Dory36 Calculator Workaround

Financial Independence/Retire Early -- Learn How!
Post Reply
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Dory36 Calculator Workaround

Post by JWR1945 »

The dory36 calculator has been updated to make things easier. Both the new and old versions are available at this time.

The dory36 calculator has a bug when making Withdrawal Change 2 (and also Withdrawal Change 3). I can duplicate my earlier results by removing the check mark for using inflation adjusted withdrawal amounts (for Withdrawal Change 2). All year by year dollar amounts are identical to those that I had previously calculated and reported in FIRE 50 Year Dual Portfolios.

I get the same results using both versions of the calculator. I can reproduce my original results with a two step procedure. Doing so is tedious. Again, all dollar amounts are identically equal to my previous results. They are not merely close.

I have contacted Captain Bill. He asked for more information and it was his comments that caused me to look at that check mark. He identifies the change as a bug. The original calculations were correct. Of course, that conclusion is subject to change. He may uncover something else.

Have fun.

John R.
raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Re: Dory36 Calculator Workaround

Post by raddr »

JWR1945 wrote: I can duplicate my earlier results by removing the check mark for using inflation adjusted withdrawal amounts (for Withdrawal Change 2). All year by year dollar amounts are identical to those that I had previously calculated and reported in FIRE 50 Year Dual Portfolios.

I get the same results using both versions of the calculator. I can reproduce my original results with a two step procedure. Doing so is tedious.
John R.


John,

I thought that all withdrawals were inflation adjusted, including withdrawal 2, no?

What is the 2 step procedure you use?

Thanks!
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Dory36 Calculator Workaround

Post by JWR1945 »

raddr, here is my latest email (with attachments) from Captain Bill. I hope that it has enough detail (from the emails that I wrote). When I first used the calculator, I got the "old" numbers with the inflation adjustment (check mark present) for Withdrawal Change 2.

Have fun.

John R.
........................
John,

Thanks for the detailed response.

There was a bug in the inflation section in withdrawal 2, that I thought I
had squashed as of the version posted Sunday morning, but maybe it has
friends. (The bug only appeared when withdrawal 2 inflation was checked.)

There is one remaining bug (to be fixed today, I hope) about the red lines
that omitted the red if there was enough optional INCOME entered into the
extra withdrawal fields or lump sum fields to bring the portfolio back into
the black after it had dipped into the red.

I'll try to walk through your info and see if I can isolate the bug.

Regards,

Bill

-----Original Message-----
From: Cox SMTP east [mailto:ranmbs@cox.net]
Sent: Monday, November 25, 2002 11:15 PM
To: Webmaster at Early-Retirement.org
Subject: Re: Suggestion Box-Fire2


Hello Bill,

I have written up the details. I will make extra sure tomorrow morning that
I have reported everything exactly as it is. The answer has something to do
with the inflation adjustment for Withdrawal Change 2. I get the old
answers if I remove the check mark. I get the old answers if I make a
manual calculation (actually, a manually assisted calculation). I get the
new answers if I include the check mark. All dollar amounts are identical
to my previous results.

Thanks. John Russell
.................
Dory36 bug (11-25-02)

1. I can get the old answers.exactly matching all dollar amounts for each
year.by removing the check mark that adjusts the Withdrawal Change 2 amount
for inflation.
2. I can get the new answers.exactly matching all dollar amounts for each
year.by including the check mark that adjusts the Withdrawal Change 2 amount
for inflation.
3. Both the old and the new calculator produce identical results.
4. When I use the manual procedure, I get the old results.
5. Inputs for basic calculations: amount = $1000, initial withdrawal
amount = zero, duration = 50 years, 80% stocks, 0.20% expenses, commercial
paper, CPI, no check mark for an initial withdrawal, Withdrawal Change 1 is
$40 starting in year 6 and it is adjusted for inflation, Withdrawal Change 2
is $40 starting in year 16 and it is adjusted for inflation (which produces
the new results). There are no other adjustments.
6. The basic calculations show portfolio failures in 1899, 1901-1907,
1909-1913, 1929, 1930, 1937, 1959-1969. At year 15 the balances are $1939
for 1899 and $1903 for 1901 and $1718 for 1902 and $1413 for 1903 and $1870
for 1904 and so forth. These failures are easy to identify. They are in
red.
7. If Withdrawal Change 2 is set to zero, the year 15 balances are all
identical.as they should be.but they are no longer in red.
8. For a manual calculation the balances at year 15 are used as starting
amounts. The initial withdrawal amount is set to $80. The duration is set
to 35 years. All Withdrawal Change amounts and other adjustments are set to
zero. We still use 80% stocks, 0.20% expenses, commercial paper, CPI, no
check mark for an initial withdrawal.
9. For 1899 the basic calculation shows a balance of $1939 at year 15.
Using this amount for the second calculation, the manually determined
balances are read from the 1914 line. (1899 + 15 = 1914) Starting with
year 1, the 1914 line reads $1796, $2164, $2251, $1852, $1970, $2122 and so
forth. They correspond to the amounts that should occur in years 16, 17,
18, 19, 20, 21 etc. along the 1899 when making the basic calculation
(including Withdrawal Change 2 as adjusted for inflation). They do not.
They are identical to the basic calculation when the inflation adjustment
check mark is removed.
10. A glance at the 1914 line shows that the portfolio succeeded according
to the manual calculation.
11. Continuing in this manner, the manual calculations indicate that all
portfolios through 1937 succeed except for 1929 and 1930. I did not check
any further.


----- Original Message -----
From: "Webmaster at Early-Retirement.org" <webmaster@early-retirement.org>
To: <ranmbs@cox.net>
Sent: Monday, November 25, 2002 6:48 PM
Subject: RE: Suggestion Box-Fire2


> Hi John,
>
> Would you mind trying your test again? I updated the computation module
> yesterday morning to correct an error in the inflation calculation in
> withdrwal 2. I hope it corrected the problem.
>
> If the problem is still there, could you send me more details of your test
> so I can try and isolate the bug?
>
> Thanks -- Bill
>
> -----Original Message-----
> From: ranmbs@cox.net [mailto:ranmbs@cox.net]
> Sent: Monday, November 25, 2002 9:32 AM
> To: suggestions@early-retirement.org
> Subject: Suggestion Box-Fire2
>
>
> >From John Russell:
>
> Good morning. I hope that you find that this is helpful.
>
> I have placed these two posts at
> http://groups.msn.com/IndexFunds/messageboard.msnw. I have included the
> second (earlier)post in case you need more details. As you can see, I
> hold the dory36 calculator in high regard.
>
> Thanks. Have fun.
>
> John Russell JWR1945
> ............
>
> FIRE Calculator Problems (11-24-02)
>
> There is a problem with the dory36 calculator at this time. The site has
> been updated. It still has the old version available. The problem
> happens on both versions.
>
> So long as you make a single change or no change, the numbers are the same
> as before. But when you put in a second change, the answers are different
> and they are wrong.
>
> I used a manual procedure to isolate the problem. It confirms that our
> previous answers are the right ones.
>
> Details
>
> This is how I verified that the second change produces wrong answers.
> First, I put in a test condition with two changes. (It was H3.) The
> initial withdrawals were zero. I made a withdrawal change of $40 at year
> 6. I made a second withdrawal change of $40 at year 16. The duration was
> 50 years. I removed the checkmark for an initial withdrawal. There were
> many years for starting a retirement that failed. I recorded the balances
> on years 15 and 16. I also recorded a couple of longer sequences.
>
> Next, I ran the same condition without the second withdrawal change.
> There were no withdrawals initially. At year 6, the withdrawals were
> changed to $40 per year. All of the results were identical up to year 15.
> Year 16 was different. It was supposed to be. I recorded the balances at
> year 15 for those years that failed up to year 1937.
>
> Finally, I ran a series of conditions, each with a 35-year duration, an
> $80 annual withdrawal and no withdrawal changes. In each case I used the
> balance from year 15 as the initial amount. I then looked at the results
> starting 15 years later. For example, initially the year 1899 had failed.
> In both of the earlier runs, the balance at year 15 was $1939. I made a
> run with an initial amount of $1939 (and a 35-year duration, an $80 per
> year withdrawal amount and no withdrawal changes). I then looked at the
> results starting in 1914 (1899 + 15 = 1914). That portfolio succeeded. I
> looked for some obvious reason that the result was different but I did not
> see one.
>
> I continued using this manual procedure long enough to show that there
> were no failures through 1937 except for 1929.
>
wanderer
*** Veteran
Posts: 363
Joined: Tue Nov 26, 2002 9:33 am
Location: anytown, usa

Post by wanderer »

christ i hope this goes thru...
regards,

wanderer

The field has eyes / the wood has ears / I will see / be silent and hear
Post Reply