One of the most often requested "features" of cFIREsim is that we change how the yearly calculations are done in regards to spending.

As it stands, "spending" for a given year is taken out at the end of the year, where all the market gains and other calculations are done. This doesn't necessarily jive with what people think is reality. People seem to think that retirees are more likely to take a chunk of money out at the beginning of the year, and spend it throughout the year.

We can certainly make that change, but it brings up a question of inflation-adjustment. If we take spending out on January 1, and then do all of the market gain and portfolio calculations on Dec 31... which inflation numbers do we use for spending calculations? It seems like using the inflation numbers for the year that we're subtracting it from doesn't make sense, since the inflation numbers represent the entire year of inflation.

Do we use the year before?

## Question on inflation

### Question on inflation

-Bo (Creator and Admin for cFIREsim)

### Re: Question on inflation

That makes sense to me. I think that models the most realistic scenario where a retiree makes an annual budget on January 1 using all available data.

### Re: Question on inflation

+1 your reasoning makes sense to me. If we're working with inflation data with a resolution of 1 year then the inflation multiplier on Jan 1st 1990 is equal to the inflation multiplier on Dec 31st 1989 (for example).

### Re: Question on inflation

Such a change (do the math at the beginning of the end) would be VERY welcome. This is kind of arbitrary, but just so much more intuitive... And yes, sure, inflation adjustments are then derived from the past year(s), not the future. That is, except for the very first withdrawal, of course.

### Re: Question on inflation

I know this forum isn't exactly booming with visitors, but I wanted to put this out there.

As I'm coding, I realize that if we take spending out on January 1 (beginning of a year)... but do all of the Market Gain calculations on Dec 31st (end of a year), then should we be applying inflation differently to spending vs. market gains? If I want to show an inflation-adjusted spending value and an inflation-adjusted ending portfolio value for a given year... things would be pretty weird pretty fast if I used 2 different values of inflation.

The easiest solution is to stick with the idea of using inflation numbers from the year before. This, however, would always make Year 1 of a simulation have 0 inflation. It effectively pushes the cumulative inflation numbers out one year into the future (than currently calculated in cFIREsim).

Is that acceptable? I'm not sure using 2 different inflation numbers for a particular year is practical.

As I'm coding, I realize that if we take spending out on January 1 (beginning of a year)... but do all of the Market Gain calculations on Dec 31st (end of a year), then should we be applying inflation differently to spending vs. market gains? If I want to show an inflation-adjusted spending value and an inflation-adjusted ending portfolio value for a given year... things would be pretty weird pretty fast if I used 2 different values of inflation.

The easiest solution is to stick with the idea of using inflation numbers from the year before. This, however, would always make Year 1 of a simulation have 0 inflation. It effectively pushes the cumulative inflation numbers out one year into the future (than currently calculated in cFIREsim).

Is that acceptable? I'm not sure using 2 different inflation numbers for a particular year is practical.

-Bo (Creator and Admin for cFIREsim)

### Re: Question on inflation

Not sure to understand why you have headaches with this. Think of it like a retiree using a piece of paper. He's not going to do half of the math on Dec 31st and the other half on January. He'll do it all on Jan 1st. And for sure, he'll use a single inflation value (i.e. previous year). And for sure, he will not know the inflation of the year to come. And yes, for the very first Jan 1st, there is no inflation whatsoever (ok, I'm assuming said retiree retired on this precise day, just to simplify!).

Personally, in my own Excel spreadsheet, I compute a cumulative inflation factor (starts from 0, gets adjusted year over year), and then I use this cumulative factor to adjust the nominal math to real math. This makes everything consistent, even for the very first year.

Did I miss your real question? Not too sure I truly got your point.

Personally, in my own Excel spreadsheet, I compute a cumulative inflation factor (starts from 0, gets adjusted year over year), and then I use this cumulative factor to adjust the nominal math to real math. This makes everything consistent, even for the very first year.

Did I miss your real question? Not too sure I truly got your point.

Return to “cFIREsim Open Source Project”

### Who is online

Users browsing this forum: No registered users and 1 guest