Simulating "long" retirements vs. shorter ones.

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
lauren_knows
Posts: 193
Joined: Mon Jan 28, 2013 2:11 pm
Location: NoVA, USA
Contact:

Simulating "long" retirements vs. shorter ones.

Post by lauren_knows »

Bogleheads,

A lot of retirement calculator sites that deal with historical data, usually have data going back to 1871 for general market (stock/bond) and maybe 1927 for more diversified asset classes. That means if you simulate every 30 year period, you get 113 "cycles". Or, if you simulate a "longer" retirement of say 55 years, you only get 88 "cycles".

For trying to simulate "longer" retirements, at least for evaluating the worst-case scenarios, would it be worth it to simulate 10yr increments, the choose one of the lower portfolio values to start the next 10 yr increment, and so on? This would yield more "cycles", but I suppose it would start to look monte-carlo-ish.

Is this sort of analysis worth it, or is current historical data analysis and monte carlo simulations just fine?
41 - Married - 2 kids - Aiming for FI/ER in early 40s - Creator of cFIREsim
User avatar
Raybo
Posts: 2244
Joined: Tue Feb 20, 2007 10:02 am
Location: San Francisco
Contact:

Re: Simulating "long" retirements vs. shorter ones.

Post by Raybo »

I'd start with spreadsheets.

Estimate your expenses, choose an inflation rate, interest rate, and investment return, start with your retirement assets, and move the data forward a year until to get the length of time you wish to estimate. By varying the unknowns, you can get a good idea of how robust your plan is.

I used spreadsheet when I was working, but wrote my own full-blown financial package that includes a Monte Carlo simulation of my withdrawals.

While a spreadsheet isn't firecalc, the future can't be predicted from the past, either.
No matter how long the hill, if you keep pedaling you'll eventually get up to the top.
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Simulating "long" retirements vs. shorter ones.

Post by staythecourse »

Raybo wrote:I'd start with spreadsheets.

Estimate your expenses, choose an inflation rate, interest rate, and investment return, start with your retirement assets, and move the data forward a year until to get the length of time you wish to estimate. By varying the unknowns, you can get a good idea of how robust your plan is.

I used spreadsheet when I was working, but wrote my own full-blown financial package that includes a Monte Carlo simulation of my withdrawals.

While a spreadsheet isn't firecalc, the future can't be predicted from the past, either.
Don't think it makes sense to randomly pick starting and end dates that don't roll.

In the end there is no assurance ANY portfolio will last. If we have hyperinflation as in Germany or Argentina no numbers are safe.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
technovelist
Posts: 3611
Joined: Wed Dec 30, 2009 8:02 pm

Re: Simulating "long" retirements vs. shorter ones.

Post by technovelist »

staythecourse wrote:
Raybo wrote:I'd start with spreadsheets.

Estimate your expenses, choose an inflation rate, interest rate, and investment return, start with your retirement assets, and move the data forward a year until to get the length of time you wish to estimate. By varying the unknowns, you can get a good idea of how robust your plan is.

I used spreadsheet when I was working, but wrote my own full-blown financial package that includes a Monte Carlo simulation of my withdrawals.

While a spreadsheet isn't firecalc, the future can't be predicted from the past, either.
Don't think it makes sense to randomly pick starting and end dates that don't roll.

In the end there is no assurance ANY portfolio will last. If we have hyperinflation as in Germany or Argentina no numbers are safe.

Good luck.
No, but there are investments that are likely to survive even those "tail-risk" events. :mrgreen:
In theory, theory and practice are identical. In practice, they often differ.
User avatar
Topic Author
lauren_knows
Posts: 193
Joined: Mon Jan 28, 2013 2:11 pm
Location: NoVA, USA
Contact:

Re: Simulating "long" retirements vs. shorter ones.

Post by lauren_knows »

staythecourse wrote: In the end there is no assurance ANY portfolio will last. If we have hyperinflation as in Germany or Argentina no numbers are safe.

Good luck.
Obviously this is true, however if we're talking about looking at a situation that will be "no worse than any time in America", then it can be useful to simulate things like this to get a warm and fuzzy feeling.
41 - Married - 2 kids - Aiming for FI/ER in early 40s - Creator of cFIREsim
Post Reply