BarryHWhite said:
is there a formula which assumes interest is paid out
and calculates the cash flows from settlement date forward
Use YIELD().
Although in bond parlance, there is usually a difference between the terms
(current) "yield" and "yield to maturity", apparently the Excel YIELD()
returns YTM, at least when the bond is held for more than one coupon period.
This is self-evident if you compare the results of YIELD() and IRR() or
RATE() in some simple cases, for example a bond held from 1/1/2009 through
1/1/2012 priced at $90 (per $100 face value) with semi-annual coupons at the
rate of 5%.
Caveat emptor: I cannot vouch for the results of YIELD() for bonds held for
one coupon period or less, or for bonds with maturity dates that are not
multple of coupon periods (!). Moreover, I have not tried to use YIELD()
for bonds that pay all the interest at maturity, although I suspect we can
make it return something reasonable by fudging the "redemption" parameter.
PS: I have also not tried to use YIELD() for bonds that are purchased after
the issue date -- ergo, we usually need to pay some accrued unpaid interest
for part of the coupon period that includes the settlement date. But
similarly, I suspect we can make it return something reasonable by fudging
the "pr" (price) parameter.
Like you, I have not had much success with YIELDMAT(), and it is unclear
what its algorithm is, notwithstanding a formula that some participants
posted in response to my inquiry in another thread.
I not yet had a chance to fully experiment with YIELD(). But it looks
promising since it returns results that are consistent with IRR() and RATE()
in circumstances that are conducive to using those functions; and those
results are very different from the YIELDMAT() results, which I believe are
incorrect in those circumstances.
FYI, it is not surprising for any valid YTM function or formula to return
results that differ from XIRR(), even when the date count mode ("basis") is
actual/actual.
The differences are due, at least in part, to apparently widely-accepted
conventions for determining YTM when the coupon frequency is more than once
per year. Also, there seems to be a wide variety of interpretations or
"rules of thumbs" that affect the details of the YTM formulation, especially
when fractional coupon periods are involved. Even the HP 12C formula is
questionable, IMHO.
But I digress. I tried to initiate an academic discussion of all this in my
thread ("Please explain YIELDMAT function"), but to no avail so far.
However, last weekend was a holiday in both the US and Canada, and some of
the participants who have demonstrated financial wizardry in the past might
still be on vacations or just returning from them.
In the meantime, I hope my comments help.
----- original message -----
BarryHWhite said:
I wonder if I have misunderstood what the formula does
I am assuming that interest on the bond is paid out annually, so if I buy
the bond on the anniversery of the issue date, interest on the period
before
my purchase is not
an issue
My first example involves cash flows as follows
-100 Sett Date (when I buy the bond) 31/12/2009
+110 Maturity (when the bond terminates and pays capital and one years
interest 31/12/2010
Clearly here the return is 10%
and the bond issue date is not relevant
The reason I showed issue date as 1 day before settlement date is the
formula will not accept the same issue date and settlement date
Is the formula assuming that no interest is paid out, but accumulates in
the
bond - if so, I see that this will give a different result depending on
issue
date
Can you tell me the underlying calculation for the YieldMat function?
If this is the case, is there a formula which assumes interest is paid out
and calculates the cash flows from settlement date forward - or should I
use
the formula with issue date one day before settlement, which (apart from
the
tiny interest difference) gives me an answer the same as I would get using
IRR with the cash flows.
Rgds
Sean Timmons said:
I must disagree.
the bond has achieved a portion of maturity prior to settlement. In the
below examples, 1 day and 1 year respectively. The clear indication is
that
those days you do not have the bond in "possession" do not accumulate
interest in your name. (Please note, in the first example above, the
return
is not precisely 100%, though the change occurs at a third decimal). Hope
thsi helps!
:
I think that the Date of Settlement was the date I purchased the bond.
Nothing to do with the original issue date, nor the maturity date
The parameters are
Syntax: YIELDMAT(S, M, I, R, PR[, B])
S = settlement date
M = maturity date
I = issue date
R = interest rate at date of issue
PR = the price of the security per $100 face value
B = (Optional) the day count basis to be used:
0 or omitted 30/360
So
My question remains
Example (dates UK format dd mm yy)
Yield to maturity 10.00%
Sett Date 31/12/2009
Maturity 31/12/2010
Issue Date 30/12/2009
Coupon yearly 10.000%
Price now 100.00
Days basis 3
But if the Issue date changes you get
Yield to maturity 9.09%
Sett Date 31/12/2009
Maturity 31/12/2010
Issue Date 31/12/2008
Coupon yearly 10.000%
Price now 100.00
Days basis 3
Which is wrong - you are paying out 100 now at 10% and will get 110 in
one
year, so the yield to maturity should be 10% - does not matter when the
bond
was issued
:
Based on MS Help, the rate and term are based on date of issue (not
date of
settlement). So, if the security was issues on 1/1/08 on a 30 year
bond, the
maturity date is 1/1/38, regardless of settlement date.
:
Why does Yieldmat include an issue date ?
The yield to maturity should be from the settlement date to the
maturity
date, subject to the rate of interest and the price of the bond
It does not matter when in the past it was issued
However I see that changing the issue date changes the yield to
maturity
Help please