Questions on FIREcalc input

[Telly]

Hi,

I have been playing with FireCalc, and other calculators too. I like FireCalc's inclusion of Soc. Sec. into the mix. Here are my questions:

Quest. #1:
For S.S., when it says "uses inflation adjusted dollars", how should I interpret that?

For example, if I knew that my annual S.S. benefit in TODAY'S dollars was $10,000, but would not actually draw it till year 12, what does it do? Does it take the $10,000 TODAY, adjust UPWARDS for whatever inflation was in the respective 12 year period in the model, and use that as the starting amount for drawing S.S? Does it do that?

Does it then also adjust it upwards each year that the benefit is taken (years 12 - and on), at 100% of the inflation index for each successive year taken?


Quest. #2
Withdrawal Change, Inflation-Adjusted or not.
I understand that if I have a defined-Benefit Pension that starts in 5 years, I would enter the amount, and year 5, and turn OFF the inflation adjustment.

But what about a predicted reduction in income needed in the future? Say I projected that in 5 years that my income needs would drop by 10,000 of TODAY's dollars. Do I input $10,000, and 5 years, and turn ON Inflation-Adjusted?
I would want the $10,000 reduction (adjusted upwards to something greater for the inflation to year 5) to occur in year 5, and in year 6, the income start to increase inflation-adjusted from the new base established in year 5. Will all that happen by turning ON the Inflation-Adjustment for that FireCalc Withdrawal Change?

Thanks for any help given

 

[dory36]

Quest. #1:
For S.S., when it says "uses inflation adjusted dollars", how should I interpret that?  

For example, if I knew that my annual S.S. benefit in TODAY'S dollars was $10,000, but would not actually draw it till year 12, what does it do?  Does it take the $10,000 TODAY, adjust UPWARDS for whatever inflation was in the respective 12 year period in the model, and use that as the starting amount for drawing S.S?  Does it do that?

Does it then also adjust it upwards each year that the benefit is taken (years 12 - and on), at 100% of the inflation index for each successive year taken?

Yes and yes.

Quest. #2
Withdrawal Change, Inflation-Adjusted or not.
I understand that if I have a defined-Benefit Pension that starts in 5 years, I would enter the amount, and year 5, and turn OFF the inflation adjustment.

But what about a predicted reduction in income needed in the future? Say I projected that in 5 years that my income needs would drop by 10,000 of TODAY's dollars.  Do I input $10,000, and 5 years, and turn ON Inflation-Adjusted?  
I would want the $10,000 reduction (adjusted upwards to something greater for the inflation to year 5) to occur in year 5, and in year 6, the income start to increase inflation-adjusted from the new base established in year 5.  Will all that happen by turning ON the Inflation-Adjustment for that FireCalc Withdrawal Change?

Yes again. Put the fixed pension in one of the income adjustment options, with inflation turned off. Put the change in spending in another, with inflation turned on.

All inflation adjustments use whichever inflation adjustment you select (PPI or CPI).

Dory36

 

[Telly]

Thanks for the help

I have been using 40 and 45 years as the length of the retirement period (I sure don't expect to live that long, but...). But due to the long period, simulations using a start year within 40 to 45 years of today cannot complete.
So in effect, even though the records and simulation input go back ~130 years, my situation is not actually using 30-35% of the records. I don't have a solution for that obviously, but it seems by omitting dog periods during those years, the result may be brighter than it really would be.

Comments or suggestions?

 

[dory36]

I suggest you not change the start year.

Firecalc uses historical performance to create a set of possibilities of what might happen in the future. If you limit that set in any way, you reduce the range of possibilities.

Firecalc's message is,

If you take $x from a portfolio of $y for z years, you'll be safe as long as the next z years is no worse than the worst z year period in history.

If you change the start year, the message becomes

If you take $x from a portfolio of $y for z years, you'll be safe as long as the next z years is no worse than some fairly recent z year periods in history.

Changing the start year was only added as an option for a few folks who wanted to explore some theories. I should take it out.


Dory36

 

[Telly]

Oops, sorry, I wasn't clear enough in what I was saying I wasn't hard-coding start dates on anything, I was letting the simulation run the normal way.

For discussion, if we just use the default input values ($30,000 Withdrawal, 30 Years, 650,000 dollars, etc.) and looking at the spreadsheet created, the last totally complete period starts at 1971 (so it ends at 2001). For the row starting in 1972, the last year (2002) is empty. For the row starting in 1973, the last two years are empty, etc. etc. continuing up to the present.

So starting out in a bad year like 1973, for example, is missed, although 1973 is included as an intermediate or ending year of rows starting 30 or less years previous. But we know that starting out in a bad year should have a much more negative effect than having a bad year in the middle or the end.

When I put in 40 or 45 years, the effect grows.

Are there any suggestions of what I could do with the data of incomplete rows into the present? I can't really create future yearly data to complete them. But it seems I am just losing a lot of start and early years for data. For a 45 year period, the latest start row that completes is 1956.

Is there anything I can mine out of the incomplete rows?

I did have a horrible idea... If there was, say, five years incomplete on the end of a row, take the last useful cell's data from that row, then successively feed it into 1871 for 5 years on, then into 1872 for 5 years on, etc. etc. In effect, re-using existing data successively for the missing data. And on and on for all of the rows that did not complete. My gut feel is that the size of the thing would become monstrous

I'm interested in any and all comments!

 

[area52]

I think you might be on to something. If the FIRECalc program is a locomotive, doing a partial loop-back like you suggest would be like using one of those little "hand-cars": Good enough to finish the trip with. And probably better than one of those Monte Carlo simulators

I just tried it. I use a 50 year lifespan for myself. 1966 (the worst starting year so far) is still "alive" out to 35 yrs. THEN I took my next annual withdrawl and the remaining bucks and plugged them into the calculator and ran out a 15 year lifespan. Gets a little tight towards the end. I'll need to die 5 yrs sooner than planned if I don't want to run out of money.

But that was the good news. While it was only 84% safe most of short-falls came right in the last 1 or 2 yrs
I figure I can probably just spend my emergency bank roll for a year or 2.

All in all not a disasterous outcome and the concept seems to work

 

[area52]

By the way.... my starting withdrawl rate was a sneeze LESS than 3.5% of the starting portfolio value.

This might make some people feel bit uncomfortable. I know makes ME uncomfortable. I thought 3.5% would be pretty much sacrosanct ! But then again we don't KNOW what the next REAL LIFE 15 yrs will be like. It might be worse or it might be quite congenial.

I'll keep my "hari kari" knife available just in case

 

[dory36]

...
I did have a horrible idea...  If there was, say, five years incomplete on the end of a row, take the last useful cell's data from that row, then successively feed it into 1871 for 5 years on, then into 1872 for 5 years on, etc. etc.  In effect, re-using existing data successively for the missing data.  And on and on for all of the rows that did not complete.  My gut feel is that the size of the thing would become monstrous

I'm interested in any and all comments!

Hmm -- that might not be so horrible after all, but I will have to look at it more closely the next time I embark on a rewrite.

Originally, the spreadsheets at REHP and FIRECalc just ignored anything after the start of the last full term. When I saw the lousy results of terms starting in the 60's, I added the partial info reporting. But as you point out, this only looks at the partial period, not the full period.

Dory36

 

[Telly]

Thanks all for all of the comments both here, and in the "FIREcalc Rewrite Suggestions" post .

I don't think the basics of FIREcalc should be changed.  I am just trying to figure out what I could do with the incomplete rows, especially since I am most often running a 45 year period.  With a 45 year period, there are 85 rows that complete.  And 44 that do not complete due to colliding with the present.  I am also assuming that FIREcalc interprets an incomplete row that has the last complete cell > $0 as a success.  If so, the success is a little (or a lot!) premature.

My choices seem to be:

1) Make up data (I wouldn't know how or what, minds much more economics-minded than mine are probably toiling away at that!).

2) Fill in the missing future years by using past data, marched through in a cyclical fashion as I described before.

3) Some as yet unknown idea (I'm listening!).

area52, how did you do the re-input idea that you described?  I can look in the last cell to see what the fund value was, to plug it back in.  But where do I get what the inflated Withdrawal Amount would be for that last year's cell?  I would need to plug that in along with the fund amount.  I am only using the web-based FIREcalc here, and I haven't tried importing it yet.

I just can't seem to give up on the idea of using the last 44 or so years of data as starting points for a fund simulation :-/

 

[area52]

area52, how did you do the re-input idea that you described? I can look in the last cell to see what the fund value was, to plug it back in. But where do I get what the inflated Withdrawal Amount would be for that last year's cell? I would need to plug that in along with the fund amount. I am only using the web-based FIREcalc here, and I haven't tried importing it yet.


Just use one of these web sites to find the current, inflation adjusted withdrawl value for whatever year you want. They're just inflation calculators based on historical data

http://www.aier.org/colcalc.html

http://www.westegg.com/inflation/

http://146.142.4.24/cgi-bin/cpicalc.pl

Note: The results I spoke of last time were incorrect and better than they should have been due to trascription errors. The 50 yr failure rates were higher BUT still not a complete disaster.

 

[wzd]

Telly, etal,

I think that Dory acknowledged this in a different thread, but my suggestion for making up data is to make the end portfolio and inflation adjusted other input parameters available for partial periods, and then run a second (recursive) firecalc run using those values. In effect, this says that the missing years should be filled in with the historical averages. The question would be how to set the survival rate for the second (and possibly later) run.

This would not give year by year data for the made up years, but would allow it to be marked as a pass or fail year.

It might be possible to do this manually from the spreadsheet - I have not looked over the data fields enough to say for sure if all the inflation adjusted fields are available.

Wayne