Allow users to compute total lifetime benefits using various assumptions

[Gregable]

One of the features that users really want, but probably don't need, is to compute the "optimal" choice for filing dates for themselves and their spouse. I personally think that the differences are small and the assumptions too large for any such calculation to be all that meaningful, it would be nice to provide especially if I can find a way to show users these facts (that the differences are small and the assumptions are large).

Ideally any such calculation would also highlight that the "longevity insurance" option.

[Gregable]

Suggestions by @stkeros include:

• user's inputs for dates of death
• claiming sliders could "jump" to the values which are optimal.
• one could set the claiming sliders as desired, but then have your site calculate the death dates which signify the breakeven point.

[Gregable]

I have an idea here that I'd like to eventually try out.

My observation is that there is some age for every user which if they die at that age or later, their optimal strategy will have been to file at 70 (or NRA if the user is receiving spousal benefits). Call this age for X. It may vary by user.

I could show a table showing ages 62 -> X for each user. One user on the x-axis of the table, One user on the y-axis of the table. Each cell in the table would show you the optimal filing strategy at that pair of filing ages. If two adjacent cells have the same optimal filing strategy, the edge between those cells would be hidden. Something like:

*--*--*--*--*--*--*
|  |  |     |     |
*  *  *--*  *     *
|  |     |  |     |
*--*--*--*--*     *
|           |     |
*--*--*--*--*--*--*

Where contiguous regions have the same optimal filing strategy.

Clicking any of the boxes will tell you what that strategy is. The hope is that there wouldn't actually be that many unique strategies in most people's cases, so there would only be say 10 different "optimal" strategies, and seeing where they fall on each person's age would give you a feel for what would be best. I don't know if this would actually work or not though! It may just be a complex mess.

[stkeros]

This is a brilliant, why-didn't-anyone-think-of-that-before, idea.

But as I was reading, I thought you were going somewhere else with this. What is actually on the x/y axis? Filing age? [Sorry, I now think that in your post you are saying what I am saying below].

Because then the box corresponding to each X/Y pair IS the strategy, so the box wouldn't have to describe the strategy. Instead, what if the x/y pair corresponded to longevity assumptions. [which is what I think you meant to write in the first place]. E.g. possible dates of death between age 62 and "very old" (your "X"). Then, each space on the grid/graph would represent/describe the filing strategy based on the longevity assumptions.

So sample ages of death at 68/90 would lead to a box with the "answer", perhaps color coded, or click in the box, etc as you mention, which then gives a the filing strategy.

I like it. Very good idea.

[Gregable]

Right. I think I mean the same thing: the X/Y pair is one where X and Y correspond to a specific longevity of the two spouses, whereas clicking on a box will show you the best strategy for that specific pair of longevities.

So, to extend my ascii-art a little further:

   *--*--*--*--*--*--*
64 |  |  |     |     |
   *  *  *--*  *     *
63 |  |     |  |     |
   *--*--*--*--*     *
62 |           |     |
   *--*--*--*--*--*--*
    62 63 64 65 66 67

The bottom-left cell would show the best filing strategy if you assume both spouses die at age 62.

The table t would need a lot more cells (going out to X for both spouses) and possibly per month of death rather than just year.

[stkeros]

Ok, we're on the same page. Again, it's a nice way to visualize the concept. You could have a color gradient, which corresponds to a given strategy. Red means both file late/70, blue means both file early. Or perhaps each cell has two halves diagonally down the middle, one for self, one for spouse, and each has its own color corresponding to early vs late filing. Assuming age 62 to 90 for each, that's about 800 cells. Might be hard to click on an individual cell unless cells were grouped by zones. Hmm.

The table t would need a lot more cells (going out to X for both spouses) and possibly per month of death rather than just year.

Perhaps calculating the strategy by month of death is excessive? First off, if you go to age 90 that's over 100,000 cells. Yes, most of them have the same strategy, but...

One could always start by just year of death (just assume one dies on their birthday, or mid-year, or whatever) and expand to month if there is demand or if you get bored. :)

[Gregable]

Well, there are at most (8*12)^2 = 9,216 strategies, since each earner can only select 1 date between 62 and 70. That's the most regions that could be displayed.

There are a lot of unknowns that need some experimentation. Given a pair of death dates, how many valid strategies are there for a typical pair of earners? Some of the 9,216 strategies will just be bad. For example, an earner with a zero PIA will never file after NRA. How slow will a browser crawl trying to actually compute each of these? 100,000 scenarios x 9,216 strategies per scenario might be a bit much for old Internet Explorer.

If per-month is too much and per-year seems too little, we can always do 6 months or 3 months.