Question about FIRECALC calculator

[FREE866]

Under the "spending models" tab
Use the following inflation assumption: PPI, CPI, or % for inflation adjustments to the historical data.

If you check the CPI button what is the assumed inflation rate?
I checked the last button and did 3%, but when I checked the CPI button my success ratio went up..any reason why that is and which is the suggested button to press?
thx

 

[GalaxyBoy]

CPI = Consumer Price Index (wikipedia link)
PPI = Producer Price Index (wikipedia link)

 

[FREE866]

CPI = Consumer Price Index (wikipedia link)
PPI = Producer Price Index (wikipedia link)

I know what their definitions are , but what I'm asking is what is the number that is applied when I select that button?

 

[COcheesehead]

It doesn't disclose the number, but I would assume it is the historic data for CPI and PPI over the same time frame as the market results it uses as the basis of the analysis.

 

[MBSC]

On the 'Investigate' tab, check the box that says optionally provide data and formulas in a spreadsheet. On the results page, open the second spreadsheet that says show the year by year inputs, data, etc. There is a column called 'inflation factor'.

Note that the default settings are a 30 year retirement starting in 1960 (year associated with the checkbox). This displays the inflation inputs for 1960-1989. You need to manipulate these values to see inflation for other time periods.

 

[ERD50]

It doesn't disclose the number, but I would assume it is the historic data for CPI and PPI over the same time frame as the market results it uses as the basis of the analysis.

Right, it's the actual historical CPI or PPI, year by year, not any fixed value.

So a 10% market gain in a 10% inflation year is a net 0% gain in purchasing power. And a 10% market gain in a 0% inflation year is a net 10% gain in purchasing power.

To me, it makes no sense to use fixed numbers when we have historical data available.

-ERD50

 

[FREE866]

Right, it's the actual historical CPI or PPI, year by year, not any fixed value.

So a 10% market gain in a 10% inflation year is a net 0% gain in purchasing power. And a 10% market gain in a 0% inflation year is a net 10% gain in purchasing power.

To me, it makes no sense to use fixed numbers when we have historical data available.

-ERD50

I agree. But if historically inflation averages 3%
Then why would the success ratio go down when you opt for the manual button of inserting3%?

 

[razztazz]

I agree. But if historically inflation averages 3%
Then why would the success ratio go down when you opt for the manual button of inserting3%?

Because you are not concomitantly averaging portfolio performance and standard deviation

 

[FREE866]

Because you are not concomitantly averaging portfolio performance and standard deviation

Bottom line. What is suggested on that page?
Just click
CPI? Or manually inserting a number?

 

[ERD50]

I agree. But if historically inflation averages 3%
Then why would the success ratio go down when you opt for the manual button of inserting3%?

Are you familiar with the story of the 6 foot tall statistician who drowned standing up in a pool with an average depth of 3 feet?

The answer is in that story.

-ERD50

 

[FREE866]

Are you familiar with the story of the 6 foot tall statistician who drowned standing up in a pool with an average depth of 3 feet?

The answer is in that story.

-ERD50

you totally lost me lol

 

[GalaxyBoy]

I don't think there's a definitive answer to the question of which inflation model is "right," except in hindsight. I run my numbers using a variety of assumptions, using different retirement calculators that use different methodologies. Bear in mind they are all just tools for deciding how much money is "enough," and none offer a crystal ball prediction of the future. However, taken together they should provide guidance in making the qualitative decision as to when to jump out of that airplane and trust the parachute.

So, I ran my numbers in firecalc using each of the inflation assumptions, and again using different straight-line inflation values to get a feel of how much high inflation rates could impact my odds. I ran cases with reduced SS benefits for the same reason. I also use i-ORP.com, Fido's RIP, Quicken, Schwab, etc. They all gave fairly similar values for retirement spending levels. I have a planned retirement spending budget created from the ground up, and since it's about $20k/year less than what the crystal balls see, I feel like I'm ready.

 

[exnavynuke]

you totally lost me lol

You have little chance of drowning in a pool that is 3 ft deep all around, which is the statistical equivalent of just using a 3% "average" for every year. However, if 20% of the pool is 10 ft deep and the rest is as shallow as 1" deep, then there's a much higher chance that someone could die in the pool. That variation in depth from the "average" is what is not taken into account if you just use "average numbers" always. How often, and by how much, there is a deviation from that average can significantly change the results. Thus using historical numbers that include deviation from the average can make the results show differently than if you just use a given % each year.

 

[FREE866]

You have little chance of drowning in a pool that is 3 ft deep all around, which is the statistical equivalent of just using a 3% "average" for every year. However, if 20% of the pool is 10 ft deep and the rest is as shallow as 1" deep, then there's a much higher chance that someone could die in the pool. That variation in depth from the "average" is what is not taken into account if you just use "average numbers" always. How often, and by how much, there is a deviation from that average can significantly change the results. Thus using historical numbers that include deviation from the average can make the results show differently than if you just use a given % each year.

Gotcha
So,by that logic it seems that one should use CPI as opposed to inserting 3% manually
Is that a correct conclusion?

 

[REWahoo]

That's what I use when running FIRECalc.

 

[FREE866]

That's what I use when running FIRECalc.

Nice. wow..using that puts me at 100%!