In microsoft.public.money, Tom Z wrote:

- quote -


Here is the Money 2006 report from your test case, agreeing with
your Money 2003 finding:


:07/13/05 Page 1

:Performance by investment account
:Investment test 2

:07/13/04 through 07/13/05


: Value on Value on Cost Realized Return % Rate Annual
:Investment 07/12/04 07/13/05 Basis Gain / Loss for Period of Return % Return

:Investment test 2
:AnMMF 0.00 2,200.00 2,200.00 0.00 0.00 0.00 0.00
:Astock1 0.00 1,100.00 1,000.00 200.00 300.00 15.00 44.75
: Total Investment test 2 0.00 3,300.00 3,200.00 200.00 300.00 7.14 19.75

:Grand Total 0.00 3,300.00 3,200.00 200.00 300.00 7.14 19.75

Note that you will be paying CG tax on the $200 if this is in a
taxable account.

- quote -

You make a good point.

It is better to reply to yourself if you want to add on to your
presentation. That allows following your information better than
starting a new thread each time.

Mine is flaky too. I have something like a -76% return in one account, even
though that account has had nothing in it but cash at small interest (not
even a money fund). Maybe it would work better if I actually bought
something.

I think what matters with ROI is what is actually "invested", and I think it
should actually be calculated only for individual buy/sell pairs. You could
calculate it this way:

Buy ST for 1000, sell for 1200, ROI = 20%
Buy ST for 1000, sell for 1100, ROI = 10%
Buy MM for 1000, assume ROI approx 1%
Buy MM for 1200, ROI ~ 1%

So I would calculate your ROI for all transactions like this:

(1000 * 0.2 + 1000 + 0.1 + 1000 * 0.01 + 1200 * 0.01) / (1000 + 1000 + 1000
+ 1200) ~ 7.7%

Of course, commissions on the trades will lower the return.

Then of course you have the possibility of ongoing investments - ones you do
not sell. In that case, you have to consider annual return over the original
investment amount. Also, is the return automatically reinvested? Should you
include the reinvestment amount as part of the investment when calculating
ROI? I don't think the answers to these questions are standardized, so ROI
should not be considered a definitive measure of investment performance.

Other measures have problems too - IRR for instance is often useful, but can
produce unreliable, even preposterous, numbers under certain circumstances.

Disclaimer: I am not a financial analyst.




"Tom Z" <TomZ[at]discussions.microsoft.com> wrote in message
news:3F9EE3E5-6C4F-487E-9219-11A258CB1C24[at]microsoft.com...
- quote -

Consider the following:
1. on 1/1/05 you put $3000 into an investment account, $1000 goes into a
Money Market (MM) and you buy 200 shares of stock (ST) [at] $10.

2. on 4/1/05 ST has gone up to $12 and you sell 100 shares and put the
proceeds into MM, so now you have $2200 in MM and $1200 worth of ST.

3. on 7/1/05 ST has gone down to $11, so now you have $2200 in MM and $1100
worth of ST and your account is worth $3300.

- Doesn't this mean that your ROI for this 6-month period should be 10%?
($3300/$3000 - 1)?

- If you do this in Money 2003 your ROI comes up as 7.1%.
- The problem seems to be the Money calculates ROI as (Return for
Period)/(Buys)-1; this is apparent if you look at the "Performance by
Investment Account" report. The "Buys" are $1000 for MM and $2000 for ST in
step 1, and $1200 for MM in step 2, so performance is $300/$4200 = 7.1%.

- so if you add another step, say you sell the remaining 100 ST on 7/1 at
$11 and put this into MM; this creates another $1100 "Buy" for MM and your
ROI goes down to 5.7% calculated as $300/$5300.

- Does any of this seem ridiculous to anyone else - shouldn't the ROI be 10%
no matter how many "Buys" there are?