Author Topic: folio ytm  (Read 8816 times)

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #15 on: February 29, 2016, 06:17:56 PM »
I have a unique and yes even perhaps bizarre way of computing ytm simply as an educational exercise (not a useful tool).
It only uses an on-line loan amortization schedule calculator such as:
http://www.bankrate.com/calculators/mortgages/amortization-calculator.aspx

There is no "closed form" equation that will yield ytm given the set of parameters. Otherwise why bother.
So, again for educational purposes, enter the Folio asking amount (in pennies) as the loan amount (pennies needed for precision like 2500 for $25).
Enter the remaining number of months as the term.
Make a guess at the interest rate xx.xx% (when you finish this will be ytm).
Hit compute and the monthly payment will be computed and displayed.
Refine your interest rate guess so that eventually the monthly payment computed equals the existing monthly payment for the original loan.
Finally when you guess the interest rate that makes the new payment the same as the original then that interest rate is ytm.

Again, for the education of a complete noob, I ran this procedure for two loans.
One was pretty much dead on the same as the Folio provided and the other significantly higher.
Given the educational nature of the exercise is my procedure correct (meaning one of the Folio displayed ytm's is off badly) or is this my approach conceptually a bag of rubbish? Naturally I'd be curious as to what is wrong with the procedure of course.

TIA.

 

Fred

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: folio ytm
« Reply #16 on: March 01, 2016, 12:32:03 AM »
http://www.bankrate.com/calculators/mortgages/amortization-calculator.aspx

The link you used implies equally-spaced payments (1-month apart).  When you purchase notes from Folio, this is not always the case.

If the note's payment is due , say, on 15th of the month, the YTM would vary if your purchased on the 14th (1 day prior to payment) vs. on the 16th (29 days before next payment).

I think RaymondG's XIRR Excel from previous page would give more accurate YTM.  It gives you precise control on the time and amount of the payment stream.

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #17 on: March 01, 2016, 10:22:16 AM »
The link you used implies equally-spaced payments (1-month apart).  When you purchase notes from Folio, this is not always the case.

If the note's payment is due , say, on 15th of the month, the YTM would vary if your purchased on the 14th (1 day prior to payment) vs. on the 16th (29 days before next payment).

I think RaymondG's XIRR Excel from previous page would give more accurate YTM.  It gives you precise control on the time and amount of the payment stream.

Yes, I see. Time value of money. If you purchase 1 day before payment you get the entire month's interest but have not had the principal invested the entire month. XIRR allows one to account for just such examples. Guess it is up to the seller to incorporate accrued interest into the ask price every day to take this into account. I'll play around with RaymonG's XIRR method to get a feel for the process. Thanks to both!

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #18 on: March 02, 2016, 04:59:01 PM »
I think RaymondG's XIRR Excel from previous page would give more accurate YTM.  It gives you precise control on the time and amount of the payment stream.

Well, I played around with RaymondG's example and came up with possibly a third different YTM so here goes:



Rows 2 through 8 are the givens. It doesn't matter if I buy the note or not the number of remaining payments, the monthly payment amount and the remaining payment due dates stay the same. So, in XIRR terms, the cash flow (payments net of fees) for sale is set (well actually very dependent only upon the borrower's future actions but that's a different problem).

What is the YTM (XIRR) of this cash flow if I buy it on 2/3/2016 and pay $66.00 for it? The Excel equation and result (9.95%) are on line 10.
It only differs from RaymondG's 8.40% result in that in his computation the last payment was calculated to be less than the others.
If that's true I clearly don't understand it and need a little help.

As a sanity check I computed in row 14 the interest rate as if it were a new loan with principal equal to the price I paid and 23 monthly payments where the payment amount is exactly the same as the original loan. Figured this should be close to the XIRR but a bit higher rate since this does not penalize me like XIRR does for the time value of money from 2/3/2016 and 2/12/2016. Line 13 annualizes the row 14 interest rate and is 9.98% and is close to the calculated XIRR. Seems to check out okay.

So, how bad did I mess this up?  :o
« Last Edit: March 02, 2016, 05:00:37 PM by Rob L »

RaymondG

  • Full Member
  • ***
  • Posts: 247
    • View Profile
    • Email
Re: folio ytm
« Reply #19 on: March 02, 2016, 10:10:26 PM »
It only differs from RaymondG's 8.40% result in that in his computation the last payment was calculated to be less than the others.
If that's true I clearly don't understand it and need a little help.

Please see notes in the attached files
« Last Edit: March 02, 2016, 10:18:48 PM by RaymondG »

Fred93

  • Hero Member
  • *****
  • Posts: 2136
    • View Profile
Re: folio ytm
« Reply #20 on: March 02, 2016, 11:44:09 PM »
I think RaymondG's XIRR Excel from previous page would give more accurate YTM.  It gives you precise control on the time and amount of the payment stream.

Well, I played around with RaymondG's example and came up with possibly a third different YTM so here goes:

Nope.  You did this funny thing where you computed a "rate" from the payment, and then attempt to apply this interest rate.  This calculation is wrong.  The interest rate was set when the loan was made.  You can't just compute a different interest rate later.  The borrower would certainly object!

My calculation matches the RaymondG calculation.  I get 8.40%

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #21 on: March 03, 2016, 09:42:32 AM »
If that's true I clearly don't understand it and need a little help.
Please see notes in the attached files

Thanks! I can now see the error of my ways. Couldn't figure out where your $65.98 came from. My mistake was to "assume" the note payments has been made according to the original loan amortization schedule. Looking back at the Folio listing I see that the borrower doubled up on the first payment. The principal amount owed on 1/12/16 was not $66.84 as it would have been if the borrower had simply paid per schedule on time, but the $65.98 you used (as shown in the Folio payment history table). Now I have at least a shot at getting the correct answer.

Nope.  You did this funny thing where you computed a "rate" from the payment, and then attempt to apply this interest rate.  This calculation is wrong.  The interest rate was set when the loan was made.  You can't just compute a different interest rate later.  The borrower would certainly object!

My calculation matches the RaymondG calculation.  I get 8.40%

The funny thing (lines 13, 14 and 15) played no part in my XIRR calculation. I did not mean to imply that the initial loan parameters were in any way magically changed. Just added it as a reasonableness test of the XIRR computation. Playing around to get a better understanding of what to expect. Of course if you start out with bad initial conditions, as I did, then the wrong answer will appear to be reasonable.

Good to know the correct answer is 8.40% so I can fix my error. Lesson learned.

Thanks to RaymondG and Fred93.

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #22 on: March 03, 2016, 12:36:54 PM »
       LoanId      Par   AskPrice   AccInt Class Mat   N   Rate   Bench      IRR   YTM    Spread   Date           Pmt
581   38528044   100   66.00   0.37       6   36   23   8.67      8.39    9.42   7.72     1.03   2016-01-27   3.16

https://www.lendingclub.com/foliofn/loanPerf.action?loan_id=38528044&order_id=52420245&note_id=67624504


Seems ironic that this "example" loan was pre-paid in full on 2/25/2016 and with a hefty service fee of $0.67.
The issue date was 1/14/2015 so it was only a few weeks beyond the 12 months LC limits its fees for loans paid off early.
If I computed XIRR correctly on the completed loan  :-\ it was 7.95%. Without the service fee it would have been 8.70%.
These numbers presume the loan was not sold on Folio but continued to be held by its original owner.

wlbsr

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: folio ytm
« Reply #23 on: March 03, 2016, 02:41:52 PM »
wow.  how did you catch that?

i started buying notes on folio in January,  have purchased 42 so far and 2 of the 42 have been fully repaid.  don't think this was one of them though.

Rob L

  • Hero Member
  • *****
  • Posts: 2039
    • View Profile
Re: folio ytm
« Reply #24 on: March 03, 2016, 02:58:46 PM »
wow.  how did you catch that?

Quite by accident. I was just using the loan you started the thread with to learn how to calculate ytm.
Had to go back and look at the Folio listing to figure things out and there it was; all paid off.
Glad you didn't get left holding the bag with the quick pay off.

wlbsr

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: folio ytm
« Reply #25 on: March 03, 2016, 03:21:18 PM »
I have had 2 payoffs out of 42 loans in my first 2 months. whats the downside of an early payoff?



I wonder if this particular note being paid off was a coincidence.  My original flag was based on an irregular ytm because of persistent payments at a level above the original amortized amount. 

Fred93

  • Hero Member
  • *****
  • Posts: 2136
    • View Profile
Re: folio ytm
« Reply #26 on: March 03, 2016, 04:27:00 PM »
I have had 2 payoffs out of 42 loans in my first 2 months. whats the downside of an early payoff?

The downside of early payoff is that you have to invest again, and this reinvestment involves a DELAY.  The money doesn't earn anything during this delay.  There are several components of this delay, and most of them are out of your control.

Fred

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: folio ytm
« Reply #27 on: March 04, 2016, 12:27:39 AM »
I have had 2 payoffs out of 42 loans in my first 2 months. whats the downside of an early payoff?

If you purchased a non-Current note (i.e., In Grace, Late16, Late31), early payoff is very desirable.  The earlier, the better!

Fred93

  • Hero Member
  • *****
  • Posts: 2136
    • View Profile
Re: folio ytm
« Reply #28 on: March 04, 2016, 02:04:18 AM »
I have had 2 payoffs out of 42 loans in my first 2 months. whats the downside of an early payoff?

The downside of early payoff is that you have to invest again, and this reinvestment involves a DELAY.  The money doesn't earn anything during this delay.  There are several components of this delay, and most of them are out of your control.

After writing this I realized I was thinking like an investor who purchases loans at the beginning. 

If you buy on Folio, sometimes you buy below par.  If you buy below par, and the payoff comes early, you win a higher return.  (In this case yield-to-early-payoff is higher than yield-to-maturity.)  Conversely, if you buy above par, then early payoff produces a lower yield.

In all cases, reinvestment involves delay, which reduces your portfolio yield, because money is idle for a time while being recycled.

wlbsr

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: folio ytm
« Reply #29 on: March 04, 2016, 03:54:56 AM »
on folio, the reinvestment lag is not a big deal, probably worth the prepay bump if you are buying at a discount.

Any thoughts on maturity while we are at it?  would a 36 mo loan with 24 months left be more desirable than a a new 36 mo loan at an equal ytm assuming no credit status degradation in the interim on the older note?



How about a 60 mo loan with 36 months left vs.a new 36 mo loan with all other credit history/info at ytms assumed to be equal.