I suspect that if you plotted net chargeoff rate vs time, you'd get a similar looking curve, with a bend up.

You can see that neither of these curves looks like it is leveling off.  I hope they're leveling off!  I suspect that they're leveling off.  But... there just isn't data yet to demonstrate that.  So far its a wish and a supposition.

One nice thing about this annualized net chargeoff rate is that you can compare it directly to the average interest rate of your portfolio.  They have the same denominator.  Each month interest comes in, and chargeoffs go out.  You can subtract one from the other.  Lets try that...

Lets do it for the LC fund.   LC's BBQ fund has an average interest rate of 12.96%, and in March had a net chargeoff rate of -13.68%.  Oops.  They're losing money.  They also charged a fee of -0.09%/month, which annualizes to -1.08%.  (The fund charges different size customers different sized fees, and this is their published number for the fund average.)  The result was that they lost money on a cash basis.  Annualized cash basis performance in March was 12.96% - 13.68% -1.08% = -1.8% 

I say "cash basis", because when LC calculates returns on their fund, they throw in an additional term, which they call "adjustment".  In accounting lingo this adjustment is a "non cash item".  The adjustment marks the portfolio down when market interest rates go up and vice versa, when credit quality goes down, and vice versa, and they also adjust for loan age.  They have this scheme where loans are given a haircut at birth because some of them are expected to default later in life.  As life goes on, there is less life left in which to default, so the value goes up.  This may be a perfectly rational system of accounting for some, but it isn't the way most of us value our own portfolios, so I've removed the adjustment, leaving the terms which actually measure movement of cash.  Their adjustment for March was +0.59% which annualizes to +7.08%, so when you add that in, that makes the return they report come out positive.

The adjustments are big, and kinda random looking.  Now let me be clear:  I'm not accusing them of anything.  These adjustments are calculated by an outside firm so that we won't have to worry about insiders making up positive adjustments to hide bad stuff.  I do think that outside firm might employ too many guys with green eyeshades and not enough people with common sense, but that's just my opinion.  I should also mention that the adjustments are negative in some months and positive in others.  I just prefer to ignore the adjustment.  Its hilarious... I charted the BBQ fund returns and the curve jumps up and down like crazy.  Looks ridiculously noisy.  I took out all their adjustments, and got a nice smooth curve! 

The BBQ fund is losing money on a cash basis.  We have every reason to believe that it will continue to do so.  Unless some magical thing makes the net chargeoff rate go down, it will continue to exceed the interest rate, and the fund will lose money.  ...no matter what the adjustments say.

But enough about the BBQ fund.  Back to the shape of the curves. 

Both of these portfolios have seen an increase in chargeoff of around +6% over the period we observe here.  That's huuuge! 

You hear LC from time to time say they've increased interest rates so everything is fine, but the increases in interest rates have been very small compared to the increases in chargeoffs.  I haven't done the math, but you know LC took interest rates down and then up again, and the net change isn't much.  Yea, I know they made some larger increases in the high risk grades, but both these portfolios are similarly middle-grade, centered on "C" grade. 

I've charted vintage chargeoff rates various ways in the past.  These show that recent vintages are performing worse than older vintages, which is nice to know, but doesn't relate directly to cash going in and out of our accounts.   Many of us have published  charts of the ratio of monthly chargeoffs to interest received.  That does something similar to what I've done here, but it's a ratio, which doesn't so easily combine with interest rates.  If you compute the chargeoff rate, you can directly subtract it from your portfolio average interest rate, which seems a little more intuitive. 

Portfolio chargeoff rate is my new favorite thing to track.

ThinleyWangchuk

  • Full Member
  • ***
  • Posts: 106
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #81 on: April 27, 2017, 12:03:22 PM »
Well, here's another way to look at the situation.  We usually look at returns.  Many of us calculate returns different ways, but all of them involve interest rates, chargeoffs, fees, and sometimes other factors.  Tonite I'm going to look at just one factor.  Its the factor that has the little problem ... chargeoffs.

I've calculated a chargeoff rate for two portfolios.  These happen to be two portfolios that I have access to.  The first is my own Fred93 portfolio.  The second is Lending Club's Broad-Based Consumer Credit Fund (which I refer to as BBQ). 

For my own account, each month I take the net chargeoff $ (which means chargeoffs minus recoveries) and divide by the account balance at the beginning of the month.  This produces a monthly chargeoff rate.  I then multiply it by 12 to get an annual rate.  Nothing fancy.  No freaky math like IRR or ANAR.  Everybody can follow along.

Since the beginning of 2016, the BBQ fund has published their monthly net chargeoff rate.  I've multipled this by 12 and plotted it on the same chart as the Fred93 chargeoff rate.



I'm very happy that my chargeoff rate is lower than LC's, but that's not the main point of this note. 

You can see that both of these curves used to run along sorta flat, and then in mid 2016, they both started climbing.  That's the cause of the degradation of loan performance that I've often talked about. 

The first thing I notice is that the net chargeoff rate for both these portfolios are moving up together.  That gives me some confidence that it isn't my bad loan selection that made the bend in the red curve.  I don't have the data for all of your portfolios, but I suspect that if you plotted net chargeoff rate vs time, you'd get a similar looking curve, with a bend up.

You can see that neither of these curves looks like it is leveling off.  I hope they're leveling off!  I suspect that they're leveling off.  But... there just isn't data yet to demonstrate that.  So far its a wish and a supposition.

One nice thing about this annualized net chargeoff rate is that you can compare it directly to the average interest rate of your portfolio.  They have the same denominator.  Each month interest comes in, and chargeoffs go out.  You can subtract one from the other.  Lets try that...

Lets do it for the LC fund.   LC's BBQ fund has an average interest rate of 12.96%, and in March had a net chargeoff rate of -13.68%.  Oops.  They're losing money.  They also charged a fee of -0.09%/month, which annualizes to -1.08%.  (The fund charges different size customers different sized fees, and this is their published number for the fund average.)  The result was that they lost money on a cash basis.  Annualized cash basis performance in March was 12.96% - 13.68% -1.08% = -1.8% 

I say "cash basis", because when LC calculates returns on their fund, they throw in an additional term, which they call "adjustment".  In accounting lingo this adjustment is a "non cash item".  The adjustment marks the portfolio down when market interest rates go up and vice versa, when credit quality goes down, and vice versa, and they also adjust for loan age.  They have this scheme where loans are given a haircut at birth because some of them are expected to default later in life.  As life goes on, there is less life left in which to default, so the value goes up.  This may be a perfectly rational system of accounting for some, but it isn't the way most of us value our own portfolios, so I've removed the adjustment, leaving the terms which actually measure movement of cash.  Their adjustment for March was +0.59% which annualizes to +7.08%, so when you add that in, that makes the return they report come out positive.

The adjustments are big, and kinda random looking.  Now let me be clear:  I'm not accusing them of anything.  These adjustments are calculated by an outside firm so that we won't have to worry about insiders making up positive adjustments to hide bad stuff.  I do think that outside firm might employ too many guys with green eyeshades and not enough people with common sense, but that's just my opinion.  I should also mention that the adjustments are negative in some months and positive in others.  I just prefer to ignore the adjustment.  Its hilarious... I charted the BBQ fund returns and the curve jumps up and down like crazy.  Looks ridiculously noisy.  I took out all their adjustments, and got a nice smooth curve! 

The BBQ fund is losing money on a cash basis.  We have every reason to believe that it will continue to do so.  Unless some magical thing makes the net chargeoff rate go down, it will continue to exceed the interest rate, and the fund will lose money.  ...no matter what the adjustments say.

But enough about the BBQ fund.  Back to the shape of the curves. 

Both of these portfolios have seen an increase in chargeoff of around +6% over the period we observe here.  That's huuuge! 

You hear LC from time to time say they've increased interest rates so everything is fine, but the increases in interest rates have been very small compared to the increases in chargeoffs.  I haven't done the math, but you know LC took interest rates down and then up again, and the net change isn't much.  Yea, I know they made some larger increases in the high risk grades, but both these portfolios are similarly middle-grade, centered on "C" grade. 

I've charted vintage chargeoff rates various ways in the past.  These show that recent vintages are performing worse than older vintages, which is nice to know, but doesn't relate directly to cash going in and out of our accounts.   Many of us have published  charts of the ratio of monthly chargeoffs to interest received.  That does something similar to what I've done here, but it's a ratio, which doesn't so easily combine with interest rates.  If you compute the chargeoff rate, you can directly subtract it from your portfolio average interest rate, which seems a little more intuitive. 

Portfolio chargeoff rate is my new favorite thing to track.

Great post! Please continue to keep us updated :)

rawraw

  • Hero Member
  • *****
  • Posts: 2794
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #82 on: April 27, 2017, 12:08:54 PM »
Thanks for sharing Fred! Just one quirk in your calculation, that probably isn't an issue but figured I would share. Your interest rate assumes that everyone is paying I think. But you actually aren't getting interest on the past due notes which aren't in your charge offs yet. This isn't a big deal for our loans since they charge off relatively quick, but for other loan types they can be nonaccrual for a very long time. So when banks have problem loans, their yields look low because they stay in the denominator but no income is in the numerator. Didn't know if you knew this quirk.

Sent from my SAMSUNG-SM-G935A using Tapatalk


investny

  • Jr. Member
  • **
  • Posts: 90
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #83 on: April 27, 2017, 01:17:38 PM »
Here is a simple summary of returns for beginning of 2016
http://screencloud.net/v/Egaqn

Also,
another way to look a defaults
http://screencloud.net/v/25biR

The around 7% return on C grade loans does not look bad to me.

mrwhizzard

  • Jr. Member
  • **
  • Posts: 55
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #84 on: April 27, 2017, 02:52:02 PM »
Portfolio chargeoff rate is my new favorite thing to track.

I liked this so much, that I made my own. I graphed my LC and Prosper accounts separately:


jheizer

  • Sr. Member
  • ****
  • Posts: 483
    • View Profile
    • LC Tools
Re: a view of LC's deteriorating investor returns
« Reply #85 on: April 27, 2017, 03:01:46 PM »
I'll join the party.  Mine excludes recoveries though as I don't have that in my spreadsheet currently.



Also my late 16 and late 31 categories both have less than half the number of notes as they did 6 months ago.  All my slopes are starting to improve.
« Last Edit: April 27, 2017, 03:03:33 PM by jheizer »
Replacement to P2P Quant's Percentile Tool http://lc.geekminute.com

rubicon

  • Full Member
  • ***
  • Posts: 103
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #86 on: April 27, 2017, 03:10:57 PM »
1. I understand the charge-off methodology but I don't really like it since if you hit a bump in your portfolio you only know 9 months later. Also you need to make sure you are comparing against the balance from 9 months ago when calculating the %. This is a cash method as it compares cash interest against actual charge-offs taken.

2. Instead I prefer to use an accrual method, using the monthly change in adjustments for late loans and adding back the cash interest paid as well change in accrued interest. This gives a more real-time read on the portfolio. You also don't have to mess about with the denominator.

Hence I saw that Jan was an absolutely awful month but Feb, Mar and April have improved. Possibly a seasonal impact (tax returns).

Fred93

  • Hero Member
  • *****
  • Posts: 2243
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #87 on: April 27, 2017, 03:13:14 PM »
Thanks for sharing Fred! Just one quirk in your calculation, that probably isn't an issue but figured I would share. Your interest rate assumes that everyone is paying I think. But you actually aren't getting interest on the past due notes which aren't in your charge offs yet. This isn't a big deal for our loans since they charge off relatively quick, but for other loan types ...t

Right.  The correct cash value for the calculation is "interest received" which is not quite the same as "interest due".

These numbers will be close, but not exactly same.

Fred93

  • Hero Member
  • *****
  • Posts: 2243
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #88 on: April 27, 2017, 07:18:19 PM »
1. I understand the charge-off methodology but I don't really like it since if you hit a bump in your portfolio you only know 9 months later.

Isn't it 4 months later? 


Quote
2. Instead I prefer to use an accrual method, using the monthly change in adjustments for late loans and adding back the cash interest paid as well change in accrued interest. This gives a more real-time read on the portfolio.

I have no problem with that sort of calculation.  I do it too.  Each has its own pros and cons.  The adjustment for late loans is a guess.  Historically, the relationship between lates and eventual chargeoff has fluctuated considerably.  Also some people (like me) like IRR, whereas others are put off by the math and prefer an ANAR-type calculation.  Others argue about details of ANAR.  Everybody does it differently. 

Sometimes tho clarity comes from backing up and looking earlier in the calculation, at the components in a more unprocessed form.

Quote
Hence I saw that Jan was an absolutely awful month but Feb, Mar and April have improved. Possibly a seasonal impact (tax returns).

I also track late ratio (Late / (late+current)) for a measure of portfolio condition.  For my Fred93 account, lates peaked in August 2016 and have been wiggling in a flat band Oct 2016 thru Mar 2017.  I can't see a trend.  For the LC BBQ fund, the trend in lates is up, with Mar 2017 the highest ever (just barely).  So I don't see improving.

Regardless of which measurement you prefer, do you agree about the large degradation in quality of LC loans over 2015/6/7 ? 

I think the portfolio chargeoff rate shows that very clearly.  Chargeoff rates have doubled or tripled for many portfolios.  This is a big deal.  The changes have been minimized in LC management's public statements, with phrases such as "pockets of underperformance", but hey, when something triples, that's the elephant in the room.

Fred93

  • Hero Member
  • *****
  • Posts: 2243
    • View Profile
Re: a view of LC's deteriorating investor returns
« Reply #89 on: April 27, 2017, 07:26:59 PM »
Here are LC management's public statements about loan quality during the past couple of years.  They consistently attempt to minimize the problem by describing changes as affecting only a small fraction of borrowers.  ... at the same time that the charegoff rate on their own fund tripled.

Quote
8K 4/20/2016
Grades A to C have continued to perform consistent with our expectations.  In some higher risk grades in early 2016, we identified some underperforming pockets of loans and made modifications to pricing and credit policies accordingly.  The population eliminated from the credit policy represents slightly les than 5% of loan volume (annualized based on Q4 2015), and was mainly characterized by high indebtedness, an increased propensity to accumulate debt and lower credit scores.

8K 6/7/2016
Effective June 7, 2016, debt to income (DT() criteria (excluding mortgage and the requested program loan amount) is being reduced to 35% (from 40%) across the standard loan program.  Lending Club expects the change to primarily impact grades E through G.  Lending Club expects standard program loan volume to be reduced by approximately 5% as a result of this change.

8K 10/14/2016
...the thresholds on borrower leverage were tightened on October 12,2016.  The platform will no longer approve loans for certain sub-sgements of borrowers who meet a combination of several risk factors such as high revolving debt, multiple recently opened installment loans, and higher risk scores on our proprietary scorecard.  Accordingly approximately 1% of borrowers who previously would have been able to obtain a loan under prior underwriting criteria will no longer be approved.

8K 1/18/2017
Effective January 11, 2017, changes were implemented to tighten credit criteria based on unique combinations of risk factors such as number of recent installment loans, revolving utilization, and higher risk scores on our proprietary scorecard.  Borrowers who meet these specific combinations of risk factors model lower risk-adjusted returns than their otherwise similar peers.  This group represents a small portion of the total borrower population but a notable portion of higher risk borrowers.

2016Q2 Earnings conference call 8/8/2016
Throughout late 2014 and 2015 we had an excess supply of capital, which allowed us to reduce platform interest rates and therefore returns. When the Fed raised rates in December of 2015 we took the decision to raise rates.  Since then rates were increased three times this year in order to increase the appeal of the asset class.

In total, rates rose by a weighted average 135 basis points since December bringing the weighted average platform rate on our standard program to just over 13%. This average rate remains a very attractive bond from secured consumer credit. 

On the credit side we reduced approval rates for certain targeted segments to eliminate roughly 9% of the higher risk personal loan population that have exhibited a propensity to accumulate debt and could have the most exposure to an economic slowdown.

Based on the above pricing and credit actions standard program returns are expected to increase from 4% to 5% to more than 6% for vintages after June.


2016Q3 Earnings conference call 11/7/2016
this year, we raised rates four times and tightened credit three times based on trends we were seeing in borrower behavior and investor expectations.

In particular, we stopped approving loans to a portion of borrowers that was exhibiting a high propensity to accumulate debt and could have the most exposure to an economic slowdown. The vintages that included these underperforming populations are still expected to deliver solid returns and we're targeting newly issued vintages as a result of the above-mentioned credit and pricing actions to deliver weighted average annual return of approximately 6%.

2016Q4 Earnings conference call 2/14/2017
Throughout 2016, we implemented policy changes involving increased prices and a limiting of certain borrower populations based on observed performance. While it is still early and the data is correspondingly thin, we are pleased to see increase gross yields and signs of stabilization in early delinquency rates through our latest data in January. 

Last month, we implemented an additional change that resulted in our seizing to offer credit to roughly 6% of our total borrower base. The change is disproportionately focused on our higher risk grades, similar to previous adjustments and targets the unique combination of risk factors. While this change will give us some short term pressure in Q1, we’re building a business for the long-term in a truly massive market and prudent credit management is our priority.

To me, management's repeated claims that changes will only affect a small fraction of the borrower base says that their primary concern is VOLUME, not quality.