### Author Topic: Stock market Monte Carlo simulation spreadsheet  (Read 1317 times)

#### SeattleCPA

• Handlebar Stache
• Posts: 1308
• Age: 59
• Location: Redmond, WA
##### Stock market Monte Carlo simulation spreadsheet
« on: September 25, 2017, 06:57:52 AM »
The Unreliability of long run stock market returns discussion included a simple stock market Monte Carlo simulation spreadsheet. But it looked at 100-year long accumulations.

I thought it might be useful to create another simple but more applicable simulation spreadsheet... so did that. Anyone interested can grab that spreadsheet here:

As title indicates, the simulation looks at 40 year long accumulations. To keep the spreadsheet to a manageable size, it only recalculates ten scenarios at a time but one can press the F9 (recalculate) key several times to see a bunch of iterations.

To use the spreadsheet, you enter a standard deviation, nominal return and initial deposit into the worksheet:

The spreadsheet calculates some familiar values:

And then most useful, probably, it draws a line chart of the scenarios to show you the ways things could turn out:

The spreadsheet is very simple. But if someone enters a decent long-run estimate of their portfolio's nominal (so unadjusted for inflation) return and their portfolio's standard deviation, they can look at the statistics and especially at the chart and get a gut-level feel for the sort of variability they could easily see.

Note: The initial inputs and the example spreadsheet values show a 10% nominal return and 10% standard deviation. These are the FIRECalc default values for doing a simulation. But I think 9% and 9% are probably better if one models the classic 60% stocks and 40% bonds asset allocation. I would think looking at a 100% stocks portfolio would mean something more like 10% return and 15% standard deviation.

Tip: The thing to really play with if someone wants to noodle around here is that standard deviation value. We are all pretty good, I think, at thinking about the annual return value. But many of us (me included) seem to sometimes kind of "glance over" the standard deviation.

P.S. If someone is interested, I did describe the steps to create this spreadsheet at my blog, link appears in sig.

#### Telecaster

• Handlebar Stache
• Posts: 1550
• Location: Seattle, WA
##### Re: Stock market Monte Carlo simulation spreadsheet
« Reply #1 on: September 26, 2017, 12:26:33 PM »
Cool!  People tend to ignore the std deviation as you say, but they also (me included) tend to think in terms of the average return.  Median return is probably better for back of the envelop calculations.

#### SeattleCPA

• Handlebar Stache
• Posts: 1308
• Age: 59
• Location: Redmond, WA
##### Re: Stock market Monte Carlo simulation spreadsheet
« Reply #2 on: September 26, 2017, 01:55:17 PM »
Cool!  People tend to ignore the std deviation as you say, but they also (me included) tend to think in terms of the average return.  Median return is probably better for back of the envelop calculations.

Agreed.

And just to clear up a couple of confusions I'm responsible for...If you use the spreadsheet, you want to enter the nominal return as the arithmetic mean (since that's the value you want to randomly adjust using the standard deviation).

And then that median value and return the spreadsheet calculates should have been labeled something more like median future value and compound average geometric return.