Talk:Retirement calculators and spending

Purpose of Article
--ThePrune 22:23, 9 September 2011 (EDT) I wrote this article to implement suggestions by LadyGeek and Blbarnitz that all references to specific retirement spending models being used in retirement calculators should be gathered together in a common table. In preparing the table I decided to also include other useful information, especially the type of calculation used to extrapolate future investment returns. With just this little bit of extra work the resulting table are immensely more useful.

I decided to incorporate as footnotes additional information about how the calculator implemented certain spending models, or to explain how to access that model if it was not readily apparent (i.e. I had to hunt around in the calculator documentation to discover the technique).

I will leave it to LadyGeek and Blbarnitz to decide how to link this article back to the Tools and Calculators article.

The formatting needs improvement, so Go To It!


 * I tweaked the table by slightly expanding the abbreviated table headings in the Spending Model as Retirement Progresses sub heading for easier comprehension, and I also made these titles into internal wiki links to the original wiki page subheading discussions. --Blbarnitz 23:48, 9 September 2011 (EDT)

I moved the forum discussions and calculator specific notes to this article. Retirement calculators now links back to here. --LadyGeek 12:20, 10 September 2011 (EDT)


 * I renamed the abbreviations, as it was difficult to interpret within the table. As for table width, I'm using a widescreen monitor. It's much easier to read if the table doesn't span across the entire page width. I reduced the maximum table width and indented from the left edge. The individual column width percentages were kept as-is; it allows both tables to align with each other. If this doesn't look good, then change the table width back to 90%.


 * Note: MS Word likes to change the character set for single and double quotes, which is ignored (format not implemented) by the wikitext interpreter. The ending double quote for: "center” is not the same as: "center" (the trailing quote is a different character). Also for single quotes: X’‘‘ is not the same as X'''


 * This was not in the original table data. I encountered this problem when I was using MS Word to edit the tables, make global changes, then paste back into the web browser. Tip: Use Notepad or equivalent.

--LadyGeek 17:42, 10 September 2011 (EDT)

--ThePrune 21:55, 10 September 2011 (EDT) I changed table width back to 90%. I like to test appearance on both laptop screen and on a large monitor. At width=60% on a laptop the table rows are way too tall; many calculator names require 3 rows! At width=90% table look much better on a laptop screen, and is still OK on a large monitor.

Good point. I check standard aspect ratio by changing the browser size, but never dropped the resolution (1900 x 1200). The laptop is widescreen, but lower resolution (1366 x 768). I have a laptop, but did not remember to check these changes. The laptop is why I stay with clip-art images, they're optimum for low resolution displays. --LadyGeek 22:36, 10 September 2011 (EDT)

Additions Needed
--ThePrune 22:38, 9 September 2011 (EDT) There are a lot more Purchased Retirement Calculators that need to be added into that Table and categorized according to Calculation type(s) and retirement spending model used. Here is the list of calculators awaiting analysis:
 * Torid Retirement Planning Software (Professional)
 * Torid Retirement Planning Software (Personal)
 * Easy Retirement Planning (Retirement-4-U)
 * 401kXtra
 * RetirementWorks II (professional)
 * RetirementWORKS2 for You (personal)
 * NestEgg Retirement Planning software
 * ExecPlan Express
 * FinPlan Inc. Family Retirement Planner
 * Harvest-Time.net Retirement Planning Software
 * AgeBander Personal Planner
 * AgeBander Planner Pro
 * OnTrack Retirement Analysis Software

--ThePrune 08:12, 19 September 2011 (EDT) I have chosen to ignore Rob Bennett's Passion Savings Retirement Calculator. I don't feel a need to be THAT inclusive in these tables, given Rob Bennett's well known reputation across financial web sites.

--ThePrune 11:35, 15 September 2011 (EDT) sscritic has recommended that we somehow indicate in these tables the computing requirement(s) to run each calculator. He had wanted to use TIP$TER, but discovered he couldn't run it using his Mac. I agree with his suggestion. In fact, I had already started adding footnotes indicating which calculators required Microsoft Excel.

Obviously there are a couple ways to go:
 * 1) Footnote every calculator name with run requirements. Cons: TOO MANY FOOTNOTES. Better to have info directly visible in the tables
 * 2) Use yet another abbreviation system and place system codes into the tables. I personally like this approach.  Rather than creating a new column, it might look better to add the abbreviation code directly after the calculator name.  But I'm open to creating an entirely new column.

If we go with an abbreviation system, here are the possilities I know of so far:
 * Microsoft Excel only
 * Any spreadsheet that accepts MS Excel files
 * Internet, any web browser
 * Java - needs latest Java Run Environment software (i.e. Voyant calculator)
 * Runs on Wintel computer, OS XXX onwards
 * Runs on Apple MacIntosh computer, OS XXX onwards.
 * Others?

I'll need to once again review all the calculators and make a detailed listing of all run requirements. From this list a minimal set of abbreviations can be created.

--LadyGeek 20:41, 15 September 2011 (EDT): First, use at least 2 or 3 letter descriptive abbreviations; Single letter abbreviations are nearly impossible to work with. For example, I changed "RR" to "RepR" because "RR" is commonly used for "RailRoad." Next, the abbreviation system is not quite accurate. You are making this too complicated. Try:
 * Xcel = Spreadsheet
 * Web = online calculator run from within your browser
 * Win = MS Windows only

- MS Excel or OpenOffice / LibreOffice can read any spreadsheet, which covers all OS.

- If Voyant needs Java to run inside a web browser (I don't have a login to try it), then your web browser should already have this installed. Java is a common software application, used everywhere - Windows, Mac, Linux, Android (smartphone). You should always be up-to-date with Java. "Latest" has no added value.

- If TIP$TER won't run on a Mac, just combine the abbreviations, e.g. Xcel, Win. Windows is the dominant market, no need to add Mac or Linux (I use both Linux and Windows 7). If not, just add a footnote - I doubt that you'll need one. (clarified comments --LadyGeek 21:05, 15 September 2011 (EDT))

Calculators
The Retirement calculators links are duplicated in this article except for one. There's not much detail. Is it worth retaining? I don't know how to categorize it.


 * Retirement Planner - Bloomberg

--LadyGeek 12:19, 10 September 2011 (EDT)


 * --ThePrune 12:27, 10 September 2011 (EDT) The Bloomberg "Retirement Calculator" isn't really a retirement calculator any more! The text makes it sound like it should be, but all the calculator itself does is extrapolate investment growth from the present to a future retirement year.  Not very useful.  So I omitted it from my Tables and recommend that you just ignore it in the Tools and Calculators article also.

The sheer number of calculators (31 free and counting) is getting confusing- I wouldn't know where to start. Would it be possible to stop adding for while, perhaps reduce the data set to only the "most popular" few in each category? Otherwise, you will need to provide a "recommendation" which adds even more complexity. --LadyGeek 19:48, 25 September 2011 (EDT)


 * --ThePrune 19:10, 27 September 2011 (EDT) Yes, I am equally amazed at just how many calculators are available. We need to think more broadly about the potential uses of this calculator list.  I feel that once the uses are clear, the list length and composition will also become clear.


 * For example, one use of the list would be for any person to come, look for the calculator they are currently using, and see how it compares to others available. If this useage is valuable, then the list needs to be kept extensive.


 * If the purpose of the list is to recommend a handful of "better" calculators, then we need to present a neutral basis for making the recommendation. Notice that the overwhelming number of these calculators use a simplistic "Constant real Spending" spending model.  If we were to take the position that these calculators are inferior, then we could do one of several things:
 * Drop all calculators that only utilize a "Constant real Spending" model
 * Highlight by background color the calculators that make available better spending models. Tell readers that such calculators will give a more realistic estimate of required savings at retirement.
 * Move all the simplistic, free "Constant real Spending" calculators to a third table at the bottom of the article. Leave the "better" free calculators in the current first table.  Add some text before the new, third table pointing out why these "Constant real Spending" calculators are less realistic.


 * Personally, I have trouble agreeing to any editing approach that "restricts knowledge" by simply shortensing the list without any explicit explanation. It comes close to violating the Neutrality principle of Wiki articles. But helping people see how to "separate the wheat from the chaff", if done based on good logic, would be fine by me!

I would not remove any calculators simply to keep the quantity down. I also would not call any calculator "better" than another, as each person has a different impression of what "better" means. Perhaps it would be more efficient from this point on to only add calculators when requested. We have a large enough list that any additional requests should be far and few between.

--LadyGeek 20:50, 28 September 2011 (EDT)

The free calculator's Constant Real model explanation helps me filter through the list much easier. --LadyGeek 16:50, 29 September 2011 (EDT)

Life cycle vs. dynamic programming
bobcat2 has posted the following comments about the requirements for categorizing a Retirement Calculator as having a Life Cycle spending model:


 * In the Wiki both the Optimal Retirement Calculator and the Pralana Retirement Calculator are listed as life-cycle models. Neither of these retirement calculators is a life-cycle model. Because of the computational complexity of the life-cycle approach, dynamic programming (DP) methods must be used to solve the computations, as the Wiki correctly notes. Neither of these calculators use DP. The Optimal Retirement Calculator uses linear programming methods. While this may sound similar to dynamic programming, it is quite different and is not appropriate for calculating retirement spending using a life-cycle approach. I don't know how the Pralana Retirement Calculator works, but given that it is a fairly straightforward spreadsheet calculation it is definitely not using DP methods and therefore is not a life-cycle model. It will be a cold day in Hades when DP can be implemented in Excel.  Anybody can call their retirement calculator a life-cycle model. That does not make it a life-cycle model.


 * One could argue that the main computational difference (calculator type) between retirement calculators is whether they use DP methods, rather than whether they use MC methods or not. I would argue that. It's a shame, at least to the best of my knowledge, that there is only one commercially available household financial planning model that uses DP and thus implements the life-cycle approach. Particularly since there are dozens, or perhaps hundreds, of academic models that use the life-cycle approach in assessing household finances.

I will research bobcat2s comments to determine the validity of his assertion that "Life Cycle" and "Dynamic Programming" are inextricably linked. First, I propose doing a search to determine the range of meanings of the expression "Life Cycle spending model" within the economics literature. Based on this literature search I will attempt to determine if Dynamic Programming is required to calculate the Life Cycle spending in ALL usages of the expression, or only in certain preferred usages of the expression. Hopefully these "data" will settle the issues that bobcat2 has raised. --ThePrune 18:16, 16 November 2011 (CST)