ESR - Am I calculating this correctly?

[intent]

I'm still trying to learn how to use the FireCalc tool and would greatly appreciate it if anyone with more experience could double check my assumptions/input.

What I want to do is get an idea of how much my wife and I would need to earn if we decided to stop working full-time in the very near future and continue to work part-time for several more years. To try to calculate this, I have entered the following into FireCalc:

* Retirement length = 60 years (I'm 42 and my wife is 40)
* Spending = $40,000
* Portfolio = $150,000
* Retirement date = 2013 and input that we would save 25K/year until then

Other Income/Spending:
* 12K/yr SS for me in 2029
* 12K/yr SS for wife in 2032
* Off chart spending reduction of $50K/yr starting in 2013
* Off chart spending increase of $50K/yr starting in 2025

So, we currently make about $100K/yr but what I want to determine is, if we reduced that by half then fully retired in 2025 would our portfolio last. With these numbers, I'm getting a 95% success rate, but I'm not convinced I'm using the tool correctly.

Appreciate the insight of those far wiser than myself.

 

[BigNick]

* Retirement length = 60 years (I'm 42 and my wife is 40)

Good luck with that... seriously, you can plan for one of you to go before then, or at least, to be spending a very different amount of money. It might be a lot less, or a lot more, depending on how demanding you are about nursing...

* Spending = $40,000
* Portfolio = $150,000
* Retirement date = 2013 and input that we would save 25K/year until then

With only $200-$220K in the bank, you're going to need a heck of a rate of return, unless your part-time income exceeds your spending.

Other Income/Spending:
* Off chart spending reduction of $50K/yr starting in 2013

Did you really mean that? Combined with your initial spending of $40K, that means you will be spending $-10K. No wonder the scenarios were successful most of the time.

* Off chart spending increase of $50K/yr starting in 2025

What does this correspond to?

 

[ziggy29]

Did you really mean that? Combined with your initial spending of $40K, that means you will be spending $-10K. No wonder the scenarios were successful most of the time.

What does this correspond to?

I think the real question he's asking here is: How the heck can I model "semi-retirement" (i.e. a reduced income that ends at some point in time) in FIRECalc?

 

[intent]

Thanks ziggy for reformatting my question in a more precise way - that is exactly what I am trying to do.

What I want to do is use the FireCalc tool, if possible, to estimate how my portfolio might perform if my wife and I were to stop contributing in a couple of years and reduce our income by half. To try to accomplish this, I used the off chart spending options, but I'm really unclear as to whether or not I'm using the tool in a way that was never intended and therefore might not be receiving very useful results.

As for the 60-year time span, well, we can all dream

 

[brewer12345]

I think the real question he's asking here is: How the heck can I model "semi-retirement" (i.e. a reduced income that ends at some point in time) in FIRECalc?

Exactly. I am grappling with a similar question. The simplification I have chosen is to model the future as if I have only my retirment account assets (about 2/3 of total) to depend upon for eventual full retirement. That way I can play with a number of potential scenarios with asset levels and periods of time to retirement and see what levels of "safety" different scenarios afford. In using this modelling approach, I am impicitly assuming that we consume all after tax assets in the next 10 years (in addition to whatever we generate by working). This is hopefully excessively conservative. Any other notions about modelling ESR would be appreciated.

 

[ziggy29]

I guess one way that it would be possible to model semi-retirement is if the income streams (like pensions and SS) had an optional *end* date. So if I semi-retired in 2015 at age 50 and retired fully in 2025 at age 60, I could enter retirement in 2015 *and* add a part-time job income stream at $X per year from 2015 to 2025. Right now the options there assume the income stream stays with you until death.

I suppose one could model a $20,000 annual part-time job in the example above by creating a $20,000 lifetime income stream starting in 2015, accompanied by an increase of expenses for $20,000 starting in 2025 to offset the temporary income stream, but I haven't really played with it that way. Though I will say that ESR within the next 5-7 years is looking a lot more likely than it did a year or two ago.

 

[ERD50]

I think you are doing it correctly. It does look strange, but that is just the way FireCalc handles these things.

To answer BigNick's Q about this one:

* Off chart spending increase of $50K/yr starting in 2025

That just offsets the...

Off chart spending reduction of $50K/yr starting in 2013

IOW: $50K spending reduction is really $50K in added income starting 2013. It 'goes away' (gets canceled out) by adding in a $50K spending increase in 2025. So, earning $50K from 2013 to 2025.


I have not played with the 'years before retiring' settings though. I always ran it as 'retiring at this time' view.

One thing I've always been confused on is how it handles adjusting a future off-chart spending amount. Does the $50,000 in 2025 get boosted by inflation, or it it a straight $50K reduction at that time. Never figured that out.

Another way to state it is - he is 42 and will be working part time for another 15 years, so will 'really' retire at 57 YO. You might want to break it into two sections as a double check - estimate where you expect to be at 57, then run a straight retirement scenario at that age.

-ERD50

 

[intent]

I suppose one could model a $20,000 annual part-time job in the example above by creating a $20,000 lifetime income stream starting in 2015, accompanied by an increase of expenses for $20,000 starting in 2025 to offset the temporary income stream, but I haven't really played with it that way.

The only other option I could think of would have been to enter a retirement date of 2025 and input $3,333 per year as the amount I will add to the portfolio until then ($3,333 x 15 = $50k), but this seemed like it would be much less accurate, as saving $50k over two years is much different than saving $50k over 15 years.

I think, in theory, the inputs should work, but this sure isn't a theory I'd want to discover the fault in *after* we quit

 

[intent]

You might want to break it into two sections as a double check - estimate where you expect to be at 57, then run a straight retirement scenario at that age.

Great idea ERD50, I'll give that a try.

 

[intent]

** Reporting back - weird results **

So, I pretended I was currently 57 years old, input and re-ran the numbers. I assumed a 6% annual return on our current portfolio to arrive at an approx. portfolio balance of $525K in 15 years. Here are the numbers I entered:

* Retirement length = 40 years (assumption is I'm 57 and my wife 55)
* Spending = $40,000
* Portfolio = $525,000

Other Income/Spending:
* 12K/yr SS for me in 2014
* 12K/yr SS for wife in 2017 (the real dates to collect when we are 62 are 2029 and 2032. As I will turn 58 in 2025, I simply added 4 and 7 to the current year to come up with these dates. Imprecise to be sure.)

This came out to an 81% success rate - significantly lower than the 95% success rate in my first post. However, if I change the off chart reduction/increases to $40K in the first calc, I end up with an 81.3% success rate - so very close indeed.

With that in mind, my initial reaction is to conclude that the way I used FireCalc in the first test was sound (thanks for the suggestion ERD50).

But wait, there's more weirdness...

In this second round of testing, I also tested a 45-year retirement period and a 50-year retirement period (no chance we'll live that long, but better safe than sorry). Oddly, the success percentage increased the more I increased the retirement length. At 45 years the success rate jumped to 82.1% and at 50 years it jumped to 83.3%. That seems counterintuitive to me, so now I'm re-questioning the tool.

 

[ERD50]

. Oddly, the success percentage increased the more I increased the retirement length. At 45 years the success rate jumped to 82.1% and at 50 years it jumped to 83.3%. That seems counterintuitive to me, so now I'm re-questioning the tool.

Yes, this can happen too. It is disturbing and counterintuitive, but it does make sense when you consider how FIRECALC works.

If you look at say, a 10 year retirement length, it can include and analyze a cycle that includes 1999-2009. But then it stops, because starting at 2000 it would have to go to 2010, and we don't have that data.

So when you look at a 40 year retirement length, it has to stop at 1969-2009. So it is possible for a 10 year cycle to include a bunch of bad cycles starting in the 70's, 80's, and 90's that just cannot be accounted for in a 40 year cycle.

Trust me, there is nothing 'warm and fuzzy' about ER. Keep running the numbers, do all you can to assure some level of comfort, try to keep some options open, but at some point, it is a giant leap of faith.

-ERD50