| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 997.1 | | POWDML::VISCONTI | | Thu Apr 18 1996 08:42 | 6 |
| Thanks for the education, I've been ready Money, Kipplinger, etc. for
several years and just assumed (silly me) that cumulative return was
something different.
Regards,
Jim
|
| 997.2 | Thanks | GRANPA::BROWN | My kids call my father Granpa Brown | Thu Apr 18 1996 09:54 | 1 |
| >I appreciate that info as well.
|
| 997.3 | A rebuttal | NCMAIL::YANUSC | | Thu Apr 18 1996 10:07 | 45 |
| Jay,
Your presumption of how cumulative return works in a simplistic way is
correct; I won't argue that. What I will argue, though, is that in the
urge to present your argument, you have committed two errors:
1. You have ignored dividends and capital gains, which even during down
periods can be critical, particularly since you are buying shares in
the mutual fund at a lower cost during those down periods. Leaving
your initial investment at $100, and never adding in these dividends
and capital gains, gives an incorrect picture of the real world.
2. Your %s used for your exercise are wildly inflated. How many funds
would have any shareholders left if they were up 100% one year, down
70% over the next two, and up 20% in the fourth. Using proportionally
the same % you used, but with a potentially more real-world touch, I
would show the following:
Initial Investment Rate of Return Ultimate Amount
$100 +10% $110
---- - 5% $105
---- - 2% $102
---- + 2% $104
Not a great return, but again, it does not factor in capital gains and
dividends as I suggested above. It also shows a generally bad market
overall for the four years, so other investments would also likely be
affected adversely.
I have a variety of mutual funds, and while not a "fanatic" of funds, a
large portion of my investments are contained in them. Some of my
funds have never had another dime added to them after my initial
investments. A good example would be my IRA mutual funds, which I have
not added to since I lost the tax deductibility. I can assure you that
they are up substantially in bottom line value over the initial amount
I deposited (much like your example above), due to increasing share
price (% return) and the addition of dividends and capital gains.
Ignore mutual funds and you are relegated to picking stocks (not an
easy endeavor for most, making them a losing proposition for most over
the long haul) or using only safe investments like Treasuries, CDs and
the like.
Chuck
|
| 997.4 | | MROA::YANNEKIS | | Thu Apr 18 1996 10:13 | 28 |
|
> This is exactly the way all fund managers report as their 2 year, 3 year,
> 5 year and 10 year returns of their funds and I believe that the are legally
> allowed to do so. In this example I have simplified the numbers for
> clarity ! But you get the picture.
Color me a sceptic. How do you know this is how they report the
returns? I have never read the details but everything I have ever
read from Fidelity implies they take the starting and ending balances
and compute the compounded return those starting and ending points
imply (BTW including the compounding which makes them look worse).
One example, you have +25% after 2 years after the fund went $100 >=>
$200 >=> $100. I've seen Fidelity finds with about the same starting
and closing balances with returns of zero. That looks more like what I
believe happens than what you say.
I find it hard to believe 1) it would be legal to do something so
deceptive and counter to common formulas financial analysis of
returns and 2) (and more importantly) that mutual fund companies would
face the rath from consumers when they were exposed for using such
deceptive practices.
Siting your sources would help this skeptic a lot.
Greg
|
| 997.5 | Not quite correct | SLOAN::HOM | | Thu Apr 18 1996 10:46 | 38 |
| Re: 0
> This is exactly the way all fund managers report as their 2 year, 3 year,
> 5 year and 10 year returns of their funds and I believe that the are legally
> allowed to do so.
The above statement is er... not quite correct.
The Vanguard family states true cumulative
1, 5, 10 year returns. (Assuming dividend reinvestments.)
Here are the 5 year return numbers:
Annual Total
Vanguard SP500 Return Return
1995 37.40% 2.131
1994 1.10% 1.551
1993 9.80% 1.534
1992 7.40% 1.397
1991 30.10% 1.301
Simple Sum 85.80%
5 Yr Average 17.16%
Cumulative Total return 113.1%
Average Return
as stated by Vanguard 16.4%
True annual return with
compounding 16.4% 1.164 ^ 5 = 2.131
As you can see, if Vanguard were using the math acccording to .0,
they would use 17.16% and not 16.4%.
Gim
|
| 997.6 | IRA still a good deal | 2155::michaud | Jeff Michaud - ObjectBroker | Thu Apr 18 1996 11:08 | 13 |
| > A good example would be my IRA mutual funds, which I have
> not added to since I lost the tax deductibility.
I know that you know that while the contribution is not tax
deductable, that all gains/dividends/interest is till tax
deferred regardless of the deductability of the contributions
(and of course assuming one properly files 8606's, the non-
deductable contributions, since they were already taxed once,
are not taxed when withdrawn (distributed)).
So my question is why not add to your IRA, assuming one would
be setting that $2k away for retirement savings/investment
anyways?
|
| 997.7 | Real Returns | IVOSS1::VILLALOBO_GI | | Thu Apr 18 1996 13:58 | 14 |
| Although I don't recall the exact details, I know the SEC requires
stock fund returns be stated in very specific ways similar to what has been
mentioned in .5. There are rules for stating returns and yields on
money market and bond funds also. Hence, .0 is not correct. The SEC
wants a uniform way for investors to be able to compare results on
funds. There are no funny numbers here.
Back to .0, don't feel bad about ignoring your office mate who says he
made 25% per year on fund xyz. I'm sure he doesn't tell you about his
entire portfolio. There are some losers in there too. He won't tell
you about those. I would suggest you keep an open mind and determine
if mutual funds are for you. Those are real returns mentioned in .5.
But those are past performance. I can't tell you what will happen in
the future.
|
| 997.8 | Monies to IRAs | NCMAIL::YANUSC | | Thu Apr 18 1996 14:33 | 15 |
| re: .6
Jeff,
No mystery to why I haven't added additional funds to my IRAs. I have a
very diversified portfolio of funds, and overall I wanted to get to a
somewhat equal proportion of monies across many of them. The IRA mutual
funds had a good head start over the others, so my newer funds (outside
of other retirement vehicles such as the 401Ks) have been securing the
bulk of my newer investment $. And while I suppose I have balked at
keeping separate records of deductible and non-deductible transactions
in those IRA funds, I will bite the bullet in that regard and begin to
move additional monies into those funds soon.
Chuck
|
| 997.9 | Distributions = compounding | SCASS1::GASSER | | Fri Apr 19 1996 10:34 | 4 |
| You all have left out end of year distributions which increases the
number of shares you own. This must be factored in. Also some funds
give distributions twice a year.
|
| 997.10 | Look to Earlier Comments | NCMAIL::YANUSC | | Fri Apr 19 1996 10:55 | 10 |
| re: .9
If you look at my rebuttal in .3, I mentioned the lack of accounting
for "dividends and capital gains." Year-end distibutions would be part
of my capital gains comment, as would other distributions that take
place during the calendar year.
Chuck
|
| 997.11 | wait | WMODEV::GERARDI_B | America's PSG | Fri Apr 19 1996 11:06 | 12 |
| I didn't think that cumulative total returns were
simply an average of the yearly returns. I thought
(to use your scenario) the cumulative total
return would be -1% new value - old value
------------------------
number of years
No?
Bart
|
| 997.12 | ... | NPSS::URVA | | Fri Apr 19 1996 13:07 | 16 |
| The cumulative total return is:
((new value - old value)/old value *100)%
So if the beginning balance was $1000 and the ending balance was $4000
the cumulative return is (4000-1000)/1000 * 100 = 300%
The yearly average of return including compounding is:
(ending balance/beginning balance)**1/n - 1 ; ** is "to the power of"
So if the beginning balance was $1000 and the ending balance was $4000
then the average yearly return is 4**1/10 = 14.8%
/Bhooshan
|
| 997.13 | oops.. | NPSS::URVA | | Fri Apr 19 1996 13:09 | 4 |
| I forgot to mention that number of years = 10 in the example in
the previous reply..
/bu
|
| 997.14 | right-o | WMODEV::GERARDI_B | America's PSG | Fri Apr 19 1996 13:22 | 7 |
| So, in .0
The manager wouldn't state +12.5, s/he would state -1 wouldn't s/he, if
s/he were ethical?
Bart
|
| 997.15 | | HELIX::SONTAKKE | | Fri Apr 19 1996 15:25 | 6 |
| Besdies, each and every fund's numbers assume that you reinvest the
gains. They also ignore the tax that you pay on the paper gains every
year. I don't think simplifying to the extent of not considering
gains/dividends/taxes gives us very distored pciture.
- Vikas
|
| 997.16 | | MROA::YANNEKIS | | Fri Apr 19 1996 16:14 | 13 |
|
> Besdies, each and every fund's numbers assume that you reinvest the
> gains. They also ignore the tax that you pay on the paper gains every
> year. I don't think simplifying to the extent of not considering
> gains/dividends/taxes gives us very distored pciture.
Our marginal tax rate is 28% (OK 34% if you include MA tax). So for
each dollar of dividend I get to keep 66 cents after taxes. It seems
that ignoring dividands and taxes misrepresents the fund's performance
quite a bit.
GReg
|
| 997.17 | taxes do affect the return you see | DECCXX::REINIG | This too shall change | Fri Apr 19 1996 16:51 | 8 |
| Yes it does. But that's not the concern of the fund managers.
Besides, if the money in the fund is IRA money there are no
year-to-year taxes to worry about.
One of the many reasons buy and hold does so well is that you don't
have year to year capital gains and thus no taxes to pay on on them.
August
|
| 997.18 | | 2155::michaud | Jeff Michaud - ObjectBroker | Fri Apr 19 1996 17:25 | 5 |
| > One of the many reasons buy and hold does so well is that you don't
> have year to year capital gains and thus no taxes to pay on on them.
Not to mention commisions. Last year I paid more in commisions
than my net profit for the year :-(
|
| 997.19 | | HELIX::SONTAKKE | | Fri Apr 19 1996 19:27 | 2 |
| Sorry about that I said it wrong; I meant to imply that not considering
gains/dividends/taxes gives us distored pciture.
|
| 997.20 | | MROA::SRINIVASAN | | Fri Apr 19 1996 19:57 | 78 |
|
Hi , I was away from the notes file for few days since our systems were
down due to moving. Any how I checked my argument with several others
who work for few mutual fund companies etc ( They are all in my class).
They are of the opinion what is stated in .0 is correct.
Now for specific questions :
Re .3
>1. You have ignored dividends and capital gains, which even during down
>periods can be critical, particularly since you are buying shares in
>the mutual fund at a lower cost during those down periods. Leaving
>your initial investment at $100, and never adding in these dividends
>and capital gains, gives an incorrect picture of the real world.
I wanted to keep it simple to prove a point Hence I ignored the
dividends and Capital gains. and also Tax effects !
>2. Your %s used for your exercise are wildly inflated. How many funds
>would have any shareholders left if they were up 100% one year, down
>70% over the next two, and up 20% in the fourth. Using proportionally
>the same % you used, but with a potentially more real-world touch, I
>would show the following:
I used a widely inflated % to make thou example much more
understandable.
> Initial Investment Rate of Return Ultimate Amount
> $100 +10% $110
> ---- - 5% $105
> ---- - 2% $102
> ---- + 2% $104
This example is not correct. If the rate of return in 2nd year is -5%
it is not 5% of 100 but 5% of 110 which will be less than 105. Just a
small nit;-)
Using your table
Initial Rate of Total Value Average
investment return return
as reported
by mutual fund
100 +10% $110 10%
- -5% $104.50 2.5%
- -2% $101.60 1%
- +2% $103.63 1.25%
If you have received 1.25% return as reported by mutual fund you should
have at least $105.09 in the fund. However what you have is 103.63 which
means the mutual fund should have reported the 4 year return as 0.91%.
In fact we had a lengthy 1 hour discussion on this subject with several
people who work for Mutual companies and their argument is that this
practice is allowed and followed by all mutual fund companies. It
appears that there are NO regulations to this effect as to how the return
is reported. ( Also I heard there is a fine print in the prospectus
which explains how the % are calculated. I was told this is as
confusing as how home shopping Network calculates the mfg.suggested
retail price ;-) So as an investor one should do their home work and
calculate the return on their own instead of blindly following what is
being mentioned by the mutual funds.
In the real world most of us invest in mutual fund through our 401K and
since we keep adding money to the fund every week and the funds are
purchased at different price points we truly lose track of actual
return.
Based on the discususion I have had with some experts, I believe my
argument is valid. If you feel otherwise, Don't worry ! Be happy ;-).
Jay
|
| 997.21 | re: Fund Managers Don't Care About Taxes You Pay | UNXA::ZASLAW | | Fri Apr 19 1996 20:01 | 18 |
| Re: .17
> -< taxes do affect the return you see >-
>
> Yes it does. But that's not the concern of the fund managers.
Well, Morningstar also reports funds' returns after Federal income taxes,
making certain bracket assumptions. And when investigating new funds for
non-IRA investments, I personally look at such figures. If fund managers do not
care about it, perhaps investors could help make them care more. That might
discourage them from churning their portfolios excessively as so many of them
do.
> One of the many reasons buy and hold does so well is that you don't
> have year to year capital gains and thus no taxes to pay on on them.
I guess you're talking about individual stocks, not funds.
-- Steve
|
| 997.22 | I Still Stand By My Earlier Comments | NCMAIL::YANUSC | | Sat Apr 20 1996 21:52 | 27 |
| re: .20
Jay,
Your math may be off somewhat still (in your original example you
showed negative when it was really a positive return, and when you
recalculated mine from .3, it is really $104.46, not your lower
figure.) But let's not quibble over the math. The bottom line that we
are all looking for is return on our hard-earned investment $. I, too,
use the mutual fund company's own returns as nothing more than a
yardstick to give me a general idea of how well they have performed. I
purchase after independently verifying their performance through other
means. And I can assure you that many, many funds give outstanding
returns to their investors. The Vanguard Group, as well as certain
PBHG funds, have rewarded many of their investors handsomely.
In a day and age where information around individual stocks is
oftentimes known well before you and I can act on them, mutual funds
are quite often the only available means of participating in the
markets. Ignore them, and you ignore them at the peril of your own net
worth.
BTW, these are great notes conferences to participate in. I hope we
continue to have more discussions like this. I'll be on vacation next
week - I'll check when I get back to see if I missed any rebuttals.
Chuck
|
| 997.23 | | CXXC::REINIG | This too shall change | Mon Apr 22 1996 11:07 | 9 |
| > If fund managers do not care about it, perhaps investors could help
> make them care more.
There was an article in the Economist about this sometime last year.
There are at least a few fund that care about tax effects. By their
very nature, index funds have fewer tax effects since they don't buy
and sell shares that often.
August
|
| 997.24 | no supporting evidence? | SLOAN::HOM | | Wed Apr 24 1996 12:18 | 26 |
| > Hi , I was away from the notes file for few days since our systems were
> down due to moving. Any how I checked my argument with several others
> who work for few mutual fund companies etc ( They are all in my class).
> They are of the opinion what is stated in .0 is correct.
It would be useful to cite an actual stock (not bonds) mutual fund
return which supports your arguments. In .5, I have provided a real
life an example of a fund which contradicts your statement.
With over 4,000 mutual funds out there, there must one
which uses the method stated in .0 unless what the author of
.7 has said is correct:
" .... the SEC requires
stock fund returns be stated in very specific ways similar to what has been
mentioned in .5. There are rules for stating returns and yields on
money market and bond funds also. Hence, .0 is not correct. The SEC
wants a uniform way for investors to be able to compare results on
funds. There are no funny numbers here."
Without a real life mutual fund example to support your statement, what
conclusions can noters draw from this discussion?
Gim
|
| 997.25 | | SHRCTR::SRINIVASAN | | Thu Apr 25 1996 02:08 | 23 |
| re .24
I used one stock as an example to simplify the example ( believe it or
not- It happens to be true. There is a vaild argument as to why it is
mutual funds to report the way they do. In teh real world there are
thousands of stocks in a fund. Suppose the funds initial value is X ,
this amount does not remain constant. People keep putting in money and
the funds are bought at differnet NAV at each day and people keep
adding/ changing /.removing the portfolio each day. So even
calculating 1 year return becomes a big task. It is not as simple as
opening balance- closing balace . Now if the mutual funds have to keep
track of such volumes of changes in a 10 year peiod to calculate the
actual return it is a difficult task. One has to keep track of every
transaction, buy sell, money added at different NAV to to funds, money
taken out at differnet NAV etc etc, it is a huge taak . So the mutual
funds take the easy easy out. Let us not forget they have one of the
strongest lobby !.
In any case if you have differnet view on this, DON"T WORRY-BE HAPPY !
Jay
|
| 997.26 | Total Return is -not- difficult to calculate | EVMS::HALLYB | Fish have no concept of fire | Thu Apr 25 1996 09:52 | 5 |
| > In any case if you have differnet view on this, DON"T WORRY-BE HAPPY !
I'd rather have the truth, even if it does make me grumpy.
John
|
| 997.27 | Need to see supporting evidence | 12680::MCCUSKER | | Thu Apr 25 1996 10:35 | 15 |
| I also agree with .24 & .26. Lets see some real numbers to back this up.
.25>In any case if you have differnet view on this, DON"T WORRY-BE HAPPY !
Its not as simple as having a different view. This is not a philosophical
discussion. These are cold, hard, numbers that don't lie. Either they
are calculated as you say in the base note or they are not. .5 has offered
evidence to support that the base note is wrong. Unless there is evidence
to the contrary, we must determine that the base note is wrong. This is not
my _view_. It is an undisputable fact.
However this whole reply is just my $.02
Brad
|
| 997.28 | ... | NPSS::URVA | | Thu Apr 25 1996 12:04 | 16 |
| I also agree with the previous replies (which disagree with the base
note).
If you can determine the new share price of a mutual fund at the end of
a trading day, then determining the return is trivial. I can imagine
the first task being compute-intensive, that is why it takes a few
hours after the market close to come up with a new price. The mutual
funds have been doing this for a long time.
If a mutual fund cannot keep up with more money/new accounts and still
does not want to buy Alpha servers, then it will probably become a
closed end fund :-) Sometimes there are hiccups, like when Vanguard gave
its investors wrong tax information.
/Bhooshan
|
| 997.29 | | MROA::SRINIVASAN | | Thu Apr 25 1996 16:03 | 21 |
| I cannot give a real life example for a mutual fund from my personal experience
since I have never kept money in a given mutual fund for a long term ( not
even 2 years ). 401K funds returns are some what clouded since I have been
putting money weekly and funds are purchased at different price points and need
lots of data such as weekly purchase price / no of shares etc etc. So
attempting to do with our 401K will be time consuming. Also most of us have
been in these funds for less than 2 years due to 401K plan change last year.
One easy way to find out the actual return is for some one in this notes
file who has kept money in a mutual fund for 10 years ( no additions -
no withdrawals etc ) and post the results here.
Another way for any one to find out the discrepancy and the real return is
to get a copy of the valueline report for given mutual fund and also the
prospectus published by the same fund. I think one may have to do some
calculations to explain this. In any case I am some what busy with my
exams and office work. Perhaps when I find some time I will
make a comparison and publish ! Till then DON'T WORRY ! BE HAPPY -
since you are making xx% annualized return for past 10 years as per
the prospectus !).
|
| 997.30 | | WMODEV::GERARDI_B | America's PSG | Thu Apr 25 1996 16:38 | 8 |
| Some funds publish the statistic:
"$1000 invested in this fund in 1986 would be worth $2700 (whatever)
today." I'm not sure if you want to trust this stat, but Morningstar
or some other independant might do it too...
Bart
|
| 997.31 | ... | NPSS::URVA | | Thu Apr 25 1996 17:22 | 14 |
|
The trustworthiness of a fund's claimed returns over a period of time
may be debatable if you do not have proof. But there is no question
about the method of computing returns. If someone did measure the rate
of return over a 10 year period for a fund and looked at what the fund
prospectus had to say about the same period, the only way to get
different numbers is for the fund managers to have lied...
Perhaps Morningstar and the like do their own benchmarks as -.1
suggested. It's good to have independent watchdogs (God forbid if they
aren't!) like them.
/bu
|
| 997.32 | Actual numbers for 2 funds | DABEAN::NEARY | Bob Neary Lexington,Mass | Fri Apr 26 1996 09:25 | 11 |
| RE: last few
I was going over this at tax time. I can get exact numbers , but as an
example,
In 1991 I bought $3000 worth of CGM Mutual fund. It's now worth $4800.
In 1992 I bought $2000 worth of Columbia Special. It's now worth $4100.
I sent a check to open each and had planned to invest monthly. Instead
I started using Fidelity, so I never added to these two.
The results include reinvested dividends.
|
| 997.33 | | PERFOM::WIBECAN | Harpoon a tomata | Fri Apr 26 1996 11:45 | 12 |
| Re: .29
So on what basis do you make the claim in .0 that mutual fund managers average
returns by averaging the percentages? I can believe that your friend may have
said such things, but nothing I've seen in reporting on mutual fund returns
leads me to believe that mutual fund managers do likewise.
Also, in your discussion in .0, you use the term "cumulative return," which I
would take to mean cumulative total return over the period, NOT annualized.
(Usually "average annualized total return" is used if it is annualized.)
Brian
|
| 997.34 | Mutual funds only report time weighted return rate | MROA::SRINIVASAN | | Sun Apr 28 1996 05:55 | 147 |
| Re .33 and several others !
Well folks,
Here is the example some of you wanted !
Money management Industry uses TIME WEIGHTED RATE OF RETURN and not
DOLLAR WEIGHTED RETURN . Attached is from a text book " INVESTMENTS"
by Bodie, Kane and Marcus - Chapter 24 Portfolio Performance Evaluation.
Some of you may have the problem understanding the concept. ( After
all this text book is for a graduate level course INVESTMENTS - 801
and not Investments101.
So again next time some one tells you they got a 10 year return of
XX% as per the information published by the mutual fund, just smile. Don't
try to explain them the Time weighted rate of return and Dollar weighted
rate of return. They may not understand it and may tell you that you are full
of ( what ever ).
Next time you talk to your broker and he tells you that fund A 's 10 year
return is XX%, ask him under what method they calculated the return.( Most of
the brokers are nothing but a used car salesman ( IHMO ) and will not know
the answer.
After reading this article below, If you still feel that you are getting the
return as reported by mutual fund managers - DON'T WORRY - BE HAPPY
Regards
Jay
PS : By the way morning star also uses the Time-Weighted rate of return !
-----------------------------------------------------------------
INVESTMENTS
( Bodie, Kane , Marcus )
Chapter 24 - Portfolio Performance Evaluation.
The rate of return of an investment is a simple concept in the case of a one
period investment. It is simply the total proceed derived from the investment
per dollar initially invested. Proceeds must be defined broadly to include
both cash distributions and capital gains. For stocks total returns are
dividends plus capital gains.
To set the stage for discussing the more subtle issues that follow let us
start with a trivial example. Consider a stock paying a dividend of $2 annually
that currently sells for $50. You purchase the stock today and collect the $2
dividend and then you sell the stock for $53 at year end. Your rate of return
is
(Total Proceeds/Initial investment) = (Income + Capital gain) / 50
= ( 2 + 3 ) /50
= .10
= 10%
Another way to derive the rate of return that is useful in the more difficult
multi-perid case is to set up the investment as a discounted cash flow
problem. Call "r" the rate of return that equates to present value of all
cash flows from the investments with the initial outlay. In our example the
stock is purchased for $50 and generates cash flow at year end of $2
( dividend ) plus 53 ( sale of stock ) . Therefore we solve
50 = ( 2 +3) / ( 1+r) to find again that r= 10%.
Time weighted Returns Versus Dollar- Weighted returns
When we consider investments over a period of time during which cash was
added to or withdrawn from the portfolio, measuring the rate becomes more
difficult. To continue our example suppose you were to purchased 2nd share
of the same stock at the end of the first year and hold both shares until
the end of year 2 at which point you sell each share for $54.
Now the total cash outlays are :
Time Outlay
0 $50 to purchase the first share.
1 $53 to purchase second share a year later.
Time Proceeds
1 $2 dividend from initial purchased share
2 $4 dividend from 2 shares held in the 2nd
year plus $108 received from selling both
shares at $54 each .
Using the discounted cash flow approach( DCF ) , we can solve
for average return over the 2 years by equating the present values of the
cash inflows and outflows.
50 + ( 53 / 1+r) = {(2/1)+r} + { 112/(1+r)( 1+r) }
resulting in r= 7.171%.
This value is called the internal rate of return or the DOLLAR-WEIGHTED
RATE OF RETURN on the investments. It is " dollar -weighted" because the
stock performance in the 2nd year when 2 shares are held has a greater
influence on the average overall return when only one share is held.
An Alternative to the internal or dollar weighted return is the TIME-WEIGHTED
RETURN. This method ignores the number of shares of stock held in each
period. The stock return in the first year was 10%. ( A $50 purchase provided
$2 in dividends and $3 in capital gains ). In the second year the stock had
a starting value of $53 and sold at the year end ( period 2 ) for $54 for a
total one year period rate of return of $3 ( $2 dividend plus $1 capital gain).
divided by $53 ( the stock price at teh start of the second year ) or 5.66%.
The time weighted average is the average of 10% and 5.66% which is 7.83%
( Note that dollar weighted rate of return computed as 7.117%. This Time
Weighted average return considers only the period by period returns without
regard to the amounts invested in the stock in each period.
Again Note that the dollar weighted return is less than the time weighted
return in this example, The reason is that the stock fared relatively poorly
in the 2nd year when the investor was holding more shares. The greater weight
that the dollar weighted average places on the second year return results in
a lower measure of investments performance. In general Dollar & Time weighted
returns will differ and the difference will be depending on the
configuration of period of returns and portfolio composition.
Which measure of performance is superior ? At first it appears that the dollar
-weighted return must be more relevant, After all more money you invest in a
stock when its performance is superior, the more money you end up with .
Certainly your performance measure should reflect this.
Time weighted returns have their own use, especially in the money management
/ Mutual funds industry. This is so because in some important applications a
portfolio manager may not directly control the timing or the amount of money
invested in securities. 401K fund management is a good example. A 401K fund
manager faces the cash inflows in to teh fund when 401K contributions are
made and cash outflows when teh pension benefits are paid. Obviously the
amount of money invested at any time can vary for reasons beyond the fund
managers control. Because the dollars invested don't depend on the managers
choice, it is inappropriate to weight returns by dollars invested when the
measuring the investment ability of the fund manager. Consequently the money
Management industry uses TIME - WEIGHTED RETURNS for performance evaluation.
ARITHMETIC Versus GEOMETRIC AVERAGES
Sharpe Performance measure,
Treynor Performace measure
Jensen Performance measure
--------------------------------------------------------------------------------
( Well This is yet another twist ! I won't go in this for now ) !!! It becomes
too complicated.
|
| 997.35 | | WMODEV::GERARDI_B | America's PSG | Mon Apr 29 1996 09:30 | 6 |
| When you are dealing with money added in over time, however, it
you can actually out perform what the manager is reporting. So, it
is hard to say which is good and which is a bad way to measure.
Bart
|
| 997.36 | it's not used because it isn't useful | MKOTS3::LEVY_J | | Mon Apr 29 1996 14:24 | 14 |
| Using dollar-weighted rate of return, a fund which had nine mediocre
years, and one GREAT year, could concoct an "average annual rate of
return" which made it look much better than it really was.
After reading the textbook excerpt, I conclude:
The money-management industry uses time-weighted rate of return
because it allows an apples-to-apples comparison of performance.
Dollar-weighted rate of return is specific to one's actual investments
over specific time periods, and therefore is a customer-unique measure
of THAT CUSTOMER'S investment performance.
It has no value in making general comparisons.
|
| 997.37 | .0 is wrong | SLOAN::HOM | | Mon Apr 29 1996 17:30 | 72 |
| The points made on time weighted vs dollar weighted returns is correct and
can be substantiated with other texts such as the Theory of Interest
by Kellison. Readers will appreciated the finer points between these
two rates of return.
In .0, the author writes:
"One of my colleagues in my office is a Mutual fund fanatic. He
used to tell me how Fund X got him 25% return for past 10 years
etc etc. He used to show me the Morning star reports for the
the funds which out performed the market in the past 10 years
etc etc. ...
However after my graduate program class in Investments - Mutual
Funds Chapter ! ) it became clear to me that the so called 25%
return in 10 years is nothing but non-sense. ! "
The author further states that the 5 year, 10 year average, etc.
return for mutual funds was just the numerical average of each of the
years return with an implication that these cumulative rates are not
achievable.
That is NOT the case for calculations of 5 years (or 10 year, etc
returns) as required for mutual funds. For example, the "5 year average
return" quoted in mutual fund annual reports, etc is NOT the numerical
average as stated by Brodie, et. al or by note .0 but the true rate of
return which is indeed achievable. These returns of course, assume a
fixed starting investment with no additions or withdrawls, reinvestment
of dividends, and ignores taxes (like an IRA account).
IF THESE RETURNS ARE ANYTHING BUT TRUE EFFECTIVE RATE OF RETURNS WHICH
RESULTS IN REAL ACHEIVEABLE RETURNS, YOU CAN BE ASSURED THAT SOMEONE
WOULD HAVE POINTED IT OUT A LONG TIME AGO. If a fund has a
10 year return of 15%, then $10K invested in that fund 10 years
ago with be worth $10K x (1.15)^10 = $40.46K.
I'm afraid that when "some one tells you they got a 10 year return of
XX% as per the information published by the mutual fund", that person
is right and has the last laugh.
Attached is real life example of where someone did acheive a modest
13.29% "5 Year average return" ( [1+.1329]^5 = 1.866 ). The mutual fund
is the Vanguard Wellesley Fund held in an IRA.
----- real returns with real data -------
Statement Statement Yr to Yr Cumulative As reported As Reported
Date Value Gain Value by Vanguard by
Morningstar
12/31/90 10,000.00 1.000
12/31/91 12,150.14 21.50% 1.215 21.6% 21.57%
12/31/92 13,211.42 8.73% 1.321 8.7% 8.67%
12/31/93 15,146.77 14.65% 1.515 14.6% 14.65%
12/31/94 14,474.90 -4.44% 1.447 -4.4% -4.44%
12/31/95 18,659.59 28.91% 1.866 28.9% 28.91%
5 Yr returns 13.29% 13.29% 13.29%
====== ====== ======
Numerical 5 yr average return as suggested by .-0 is 13.88%
If indeed note .0 is correct, then simply posting an example of a real
mutual fund which supports your statement in .0 will prove the point.
I would be very surprised if such an example exists.
Unfortunately the concepts of time-weighted vs dollar weighted returns
(which are indeed used in portfolio management) merely obfuscate the
issues raised in .0: Are the 5 year and 10 year returns
reported real or not? They are indeed real and achievable.
Gim
|
| 997.38 | I disagree with .37 | SHRCTR::SRINIVASAN | | Mon Apr 29 1996 18:04 | 48 |
| re .37
IMHO the Time weighted average is used by the
Mututal funds for 5 year and 10 years. Also the example given in .0 is
one example to prove the point. In fact the text book calls for similar
example in Arthimatic and Geometric averages ( Mutual funds report only
report Arithmatic Averages and NOT geormetric averages. More generally
for an n period investment the geometric average rate of return is
given by
1+ rg = [ (1+r1)(1+r2 )....... ( 1=rn)} to the power of 1/n where rt
is the time perid. As always geomertic mean is lower than the
arithmatic mean.
The intutive advantage here is that this is good for portfolio manages
who have no control over the cash flows, For example a pension funs
manager may see a large cash outflow to pay off benefits which would
hurt the dolalr weighted average return through no fault of the
portfolio manager. Also for dolalr weighted averages you need portfolio
values at eash date tehre is a cash flow, which may be difficult in
practice. So which is a better measure. If you are trying to perdit
returns ( Short term 1 - 2years or long term 5 - 10 years, the mutual
funds use the Arithmatic avearge becasue Geometic avearge is downward
biased. So Mutual funds wants to llok good and report what is
considered a better number. ( This is perfectly legal and they follow
all follow the guidelines of AIMR ( Association of Investments &
Reaserch ) Performace presenttation standards.
Now in conclusion, I am not disputing the fact some fund may give XX
return ( Say 20% ) AS reported in their prospectus or morning star- But
siince they used time weighted return rate and arithmatic average as
opposed to dollar weighted return and Geormetric average, the actual
return is less always less than the 20%.
Here are some information from AIMR whose guidelines are followed by
the mutual funds.
Annual returns ( 1 year- 2 year, 5 year and 10 year ) should be
reported for all years individually as wel as for longer periods. Firms
should present TIME WEIGHTED - AVERAGE RATES of Returns with
portfolios valued at least quarterly ". There are 3 pages fo
guidlelines, which has interesting fine prints !
Well ! I would like to give more examples on this _ But it is enough
for now !
Regards
Jay
|
| 997.39 | I won't worry, I'm happy | DECWET::ONO | The Wrong Stuff | Mon Apr 29 1996 21:55 | 70 |
| re: .34
.34> Some of you may have the problem understanding the concept. ( After
.34> all this text book is for a graduate level course INVESTMENTS - 801
.34> and not Investments101.
I find this statement incredibly insulting!
re: .38 and AIMR
.38> practice. So which is a better measure. If you are trying to perdit
.38> returns ( Short term 1 - 2years or long term 5 - 10 years, the mutual
.38> funds use the Arithmatic avearge becasue Geometic avearge is downward
.38> biased. So Mutual funds wants to llok good and report what is
.38> considered a better number. ( This is perfectly legal and they follow
.38> all follow the guidelines of AIMR ( Association of Investments &
.38> Reaserch ) Performace presenttation standards.
I looked at the AIMR Performance Presentation Standards too.
(See http://www.aimr.com/aimr/advocacy/pps/pps-home.html). This
is some text from the summary:
"To be considered in compliance, a manager's presentations must
incorporate the following practices:
". Use of time-weighted rates of return, with valuation on at
least a quarterly basis AND GEOMETRIC LINKING OF PERIOD
RETURNS. [emphasis added]"
Guess what, the AIMR standards call for geometric averaging --
not arithmetic averaging as stated in .38.
re: .38 and the averaging funds use
.38> example in Arthimatic and Geometric averages ( Mutual funds report only
.38> report Arithmatic Averages and NOT geormetric averages. More generally
Vanguard's web site includes a set of pages called Vanguard
University. In the module on "Selecting Specific Mutual Funds"
there is a page describing the difference between cumulative and
average total return. The page is at
(http://www.vanguard.com/educ/module3/m3_4_2.html)
On that page is the following text:
Mutual funds report total return both on an average annual basis
and a cumulative basis. An example is perhaps the best way to
illustrate these concepts. Suppose your $1,000 investment in a
fund increased in value to $2,000 over a period of ten years.
Your cumulative return over that ten year period would by +100%.
Expressed on an average annual basis, your return would be +7.2%.
(Note that the average annual return takes into account yearly
compounding.)
Again, here is a real mutual fund family that uses geometric
averaging, rather than arithmetic averaging.
re: .38 and real-world examples
> Well ! I would like to give more examples on this _ But it is enough
> for now !
I haven't seen a real example yet. Gim has given two. The
textbook does not provide an example of a real mutual fund. It
provides an example of how the calculations can be done.
I've convinced myself that .0, .34, .38 are bogus, as far as
mutual fund comparative performance is concerned.
Wes
|
| 997.40 | Still no supporting evidence... | 12680::MCCUSKER | | Tue Apr 30 1996 10:13 | 46 |
| .38> IMHO the Time weighted average is used by the
.38> Mututal funds for 5 year and 10 years.
This is not something that one can have an opinion on. Either they do or they
do not. We have seen examples where they do not. If you insist that your
statements are correct, then give us an example using real data, for real funds.
.38> Also the example given in .0 is one example to prove the point.
First of all ther is no example in .0. The only thing the example in .0
is is a poor selection of data to illustrate a method. But regardless, since when
do hypothetical examples prove anything?
.34> Some of you may have the problem understanding the concept. ( After
.34> all this text book is for a graduate level course INVESTMENTS - 801
.34> and not Investments101.
I also find this to be arrogant and offensive. I have taken graduate level
courses in many areas (as have many others in this conference I'm sure) and
to be quite honest, graduate level courses, and the ability to pass them
does not impress me. What would impress me is the ability to take the knowledge,
understand it, and apply it to real world situations. I have not seen that
yet from the base noter.
(Step onto soapbox, mild flame on)
In recent years (since the mid-late 80s) MBA's and the people who hold them
have lost a lot of the respect they once had. I think that if you re-read
this string, one can see one of the reasons why. Here is an example of
someone making a statement, having a number of people say he's wrong, and
instead of backing up and re-evaluating the situation, he continues to stand
firm, pointing to his text books and the piece of parchmant on the wall: "I'm
right, it says so right here in my text book, Professor KnowItAll said so in
class and I've got the degree to prove it." IMHO, it is this inability to
_apply_ the knowledge gained from MBA courses (instead of regurgitating it)
which has given such a bad name to MBA's in recent years.
(step off soapbox, extinguish candle)
I do not understand why the base noter doesn't provide us with a real example
to back up all the theoretical data he has provided. The further this string
goes, the more evidence we see to the contrary.
another $.02
Brad
|
| 997.41 | This just doesn't make any sense | EVMS::HALLYB | Fish have no concept of fire | Tue Apr 30 1996 11:44 | 14 |
| .38> The intutive advantage here is that this is good for portfolio manages
.38> who have no control over the cash flows, For example a pension funs
.38> manager may see a large cash outflow to pay off benefits which would
.38> hurt the dolalr weighted average return through no fault of the
.38> portfolio manager.
I do not understand this. If there is a large cash inflow, the fund has
a lot more money but an equally larger number of shares. Conversely when
there is a large outflow the number of shares shrinks correspondingly.
The fund's per-share price is adjusted to relect the dollar flows!
The portfolio manager should be rated on the per-share value of the fund,
not the capitalization level.
John
|
| 997.42 | .34 either forgot a :-), or is naive :-) | 2155::michaud | Jeff Michaud - ObjectBroker | Tue Apr 30 1996 12:14 | 13 |
| > .34> Some of you may have the problem understanding the concept. ( After
> .34> all this text book is for a graduate level course INVESTMENTS - 801
> .34> and not Investments101.
> I find this statement incredibly insulting!
FWIW, I don't find it insulting. How can one be insulted by a
statement that shows the naivity of the speaker of said statement?
Ie. A graduate course numbered "801" is just an "intro" course itself,
not a high level course. Remember that a Graduate program in a
given area does not require one to have an undergrad degree in
the same area. In fact I believe some programs don't require one
to take the "801" (or equiv) course if one has an undergrad in
the same disipline.
|
| 997.43 | Last thoughts | SLOAN::HOM | | Tue Apr 30 1996 17:21 | 32 |
| re: .40
> I do not understand why the base noter doesn't provide us with a real
> example to back up all the theoretical data he has provided. The
> further this string goes, the more evidence we see to the contrary.
One explanation on why a real example can't be provided is that the
author of .0 is wrong. Citing an example of a real PUBLICLY TRADED
MUTUAL fund (where the results are available to all for inspection)
which reports performance as described in .0 will certainly resolve this
string.
Until that example is provided, the conclusion as .39 has stated is
"that .0, .34, .38 are bogus, as far as mutual fund comparative
performance is concerned."
Re: .34
> Next time you talk to your broker and he tells you that fund A 's 10 year
> return is XX%, ask him under what method they calculated the return.( Most of
> the brokers are nothing but a used car salesman ( IHMO ) and will not know
> the answer.
This is a nit but I would be very surprised if any readers of this
conference buys mutual fund through a broker - except maybe for Charles Schwab
or Fidelity.
Gim
|