Possible Defect: Sensitivity Analysis with Life Expectancy

Post questions about how to use the planner, user inputs, how the planner works, and comments and suggestions.
Post Reply
frp user
Posts: 65
Joined: Fri Feb 29, 2008 12:55 pm

Possible Defect: Sensitivity Analysis with Life Expectancy

Post by frp user »

Hello,

When using this great tool I ran into an issue running a sensitivity analysis using life expectancy as a parameter. I did not see anything on the website so decided to report this possible defect. Hopefully this is not a known issue already. I'm using version 4.01.02.

I found that a run of sensitivity analysis would "exit" without producing any results when there is an additional input cash flow entry with a start year > the life expectancy "min" AND an end year of "end of plan" (EOP).

The report works as expected with start year <= min and end year = EOP

The report also works with start year > min and end year some specific age, e.g. 100.

To duplicate, create an additional inputs cash flow that starts at age 71 and ends at end of plan, then run sensitivity analysis with Param 1 = Annual Retirement Spending, and Param 2 = Life Expectancy Min=70 Max=120

Dave

jimr
Posts: 614
Joined: Thu Feb 28, 2008 6:48 pm

Re: Possible Defect: Sensitivity Analysis with Life Expectan

Post by jimr »

Dave,

Thanks for the bug report and especially for the great writup. That made understanding and interpreting the issue much easier.

I plan to treat this as an input validation error, but first I want to verify that's what you'd expect. With my proposed change, clicking run sensitivity analysis in this case will result in a pop up error message with something like:
Sensitivity Analysis failed because of a conflicting or invalid entry in Additional Inputs.

Please check all enabled Additional Inputs entries for errors or conflicts with the currently configured Sensitivity Parameters.
I realize it's confusing that the error only happens on cash flows using EOP as the end date, and not cash flows with a fixed end. In truth, the fixed end date should probably also give an error because it's 'out of range' of the plan boundaries, but I don't currently have a check for 'end age > life expectancy'.

The reason using end of plan gives an error is that there's a 'start can't be > end' check and end of plan converts to age 70, which is less than the start age of 71. With a fixed end instead of end of plan, the end date is out past the end of plan, but the 'start > end' check is ok, since the fixed end is greater than the start.

Anyhow, does the message make sense and if you saw it, do you think it would have helped you track down the issue sooner?

Thanks again,

Jim

frp user
Posts: 65
Joined: Fri Feb 29, 2008 12:55 pm

Re: Possible Defect: Sensitivity Analysis with Life Expectan

Post by frp user »

Jim,

The error message does provide some additional pointers to the possible error, "Sensitivity Parameters," so I think that would have helped. This approach seems reasonable given the current/past end user experience with your tool.

jimr
Posts: 614
Joined: Thu Feb 28, 2008 6:48 pm

Re: Possible Defect: Sensitivity Analysis with Life Expectan

Post by jimr »

Ok. Great. This change will appear in the next release of the planner.

It will probably be called version 04.01.03, but any version newer than 04.01.02 should display the new error popup when this condition is encountered.

Jim

Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests