tag:infinitekind.tenderapp.com,2009-01-14:/discussions/investments/17-report-calculations
Infinite Kind: Discussion
2018-03-17T09:23:09Z
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-19T18:43:39Z
2015-06-19T18:43:39Z
Report Calculations
<div><p>"Semi-actively" means I'm working on it at the same time as
other issues while also doing my main job, which is customer
support. It's not the same as not working on it at all. I guess I
could have just said "actively" and saved the confusion.</p>
<p>That said, I think I've fixed the ROI bug in build 1235, which
which is now available on the <a href="http://infinitekind.com/preview">preview site</a> . I'm still
looking at edge cases that are harder to test in simple test files,
so I'd love for you all to take a look and let me know how it looks
to you.</p>
<p>Still working on the cost basis reported on the Capital Gains
report.</p>
<p>s©tt, Moneydance has had the same lead developer since it
was founded, along with a rotating cast of supporting developers.
So, yes, your assumption is wrong. I would appreciate it if you
could stop making assumptions about the people who work here and
how we work internally. I believe Ben has made this request of you
before, and I'd like to make it again.</p>
<p>Ian<br>
Infinite Kind Support</p></div>
Ian L
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-19T19:30:24Z
2015-06-19T19:30:24Z
Report Calculations
<div><p>Thanks for the response Ian. Rather than being defensive, you
might start trying to be more open and inclusive with your user
community. Despite YOUR assumptions, I am a big supporter of
Moneydance and have been using it - and buying upgrades - for many
years. That's the reason I am immensely frustrated at how your team
has turned out "upgrades" for the past couple of years which have
been VERY bad, in terms of user engagement and testing and bugs.
Zero transparency. I am vocal, precisely because I am passionate
about the product. But time and time again, I see your team respond
defensively to a number of serious complaints from long-time users.
You need to consider that.</p></div>
JFG
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-19T22:28:09Z
2015-06-19T22:28:09Z
Report Calculations
<div><p>I'm not quite convinced.</p>
<p>This is from one line of a report that shows changes for a
single day, although Dividends stuff it up I'm using it to see the
change in market value between days.</p>
<p>19/6/15: 27,387.88<br>
Buys: 0<br>
Sells: 0<br>
Income: 0<br>
20/6/15: 27,642.73<br>
Gains: 0<br>
Return Amt: 254.85<br>
Return %: 0.93<br>
Annual % (ROI): 2839.13</p>
<p>I still have a couple showing annual ROI with 12 significant
digits, for example:</p>
<p>19/6/15: 615.00<br>
Buys: 0<br>
Sells: 0<br>
Income: 0<br>
20/6/15: 660.00<br>
Gains: 0<br>
Return Amt: 45.00<br>
Return %: 7.32<br>
Annual % (ROI): 335763727615.67</p>
<p>Des</p></div>
dwg
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-20T19:41:26Z
2015-06-20T19:41:26Z
Report Calculations
<div><p>It is not far off? I would really like an explanation of exactly
how MD is doing the calculation. So an estimated 1 year return
could be done in a couple of ways. What I think is being done
is<br>
Annual return = exp (365*daily return)<br>
This gives for the first example 2980% annual ROI which is close
but not exactly what MD gets. Or if I work backwards, MD assumes a
359.8 day year.</p>
<p>Again transparency. But this is not really the right way. See
wikipedia for IRR. If I buy 1 share a $1 and then a year later buy
another share at the current price of $2, I now have $4 worth of
stock purchased for $3 and the return is 33%. BUT it is actually
much higher since that $1 was held for a year and went to $2 which
is a 100% return.</p>
<p>DOES MD CALCULATE THE RETURN USING IRR OR DOES IT just take the
beginning and ending values. The IRR is a valid way to calculate
this annual return but one needs a more complicated sample than a 1
day report.</p>
<p>I don't speculate on internal personnel at MD, but I really
don't want to speculate on how MD does its calculations. I want to
know.</p></div>
sth
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-21T00:30:43Z
2015-06-21T00:30:43Z
Report Calculations
<div><p>Initial results with 1236 are very good. My 3 simple test cases
(which gave way-off results before) now get exactly the same ROI in
the MD report as the XIRR() spreadsheet function.</p>
<p>I will try some more cases as well as some from this thread.</p></div>
ljb
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-21T01:32:21Z
2015-06-21T01:32:21Z
Report Calculations
<div><p>Ian, Can I email you privately with some calculations that seem
to be off?</p></div>
sth
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-21T02:06:40Z
2015-06-21T22:02:44Z
Report Calculations
<div><p>Des,</p>
<p>Your first example is: 2015-06-19 value 27387.88; 2015-06-20
value 27642.73 and MD2015 is reporting Annual % (ROI) = 2839.13</p>
<p>MD2015(1236) is right - that is the correct value for IRR
(internal rate of return). This assumes you are reporting through
the second date, or you sell everything on that date for that
amount and report to some future date.</p>
<p>The calculation goes like this: NPV (net present value) of
27642.73 with a return of 2839.13% for 1 day (0.00274 of a year)
is:</p>
<pre>
<code> 27642.73 / (28.3913+1)**0.00274 = 27387.85</code>
</pre>
<p>which is (almost exactly) what you started with, so the NPVs sum
to zero. Which means the IRR is 2839.13%.</p>
<p>But it makes more sense to look at it this way. You earned 0.93%
in a day. Compound that for a year: (1+0.0093)**365-1 = about
2830%.</p>
<p>(Edit: corrected last equation, forgot to subtract 1 at the
end.)</p></div>
ljb
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-21T21:44:19Z
2015-06-21T21:44:19Z
Report Calculations
<div><p>Hmm I think my rough mental arithmetic I did at the time was out
of whack.</p></div>
dwg
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-22T01:15:19Z
2015-06-22T01:15:19Z
Report Calculations
<div><p>Des, I tried your second case (2015-06-19 value 615.00;
2015-06-20 value 660.00)</p>
<p>The spreadsheet XIRR() function returns an error code, and
MD2015 reports a huge, apparently random number for ROI in the
report.</p>
<p>It's a 7.32% return in 1 day. Compound that for a year:
(1+0.0732)**365-1 and you get a huge number, around 1.58e11.</p>
<p>Since ROI is calculated by iterative approximation, a case like
this will probably exceed the iteration limit without converging
(that's what the spreadsheet error message I get means). Or it will
just give a wild number that may or may not be close to the actual
answer.</p>
<p>In my opinion, the right thing for MD to do in this case is to
not report an ROI at all. But the calculation itself isn't at
fault.</p>
<p>By the way... I had trouble getting these tests through MD2015.
My security history entries kept getting lost, or old ones
re-appeared, and changes to my memorized reports weren't sticking.
(Intermittent - not repeatable - so these will be very hard to
fix.) I ended up making a change, restarting MD, check, repeat. So
even if ROI is fixed, there a long way to go with this before I
would use it.</p></div>
ljb
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-22T01:38:14Z
2015-06-22T01:38:14Z
Report Calculations
<div><p>ljb,</p>
<p>I only use this report to get a guide whether the share prices
have updated for the day. I would tend to only look at overall ROI
after I have sold out of an investment.</p>
<p>While I hold the investment I'm more interested in separating
income from capital gains so would tend to look at dividend Yield
on one hand, so what the income return is like, then separately
look at potential capital gains from selling.</p>
<p>Des</p></div>
dwg
tag:infinitekind.tenderapp.com,2009-01-14:Comment/36285387
2015-06-24T21:58:43Z
2015-06-24T21:58:43Z
Report Calculations
<div><p>Hi all,</p>
<p>A couple more notes, though I'm still working on cleaning things
up.</p>
<p>sth, Moneydance does use IRR internally. So the reports should
match what Excel's XIRR function spits out. Feel free to email me
at ian @infinitekind.com if you have example cases that aren't
working in the current preview.</p>
<p>ljb, you're right that Moneydance should not report a number
when the equation fails to converge. I've flagged that to work into
a future update. Regarding issues with changes not sticking, could
you start a separate conversation on that? We've been stamping out
those bugs when they pop up and I'd like to work with you on that,
but I don't want to clutter up this discussion.</p>
<p>Also, I believe the Capital Gains report should now calculate
the average cost basis correctly as of build 1239, which should be
available on preview site shortly. This also fixes the gains
reported on the Investment Performance report. The Portfolio report
is still a bit off, but that's next on my list.</p>
<p>Ian<br>
Infinite Kind Support</p></div>
Ian L