Forum Archive Index - December 2001

["tennyson@caverock.net.nz" <tennyson@caverock.net.nz>]

Hi John,

I see no-one has commented on your little dissertation on internal 
rate of return so far, so I will!



> Internal Rate of Return (IRR)
> 
> A graph tracking the unit price over time of an investment tells us
> a certain amount about the investment, but it does not tell us the
> full story.
>
>
Right!  Looking just at the unit price leaves the dividend stream, 
and any cash issues, out of the equation! 
>
> 
> The Internal Rate of Return (IRR) on an investment brings into
> account the cash flows pertaining to the investment, as well as the
> value of the investment, so provides a more comprehensive measure
> how well an investment is performing.
>
>
Yes, although a graph of the gross sharemarket index (including 
dividends) can have a similar effect.
>
> 
> I haven't tired it but, I suspect, graphing the IRR over time
> alongside the unit price would be an interesting exercise.
> 
>
Why haven't you tried it?

'IRR' is basically just an averaged return on shareholders equity.
  
Or, put another way, 'ROE' is just the internal rate of return on a 
one year basis.   Looking at the trend of ROE is somewhat akin to 
looking at the 'IRR'.   This is one of the screening criteria on the 
focus investment group. 

I'll do a couple of 'thought' share graphs for you.  Suppose a 
company had shareholders equity of $1.00, and paid out all of its 
profits, amounting to 8c per share, every year.  The profit was the 
same year in year out.   Assuming full imputation credits applied, 
the Internal Rate of Return to you would be 8%, year in, year out.

The Return on Shareholders Equity would be 8% every year too.   The 
share price would probably have a long term average of about $1 but 
might go up and down a bit with prevailing interest rates.

Now for the second 'thought' graph.

Suppose this same company, rather than paying out the 8c per share 
profit as a dividend, kept the money, then reinvested it back in the 
company.  Let's suppose they were able to keep earning that 8% on the 
reinvested capital.   We cannot forecast exactly where the share 
price will be at any particular date in the future.  But the expected 
price in a years time would be $1.00 + 8c = $1.08.  The earnings for 
the following year would be 8% of $1.08, which is 8.6c.  All the 
earnings are retained so the expected share price two years out in 
the future is $1.16.6c, close enough to $1.17.   Notice that in the 
first year we expect the share price to go up 8c, whereas in the 
second year we expect the share price to go up by 9c.   So this means 
the shareholder can expect a better investment performance in the 
second year compared with the first year, right?

Not really.   Although you have earned more money in the second year 
than the first year, the first years result was attained on capital 
of $1.00 per share and the second years result was attained using 
capital of $1.08.   So looking at the bare share price graph which 
went up 8c in the first year and 9c in the second year might mislead 
you into thinking that the share had done better in the second year.  

If instead you had looked at the IRR and had found this to be 8% on 
each year (which was the actual figure), you would see that the the 
financial performance of each year was in fact identical.

So you are quite correct John.   IRR will give you a better 
measurement of company performance than bare share price movements.

I have picked to simple examples.  One where the all of the earnings 
are paid out as dividends and the other where none of them are.  If 
some of the earnings are paid out as dividends and some retained then 
the calculations become slightly more complex.  But not grossly so.

In fact, projecting internal rate of return forward into the future 
is one of the exact same techniques that Warren Buffett uses.  This 
is the basis behind the 'Let's Ask Warren' series of calculations on 
the Focus Investment Board.   You can use a spreadsheet to do these 
calculations, but you don't desperately need to.

So maybe now as a confirmed 'IRR Man', you might like to participate 
in our Focus Investment group?   


>
> 
> Now if one has a portfolio of investments and wants to see the
> overall IRR it seems to me that one needs to copy all the data onto
> another sheet, sort it into date order and then apply the IRR
> function to this "Total" table. To keep calculations up to date one
> would have to be entering data into the tables for the individual
> investments and then again into the "Total" table.
>
>

Yes, I think you are correct, you do need a separate 'total' 
spreadsheet.   But I am sure it would be possible to automate the 
process.   You know how to work out IRR for a single company.   In 
'Excel' each spreadsheet is set up as part of a workbook.   So if you 
had, say, ten individual companies, you could program a master 
spreadsheet so that the master spreadsheet totalled the input figures 
from the individual company spreadsheets, saving you entering the 
information twice. 

My own view is that this would be a waste of time, as the most 
important point I would be interested in is what parts of my 
investments are performing and what are not.   Combining them into a 
'total' doesn't allow me to fix anything that isn't performing to my 
satisfaction.
>
> 
> The UNIT PRICE of the Portfolio
> 
> Unless cash inflows and outflows during a period have been zero, you
> cannot use the difference between the value of a portfolio at the
> end of a period and that at the start, as a measurement of how well
> the portfolio has done.
>
>
Yes.
>
> 
> I presume that if you take the cost of the portfolio and divide it
> into its value you will get a Unit Price at that point in time. Is
> that how Unit Fund prices are calculated?
>
>
If the fund itself is not listed on the sharemarket, the unit value, 
(as measured by the constituent parts of all the shares that make up 
the fund) becomes, in effect, the unit price.

The historic cost of each unit is not necessarily related at all to 
its present day value.
>
> 
> Can one then take this Unit Price at the start and end of a period
> and then use it as a measure of performance?
>
>
Not as a measure of the fund performance.   The timing of you buying 
in and selling out of a fund should have no effect on the long term 
performance of the fund.
>
>
> But what if you put a lump of new cash into the portfolio just
> before the end of the period. Would that not dilute what you are
> trying to measure?
> 
>
Yes.  So the best way to measure the performance of a fund is to 
adjust the figures to 'what they would have read' if you had made no 
deposits or withdrawals.   For calculation purposes, you take your 
deposits you made during the year back out, and put any withdrawals 
you made during the year back in.
> 
>
> What about distributions received. Are they deducted from the cost
> of the portfolio?
> 
>
You could do that.   But wouldn't it be easier just to add any 
distributions received onto the value of your portfolio at the end of 
the year?   In other words treat any income you received as 
equivalent to a one off capital gain
 
Has that made things any clearer?  SNOOPY



---------------------------------
Message sent by Snoopy 
e-mail  tennyson@caverock.net.nz
on Pegasus Mail version 2.55
----------------------------------
"You can tell me I'm wrong twice, 
but that still only makes me wrong once."

----------------------------------------------------------------------------
To remove yourself from this list, please use the form at
http://www.sharechat.co.nz/chat/forum/