Author Topic: Cumulative ROI by Vintage Beginning 14Q1  (Read 3750 times)

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Cumulative ROI by Vintage Beginning 14Q1
« on: January 06, 2018, 04:53:02 PM »
When I posted these tables last quarter I trampled on another thread so I figured it would be best to start a new thread on the topic.
The tables are another way to look at the Cumulative ROI numbers provided in graph form by Insikt's excellent web site.
I think the data speaks for itself.
It may be that 16Q2 was the worst vintage, but I've not seen anything to indicate a major turnaround for lenders is underway.










MarinBB

  • Jr. Member
  • **
  • Posts: 83
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #1 on: January 06, 2018, 07:24:35 PM »
When I posted these tables last quarter I trampled on another thread so I figured it would be best to start a new thread on the topic.
The tables are another way to look at the Cumulative ROI numbers provided in graph form by Insikt's excellent web site.
I think the data speaks for itself.
These are beautiful, thanks for doing this. Do you have a summary tab for all LC loans that is not sliced by Grade?

sensij

  • Newbie
  • *
  • Posts: 30
    • View Profile
    • Email
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #2 on: January 07, 2018, 02:13:22 AM »
These are beautiful, thanks for doing this. Do you have a summary tab for all LC loans that is not sliced by Grade?

I think this covers your request (36 mo loans only, all grades)



Also, for those interested, here is Insikt's definition for this metric:

Quote
For any given period, this return metric equals the monthly interest and borrower late fees you have received on your loan investment, net of servicing fees and principal losses (all annualized) divided by the average loan balance outstanding during that period.

« Last Edit: January 07, 2018, 02:18:01 AM by sensij »

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #3 on: January 07, 2018, 09:34:37 AM »
These are beautiful, thanks for doing this. Do you have a summary tab for all LC loans that is not sliced by Grade?

I think this covers your request (36 mo loans only, all grades)


Very nice addition, thanks!

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #4 on: January 07, 2018, 10:53:01 AM »
Thought it would be interesting to compare the cumulative ROI of C, D and E grade loans to B grade loans. In the tables below the ROI of B grade loans is subtracted from C, D and E grade loans respectively. Since C, D and E are riskier by definition their ROI should include a risk premium to compensate the lender for owning them. As shown below this has been far from the case. Even C grade loans have little or no risk premium. Best case there has been very little or no risk premium but for the majority of vintages there are very substantial risk discounts! Of course this isn't new news, but it's interesting to see the magnitude. To me it's a persuasive argument to only invest in B loans if one invests at all (or maybe A grade as I haven't looked at them).



SLCPaladin

  • Full Member
  • ***
  • Posts: 202
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #5 on: January 07, 2018, 11:03:54 AM »
This is excellent insight Rob. If I read these tables correctly, would I be correct in assuming that loans that were originated in Q3 of 2017 would be "expected" to perform similar loans originated in Q4 of 2014? I know major economic shocks and external events would change the trajectory and of course nothing is guaranteed, but it seems like if early indications of the trajectory look like the loan quality has improved relative to what was originated during 2015 and 2016. Is that a fair conclusion?

I am specifically thinking about reinvesting in B grade notes of 36 month duration. I have run down about $120k to about $40k now.

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #6 on: January 07, 2018, 12:36:02 PM »
This is excellent insight Rob. If I read these tables correctly, would I be correct in assuming that loans that were originated in Q3 of 2017 would be "expected" to perform similar loans originated in Q4 of 2014? I know major economic shocks and external events would change the trajectory and of course nothing is guaranteed, but it seems like if early indications of the trajectory look like the loan quality has improved relative to what was originated during 2015 and 2016. Is that a fair conclusion?

I am specifically thinking about reinvesting in B grade notes of 36 month duration. I have run down about $120k to about $40k now.

If I had to make a WAG I'd guess we never get back to 14Q4, but land somewhere between 15Q1 and 15Q4 (say 5.8% to 4.8% ROI on B grade loans). LC has competition now. Back in 14Q4 it was either LC or Prosper. Times are so different and banks are lending again. Someday most banks will probably have their own version of Marcus. It appears clear LC lowered standards way too far in the 16Q1 - 16Q4 vintages, so they probably have since needed to swing the pendulum back to keep and attract lenders. Of course they can only do this if competition permits. And yeah if major shocks come along we don't know how bad it could get. This analysis and a buck won't get you a coffee so take it for what it's worth.

edit
Thought I'd add one more comment. Lately I've been looking at the A,B and C loans being offered at feeding times and IMO there are some really really lousy ones in all the grades. You've seen some mentioned in the "Worst Loan of the Day" thread https://forum.lendacademy.com/index.php/topic,4767.0.html. That said, I think now, more than ever, loan selection by retail lenders (either by 3rd parties or do-it-yourself) can be of substantial benefit. Being extremely picky will pay off. Again I must say this is my unproven speculation at best, so FWIW ...
« Last Edit: January 07, 2018, 03:06:32 PM by Rob L »

rawraw

  • Hero Member
  • *****
  • Posts: 2759
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #7 on: January 07, 2018, 03:21:19 PM »
Best case there has been very little or no risk premium but for the majority of vintages there are very substantial risk discounts! Of course this isn't new news, but it's interesting to see the magnitude. To me it's a persuasive argument to only invest in B loans if one invests at all (or maybe A grade as I haven't looked at them).


Thanks for the great analysis.  Makes me very happy.

But I disagree with the above conclusion.  You've included 4 years below, two or so of which contains a low point in a cycle.  If you think consumer cycle's occur over 4 years in a similar fashion, then yes I think the above statement is correct.  But without expliciting addressing this assumption, I don't think you can conclude that B is the best bet.  If we took the same analysis for E grade loans several years ago when this forum was complaining there wasn't enough high interest rate notes available (oh my how things have changed!), then it would have looked stupid to ever invest in B grade notes (I'm guessing).  You have to judge profitability of loans "through a cycle."  But you have the beginning blocks to do such an analysis.

rawraw

  • Hero Member
  • *****
  • Posts: 2759
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #8 on: January 07, 2018, 03:23:12 PM »
Best case there has been very little or no risk premium but for the majority of vintages there are very substantial risk discounts! Of course this isn't new news, but it's interesting to see the magnitude. To me it's a persuasive argument to only invest in B loans if one invests at all (or maybe A grade as I haven't looked at them).


Thanks for the great analysis.  Makes me very happy.

But I disagree with the above conclusion.  You've included 4 years below, two or so of which contains a low point in a cycle.  If you think consumer cycle's occur over 4 years in a similar fashion, then yes I think the above statement is correct.  But without expliciting addressing this assumption, I don't think you can conclude that B is the best bet.  If we took the same analysis for E grade loans several years ago when this forum was complaining there wasn't enough high interest rate notes available (oh my how things have changed!), then it would have looked stupid to ever invest in B grade notes (I'm guessing).  You have to judge profitability of loans "through a cycle."  This means including the full range of a cycle, not just sections of it.  But you have the beginning blocks to do such an analysis.

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #9 on: January 07, 2018, 05:03:36 PM »
Best case there has been very little or no risk premium but for the majority of vintages there are very substantial risk discounts! Of course this isn't new news, but it's interesting to see the magnitude. To me it's a persuasive argument to only invest in B loans if one invests at all (or maybe A grade as I haven't looked at them).


Thanks for the great analysis.  Makes me very happy.

But I disagree with the above conclusion.  You've included 4 years below, two or so of which contains a low point in a cycle.  If you think consumer cycle's occur over 4 years in a similar fashion, then yes I think the above statement is correct.  But without expliciting addressing this assumption, I don't think you can conclude that B is the best bet.  If we took the same analysis for E grade loans several years ago when this forum was complaining there wasn't enough high interest rate notes available (oh my how things have changed!), then it would have looked stupid to ever invest in B grade notes (I'm guessing).  You have to judge profitability of loans "through a cycle."  But you have the beginning blocks to do such an analysis.

I appreciate the complement and agree with your loan cycle criticism. It makes a very important point. Look at what's been going on with ROI in the various LC grades over just over the past four years. Huge drops. No way has the loan cycle (i.e. consumer repayment behavior) been the dominant factor. I would even argue (speculatively) that the loan cycle has played a rather minor role. Instead my theory FWIW is that LC's management of its underwriting criteria has had a much larger impact over this period. They have a balancing act to perform between offering lower interest rates (to maximize originations and their profits) versus higher interest rates (to keep and attract investors). Not only did LC overtly lower interest rates 14Q1 through 16Q1 but I suspect it simultaneously changed its lending criteria to offer even lower rates to more risky borrowers. They made a judgement that the borrower / lender balance was skewed too far towards the lender. As you said, lenders were knocking down their doors to give them money. It takes quite a bit of time (maybe a year) for the impact of these changes on ROI to become apparent (to lenders and probably LC itself). Sometime around 16Q2 LC not only had its scandal but realized it had taken things way too far in the borrower / lender balance. It may be that by that time the lending cycle contributed to this imbalance as well. Since then LC has increased interest rates a bit but we don't have enough data to know what if any changes they have made to tighten their underwriting. To further complicate life for them they now have serious competition and there will be more rather than less of that in the future.

Somewhere in the past I wrote that LC would always have borrowers but not necessarily lenders. The year or more time lag between lending and results requires lenders to bestow a certain degree of confidence in the underwriting of their intermediary. I don't have that any more. However if LC continues to provide transparent data it seems to me that, in theory, lenders may be able to cherry pick loans that meet their own criteria (if any) and have only themselves to blame if it doesn't work out.

Fred93

  • Hero Member
  • *****
  • Posts: 2136
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #10 on: January 07, 2018, 05:39:08 PM »
This is excellent insight Rob. If I read these tables correctly, would I be correct in assuming that loans that were originated in Q3 of 2017 would be "expected" to perform similar loans originated in Q4 of 2014?

Depends on how you develop your expectations.  Every vintage does not have the same shape vs time.  You're extrapolating from a very early point in the curve.  Just a few months of behavior. 

So, the answer is "maybe".

Lovinglifestyle

  • Hero Member
  • *****
  • Posts: 898
    • View Profile
    • Email
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #11 on: January 07, 2018, 07:31:20 PM »
My confidence in LC's underwriting hit a new bottom when all the loans started to come out as "approved", but didn't look any better to me.  Seems like all they had to do to get approved was fill out an application and have sufficient FICO.

SLCPaladin

  • Full Member
  • ***
  • Posts: 202
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #12 on: January 07, 2018, 10:43:24 PM »
I was playing around with the charts on Insikt to see the different vintage curves. I agree with your point Fred, it does seem way too early to draw any meaningful conclusions about the shape of the charge offs for these early vintages. I'm comparing cumulative ROI curves of various vintages and net charge offs. The two correlate, but not exactly. As best I can tell, the reason these most recent ROI curves look better is because the rates are better, not because there are fewer charge offs. What I suppose we don't know is whether the rate increases that happened post scandal will offset the higher delinquencies.

Lovinglifestyle

  • Hero Member
  • *****
  • Posts: 898
    • View Profile
    • Email
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #13 on: January 08, 2018, 11:34:57 AM »
These are beautiful, thanks for doing this. Do you have a summary tab for all LC loans that is not sliced by Grade?

I think this covers your request (36 mo loans only, all grades)



Also, for those interested, here is Insikt's definition for this metric:

Quote
For any given period, this return metric equals the monthly interest and borrower late fees you have received on your loan investment, net of servicing fees and principal losses (all annualized) divided by the average loan balance outstanding during that period.

I need a little help understanding what I think I'm seeing.  Does the bolded quote apply to this chart?  If so, the chart makes investing in LC look great to me.  There are no minuses at the bottom.  So the outcome of all grades combined is positive, regardless of individual risk?

arcee49

  • Newbie
  • *
  • Posts: 30
    • View Profile
Re: Cumulative ROI by Vintage Beginning 14Q1
« Reply #14 on: January 08, 2018, 01:37:37 PM »
My confidence in LC's underwriting hit a new bottom when all the loans started to come out as "approved", but didn't look any better to me.  Seems like all they had to do to get approved was fill out an application and have sufficient FICO.

IIRC the reason all loans started coming out as "Approved" was a change LC made to keep people from investing in a loan that later got rejected and thereby not tying up investor's money in a loan that would never be issued.