Sensitivity Analysis - Basic Questions

Post questions about how to use the planner, user inputs, how the planner works, and comments and suggestions.
Post Reply
stevecle
Posts: 2
Joined: Fri Dec 18, 2020 3:39 pm

Sensitivity Analysis - Basic Questions

Post by stevecle »

First, I really like the FRP program - It's much better (accurate & flexible) than the others I've tried!

I have a basic question regarding the Sensitivity Analysis, specifically the "Portfolio Return" parameter... What does the "Min" and "Max" represent? To take an easy example, let's say one was 100% invested in the SP500. Historically, the minimum / maximum annual returns were ~40% and 60%, respectively, but the "min" amount entered into the Sensitivity Analysis cannot be negative. Therefore, my basic question is what do the "Min" and "Max" represent?

Also, what are the suggested parameters for the St Dev "Min" and "Max"? Historically, the SP500 standard deviation is apx 20%. Realizing that we can't predict the future, what would be some suggested "reasonable" tests? For example, would a min of 10% and max of 30% standard deviation be an obvious and reasonable test?

Many "THANK YOUs" for a fine product! ~ Steve
jimr
Posts: 821
Joined: Thu Feb 28, 2008 6:48 pm

Re: Sensitivity Analysis - Basic Questions

Post by jimr »

In general, with sensitivity analysis, the min and max for parameter 1 and parameter 2 are bound the range of possible inputs for the given parameter.

What sensitivity analysis is trying to do is run the simulation a ton of times with different values for the parameter 1 and parameter 2 inputs each time. The idea is that by doing this you can get an idea of how sensitive your plan's success is to changes in your plans inputs (eg parameter 1 and parameter 2).

The heatmap x-axis plots what happens to the plan's success rate as parameter 1 is varied and the heatmap y-axis plots what happens as parameter 2 is varied. So the bottom left cell of the heatmap shows the probability of success when the simulation is run with parameter 1 and parameter 2 both set at their minimum values. The top right cell of the heapmap shows the outcome with parameter 1 and parameter 2 both set at their max values.

You can click on cells in the heatmap and the right-hand panel shows the data for the run corresponding to the clicked cell (you can see for yourself what values were used for parameter 1 and parameter 2 for the cell (run) you clicked on).

So with parameter 1 set to return/std dev, the min and max determine the range of possibilities that sensitivity analysis tests out. This isn't intended to be the range of possible returns you might see from year-to-year because these values are inputs to the simulation that are used to generate the sequence of returns over each retirement path that gets simulated.

I don't give financial planning advice here, but one thing this could be used for is to help gauge your need to take risk to have a successful retirement. Usually a more aggressive a portfolio will have a higher standard deviation (more volatility). So with the default min/max vals (eg 3-12% for return, 2-18% std dev) you can get an idea of how successful your plan might be with different portfolios that have differing levels of risk.

One decent source of information about historical returns for a given portfolio is something called Simba's Backtesting Spreadsheet from bogleheads.org. This spreadsheet lets you input a portfolio mix and it spits out the historical average return and standard deviation.
https://www.bogleheads.org/wiki/Simba%2 ... preadsheet

This is just one source of information on this, and you'll probably want to do more research and not just rely on one source, but it's a good start.
stevecle
Posts: 2
Joined: Fri Dec 18, 2020 3:39 pm

Re: Sensitivity Analysis - Basic Questions

Post by stevecle »

Thanks so much for the complete answer! Got it! And I have been working with the Simba spreadsheet. ~ Best Regards
Post Reply

Who is online

Users browsing this forum: Ahrefs [Bot] and 23 guests