All of the following are real questions received from PRC2016/Gold users along with the responses provided by Stuart Matthews of Pralana Consulting LLC.

Scenario Naming

Q: The Scenario names are truncating to S1, S2, S3 instead of either “Scenario 1”, “Scenario 2”, etc.

A: On the Special Controls page, there's a table that allows you to associate a brief name with each scenario to help you distinguish one from another.  If you leave this table blank, you'll just see the S1, S2 and S3 names.

Social Security

Q: Does 2016 version include changes to the Social Security Optimzer program to reflect the recent elimination of File&Suspend and Restricted Application by Congress?

A: Yes.

Q: Does the social security optimization section of the tool properly estimate retirement benefits at various ages, reflecting a decline in the benefit if I worked part-time from 62-67 and delayed claiming my retirement benefit until age 70 (compared to the standard assumption that my current earnings level continues to FRA)? Or do I need to use the calculator on ssa.gov and then plug in the resulting benefit amounts into the PRC tool?

A: Yes.  It asks you to specify two things: benefit at FRA and start age.  Based on the specified start age, your benefits will be adjusted upward or downward depending upon whether you start collecting prior to or after FRA.  If you start benefits prior to FRA and have earned income, your benefits will be reduced according to IRS rules.  The optimization algorithm takes all of this into account.  One final note: this optimization algorithm does the optimization based on the start ages that yield the highest savings balance at the end of your life.

Q: I understand that my age for PRC calculations is my age on Jan 1 of the subject year. For example I'm 64 on Jan 1 2016, so my age for PRC is 64 for 2016, even though I turn 65 in 2016. But does that go for Social Security optimization as well? For example for one scenario the recommended age to begin SS is 68. On Jan 1 of 2020 I'll be 68. So does the recommended age mean I begin taking SS in 2019 (when I turn 68) or 2020? If 2020, when in 2020 - Jan 1 or June (when I turn 69)?

Part of my confusion is I don't understand how the SS optimizer figures the age for calculated benefits. In the real world SS calculates benefits based on your birth month (or thereabouts). So if I were to begin withdrawals at the beginning of a year it seems to me I'd be getting a smaller benefit than if I waited till my birth month to apply.

A: Yes, PRC always assumes that birthdays are on January 1.  If you tell it to start SS benefits at age 68, you'll receive an entire year's worth of those benefits in the year it shows you being age 68.  That's also how the SS optimization algorithm works.  If in reality you turn 68 in the middle of the year then, yes, PRC will not model the benefit for the part of the first year in which you reach age 68.  I doubt if this matters much in the grand scheme of things, but it might be worthy of an enhancement.  So, I'm putting it on my list of things to consider for 2017.

Q: I ran analysis and noticed that even though I put in SS amounts estimated by SS website at my selected ages, the benefits show up to be zero next to the start age for both myself and my wife? Is this something that you recognize as an error I made? No difference on the SS details page/tab. 

A: Make sure you haven't entered a zero in the current benefits amount. If there's anything in that field other than a blank, it'll use that instead of the amount you entered for your FRA amount.  You enter the blank via the Delete button on your keyboard.

“Not Authorized” Errors and Moving PRC to From Computer to Another

Q: The new package doesn’t seem to be updating properly.  I did get a message that the old version was not authorized on this computer.  I do have a new computer.

A: It sounds like you've copied your 2015 version over to your new computer, which you cannot do because it is registered to run only on the first computer it is loaded on.  To move PRC data from one computer to another, you have to use PRC's Export function.  So, the way that works is that you bring up the source copy of PRC (in this case PRC2015) on your first computer, do an export to the PRC2015 Export Data file (which you can download from our website) and then transport that file to your new computer (via email or some other method).  Then, bring up PRC2016 and import the 2015 Export data file using the Import button on the Home page.  If your original computer is no longer available and you don't have a back-up copy (again, created via the export function), then you have two choices: start over or send me a copy of your 2015 file and I can reset the registration code so that it'll run on your new computer.

Q: I am a little confused on how to transfer my data from an old version of 2015Gold which I was using on my old computer to the 2016Gold version on my new computer. I have had several attempts of creating an export file and have failed each time as I get the following message when I try to import the file: This copy of PRC is not licensed to run on this computer.

A: The thing you need to understand is that the copy of PRC2015 that was on your old computer will not run on any other computer.  When it finds itself running on any computer other than the one it was initially registered on, it generates the message you included below.  What you need to do is EXPORT your data from your old computer into a PRC2015 Export Data file (which you can download from the User Support page of the website) by bringing up PRC2015 on your old computer, going to the Special Controls page and clicking the Export button and selecting the Export Data file.  Once that process is completed, then you can move that Export file over to your new computer, bring up PRC2016, go to the Home page, click the Import Data button and select the Export Data file as the source of the data.  Going forward.....when you elect to take further updates of PRC2016, you can import data directly from your older copy of PRC2016 by simply clicking the Import Data button.  This works because your older copy of PRC2016 will be licensed to run on your new computer and will be come up with no issues during the import process (unlike your old copy of PRC2015).

Monte Carlo and Historical Analysis

Q: I assume the projection reports are only based on the fixed rate assumptions.  Is this correct?

A: Yes, all projection reports are based on fixed rate assumptions.

Q: I’m up and running PRC2016 Gold, importing from my PRC2015 data file, and my initial read of it is that it looks great, but I can’t seem under the Projections tab any tabular output for the Monte Carlo and Historical analyses for the sensitivities I’ve selected. Have these tabular outputs been eliminated? In addition and relatedly, when you mouse over the upper and lower borderlines of the banded graphs under the Basic Analysis tab, it doesn’t give present any dollar values for the Monte Carlo and Historical displays, only output points numbers.

A: Yes, the tabular projection of Monte Carlo and historical analysis results was eliminated because I didn't perceive it as having much value. Also, I'm guessing that Excel doesn't show the numeric value of the edges of area graphs like it does on line graphs.

Q: Having access to the actual output of the Monte Carlo and Historical analyses was very helpful. In PRC2015 Gold, as you know, you could see that data for the sensitivities selected in either of two ways — via the tabular output (under Projections tab) and by “mousing” over the line graphs. In PRC2016 Gold, with both of these features eliminated, from what I can see, I can only estimate, at best, these values year-to-year from a visual inspection to “pick-off" values from the borderlines of the banded graphs against the axes.  On a related note, I know from your writings about the operating philosophy of Pralana, having verifiable outputs of the features it contains (or the financial calculators of others) is important. On that basis, it would seem to me that perhaps there would be a loss of fidelity without some sort of actual (numeric) data output.

A: Taking your last question first, yes, I believe transparency in a calculator is important but Monte Carlo and historical analysis results are WAY BEYOND any practical limits for user verification (because there's a ton of data).  Simply showing you the tabular listing of the year-by-year results is still very much a distillation of huge amounts of data.  Furthermore, the Monte Carlo data changes on each and every run which is why I personally only see value in the big picture it provides rather than any particular detail.

Q: No matter how much I read about the subject, I still don't understand the value of Monte Carlo analysis vs historical as applied to investment returns. Isn't Monte Carlo strictly for *random* processes?

A: You probably know that market volatility and the sequence of returns can have a dramatic long term effect on your savings which will vary to a large extent from fixed rate projections, even when the long term average rate of return is equal to some specified value.  Monte Carlo (MC) and historical analysis methods are two (imperfect) ways to simulate market volatility.  Yes, MC analysis is best suited for random processes but it does have value, at least in many peoples' opinions, in financial forecasting because it's a lot better than nothing.  Its weakness is due to the reason you stated: equities market returns are not really random, but neither are they predictable.  The reason I include it in PRC and the reason I think that many other calculators also include it is because we need SOME mechanism for simulating market volatility and trying to forecast the future rather than simply relying on fixed rate projections (which are also flawed).  There is no mechanism that's perfect, so I've incorporated three different and generally independent methods for attempting to establish a range of likely outcomes.  I couldn't say whether MC or historical analysis is better; they're just different and both have strengths and weaknesses.  MC generates random returns around some long term average, which doesn't simulate the cycles we actually see in the market, while historic analysis uses sequences of returns that have actually happened in the past but the issue there is that the past is not the future.

Q: The equities market has plenty of patterns, dependencies and correlations that seem to make it anything but random.  And what's worse is that the Monte Carlo results always seem to be significantly worse than the historical ones.

A: Regarding your point about MC being worse than historical: I agree, but that's because we probably tend to be more conservative in selecting average rates of return than historic averages would suggest.

Q: I get the “Scenario 1 analysis results are invalid” message on the Analysis pages of my file. What is the overall issue that would cause this message to be generated?

A: This error message only comes up after doing an import and prior to performing the scenario analysis, which is done by you clicking the “Analyze Scenario n” button.  So, if you haven’t already done so, just click that button and see if that takes care of it. 

Q: So, I know this likely gets into hidden columns, so I'll tread lightly ... but I'm working on my "score card" for the user worksheet and want to get at the Monte Carlo and historical numbers, maybe something like: what is the lowest low percentile value, and in what year, and what is the first year where the low percentile goes below zero? Or what percentage of the low percentile values go below zero? I assume those shaded blue areas come from the Monte Carlo and historical simulation Savings Projections on the hidden Projection tab?

 I know you already said you don't like the "Funded Ratio", and maybe I'm getting too close to something like that. Is it a fool’s errand? Or do you have some ideas on this? I know I need to look at the graphics, and the tables, and take it all in to gauge success. But the idea of my score card is not just getting a "quick answer" but comparing results over time.

A: The Monte Carlo data is on the Basic Analysis page and the historical data is on a hidden page and it is distilled via the percentile calculations to turn it into the blue shaded areas and the columns on the hidden projection page.  By its nature, the MC data changes from one run to the next, so I think trying to glean any useful details from it is a waste of time.  It's the big picture as presented in PRC's charts that contain the useful information, in my opinion.  The historical analysis outputs do not change from one run to the next but I'm still not clear on any value to be gained from looking at the details.  That's why those columns on the hidden projection page don't appear on any visible pages.  A couple of other users have asked for them and I've stated that I would consider it for a future upgrade, with the intent of seeing how many other users express interest. 

 A suggestion for now is to play with the percentile settings and re-run the analysis to get a better feel for the results.

User follow-up.: That is fine for now. I am in favor of any set of objective measurements (perhaps accompanied with disclaimers), that you might eventually agree to put under the Analysis Graphs. I'm assuming there are some relatively simple ones, especially for fixed rate, that put into words what the graph already says. Examples might include "savings left at death", or "ran out of savings in 2035"; notice these do not interpret or predict, just take key data from the graph and put it in plain sight as a number.

User-Specified Savings Contributions

Q: My wife doesn't qualify for Long Term Care Insurance, so we are adding $6,000 per year to a taxable equity account. How can I enter that in your program?  I don't think just adding it to Cash is the same.

A: You can't do this directly for two fundamental reasons: first, asking users to specify how much they plan to contribute to savings is counter to PRC's design philosophy and, second, PRC doesn't have the ability to track the balance of sub-accounts within the larger categories of regular, tax-deferred and Roth savings.  To elaborate, PRC is built around this principle: based on user inputs it build detailed income and expense profiles, and then subtracts expenses from income each year.  If the result is positive, it's treated as a contribution to cash/regular savings accounts.  If it's negative, it's treated as a withdrawal.  There's a control at the top of the Special Controls page that enables you to specify the percentage of these contributions that go into cash accounts vs. regular investment accounts. So, as long as this control is set appropriately, PRC will accomplish your goal of contributing $6000 (or more) to a taxable equity account each year, but you won't have the ability to see the specific growth of that individual account.

Tax Calculations

Q: I can play with our specific income tax situation in TurboTax. Can I input those results in your program instead of your automated results?

A: No.  PRC calculates your taxes in detail using accurate tax tables and, unlike TurboTax, does it for every year in the future based on potentially changing circumstances in the future. 

Q: When putting in a future increase to the Federal (or other) tax rate, if I enter 10% (0.1) would I be going for example from 25% to 35%, or from 25% to 25%x1.1=27.5% ?  I’m assuming it’s the former.  An example in your Manual or cell comment would help. 

For the decrease in SS: from your comment it looks like we enter a positive number, so if I expect a 25% reduction in the future I enter 25% (0.25)?

A: The tax calculation formulas apply this as a multiplier.  So, if you specify a 10% increase, PRC will calculate the taxes and then multiply them by 1.1. 

Q: I’m not following the AGI calculation. I’m retired with no earned income, have no regular savings, no 529’s; shouldn’t AGI simply match my tax-deferred withdrawals? Why is it more than double normal in the year I turn 70?

A: There are two reasons for this:  First, AGI includes RMD’s, which begin at age 70. Second, PRC pays the taxes on withdrawals from tax-deferred savings which are driven by negative cash flows in the year after they occur.  You can also see this in 2016 where you have that big withdrawal but no taxes (because the tax on those withdrawals is paid in 2017).  I think you'll easily understand the reason for this:  PRC determines cash flow by subtracting expenses from income, where expenses include taxes.  Then, it adjusts account balances such that "deposits" are made if the cash flow is positive and "withdrawals" are made if the cash flow is negative.  But withdrawals from tax-deferred accounts are taxable events, and rather than get into some iterative tax calculations I chose to simply include those withdrawals in the AGI for the subsequent year.  The same thing is true when making withdrawals from taxable accounts that result in capital gains. 

Q: We are a couple, not married, and I have everything entered separately for us, no survivor benefits. And we are contributing equally to our expenses in retirement. On Taxes tab, M for married automatically populates. Is that affecting tax calculations differently than it would for two single people filing separately, such as we do?

A: PRC doesn't model the taxes of unmarried couples and it also doesn't model the filing status of married filing separately, so when it sees that two people are being modeled it assumes they're married and, thus, automatically selects the M filing status which will probably result in lower taxes than what you're actually experiencing.  My best thought at the moment is to try to estimate what the error is (with Turbo Tax or manually) and then compensate for that with an additional expense on the Miscellaneous Expense page.

Graphs

Q: What is the "Fixed Rate Savings" on the graphs in the Analysis Results?  Is it the combined ROR from the asset allocation & historical returns?  Manual is unclear.

A: I think you're referring to the dotted line on the Analysis graphs.  I guess maybe I made a poor choice in choosing this name for the legend and should probably have called it Fixed Rate Projection.  That's what it is.  In the paragraph on Presentation of Analysis Results on page 55 of the user manual, it states that "fixed rate results are superimposed as a dotted line".  This is the same data shown in tabular form in the Savings column on the Projection pages.

Q: For Custom #2 graphs, with nothing selected, I get multiple earliest death indicators (4 to be exact).   The multiple indicators seem to persist for some data selections (income and expenses for example) but by selecting some additional parameter (such as Net Worth and Net Savings) the indicator (sometimes) presented correctly only once.

A: I know what you're referring to and I simply haven't addressed that because I'd associated it with no data having been selected for that graph.  I just now looked at it and it seems to occur anytime less than four data sources are selected.  So, I'll see about fixing it in conjunction with the next update.

Q: This sounds like a silly question as I ask it, but what values are being plotted in these Basic Analysis graphs? Is there a column in one of the projections tables that corresponds to the graph? I assumed this was net worth, but I cannot get the plotted values to match net worth or anything else. Right now, I'm specifically talking about the dotted-line "Fixed Rate Savings", but am interested in the Monte Carlo and Historical analysis results also. If it is in the manual, please point me to it, but I see things like "Fixed-rate analysis generates a single projection" but not what projected value is plotted. 

A: I went back and checked both the help info in the tool itself as well as the user manual and came away with the same conclusion you did: it's definitely not clear.  The answer, though, is that it's the Net Savings line from the Summary projection page.  The only hint given is the legend on the analysis graphs which says Fixed Rate Savings (and it'll only match when the analysis has been run with no further input data changes). 

Q: I saw the "clue" and tried to match the graph to Net Savings by hovering the mouse over the graph and comparing to the projections table, and it was close, but not exact.

For example, using Bob and Dorothy, Scenario #1, Today's Dollars:

Year

Graph

Table

2015

600,000

600,000

2025

1,230,038

1,239,913

2035

1,613,798

1,596,289

2045

1,998,468

1,988,417

2055

2,316,630

2,312,609

2065

2,577,415

2,577,415

What am I missing?

A: An interpolation process is used for dynamic control of the length of the horizontal axis to match your lifetime, resulting in slight differences at intermediate points. 

Withdrawal Prioritization

Q: If I choose to withdraw in this order: Tax Deferred then Regular Savings then Roth, if the PRC starts drawing from Tax Deferred when we’re 56/55, is the 10% penalty being captured (where?), as well as the income tax rate?  Is there a way to start with Regular Savings til 59-1/2, then Tax Def, Reg & Roth?  I see it starts with Cash when we have enough of it over our minimum level.

A: Yes, the 10% penalty is applied to your taxes but you can't specifically see it in any dedicated column.  At least for now, there's no way to associate different withdrawal priorities for different time periods.  I think that might be a reasonable enhancement for sometime in the future, though.

Q: Does the tool allow for comparison of various withdrawal strategies by tax “bucket” – for example if I wish to draw/convert my IRA/401k assets from age 62 to 69 to meet my annual income needs, then switch to drawing just RMDs and take the remainder from either capital gain or tax-free assets?

A:

  • It WILL let you start smooth withdrawals from your tax-deferred accounts at a specified age on a per scenario basis, and this will cease at age 70 and be replaced with RMD’s.  The size of the pre-70 smooth withdrawals are not user-controllable but, rather, are calculated for a constant annual distribution over your remaining lifespan.  This won’t actually be a constant dollar value, though, because the balance is affected by ongoing contributions and annual growth.  These distributions will, in effect, go into your taxable savings from which withdrawals will be made, as required, to cover negative cash flows.
  • It WILL let you specify a withdrawal priority on a per scenario basis, but it WILL NOT let you change that strategy at some point in the future.  It’s constant throughout the entire modeling timeframe for a given scenario.  This prioritization scheme allows you to put taxable, tax-deferred and tax-free (Roth) accounts in any desired order.
  • It WILL let you do rollovers from tax-deferred to Roth accounts using the timing of your choice on a scenario basis.  You can select the start date, number of years over which the rollovers will occur and the percentage of the tax-deferred funds to roll over.]

Inherited IRA’s

Q: Do you have any advice on how to model an inherited IRA? I get current income from it with a defined (by the IRS) withdrawal. It's tax-defered, but I have no choice as to when I take the withdrawals (RMD).

A: Since PRC doesn't directly model an inherited IRA, I suppose the best way to handle it would be to establish another stream of income on the Income page, using one of the five Other Income choices.  With that, you can specify the current amount, future growth rate and its taxability, which in this case would be as regular income.  The biggest downside of this I see is that the IRA will not be included in your total savings amount but, rather, only as a source of future income.

Required Minimum Distributions (RMD’s)

Q: I am trying to figure out Pralana handles RMDs or any other distribution from IRAs.  In particular, in the Income Projection summary I would have expected to see some income coming out of our IRAs, but I do not.  The only retirement income I see is my wife’s (modest) taxable pension treated as regular income and our Social Security income.  Hence a large part of what will be our income starting in 2017 when we retire is not appearing, and hence the projected total spendable income is much lower than it will or should be.

A: The Income Projection page includes only income from external sources as opposed to growth of and distributions from accounts you already hold.  You'll find RMD's on the Accounts Projection page. On a related note, the Total Income column on the Summary Projection page doesn't include RMD's and growth of accounts as income.

Q: It appears that all of the required minimum distributions from my tax-deferred accounts are being treated as ordinary income for tax purposes. A significant share of my contributions to these accounts was made with after-tax money which shouldn’t be taxed again when I withdraw it. Do you have any suggestions as to how to accurately reflect that in the tax projections?

A: There's no way to do that within the tax-deferred category, so I’d recommend modeling the after-tax portion as a Roth IRA.

Sensitivities Page and Asset Allocation Balancing

Q1: I don’t understand why the what-if analysis on the Sensitivities page shows a deviation when the key assumptions are the same as the baseline? See attached screen capture…

q and a pic1

A1: My guess is because you have changed one or more parameters that are not included on the Sensitivities page since the last time you initiated an Analysis of the selected scenario.  As you are probably aware by now, there are many pieces of input data that PRC uses in generating projections and in performing the analyses that are commanded from the Basic Analysis page.  When the analysis is performed (after you click the Initiate Analysis button), PRC saves the resulting projection and the key parameters that are shown on the Sensitivities page.  If you subsequently go back and change some other input data, it will affect the what-if projections but not the saved baseline projection.  So, with no further information to work with, that's my best guess as to what you're seeing.  I'd suggest playing with it some more and get back with me if you observe some behavior not explained by my supposition.

Q2: I didn't change anything after running the analysis. I just go to the analysis section, click on analysis of scenario 1, then go to the sensitivities and see the discrepancy. I noticed that when I don't have a period 2 with different allocations in the allocation table, the problem disappears. When I have a date for period 2 in the allocation table, the curves start diverging at that date. So the what-if parameters are not including a change in allocation.

A2: That's exactly right! Mystery solved.  Please elaborate on the importance of this design simplification to your use of the tool.

Q3: I think it’s important as the distribution of asset classes needs to change as taxable accounts are depleted. Tax-deferred accounts are typically initially loaded with non-tax advantageous investments like bonds as long as taxable accounts can carry enough in stocks to preserve an overall asset allocation goal. As taxable accounts are depleted, the asset allocation of the tax deferred accounts need to be adjusted back to the target allocation.

The current model doesn’t allow for this progressive rebalancing. Since my discovery shows pretty dramatic impact on the results, I think there is value in addressing the issue.

Maybe a solution is to simplify the target allocation to make it an overall portfolio allocation.

Then, depending on the spending strategy in the special controls tabs, the model would maintain the target allocation by rebalancing the allocation across account types.

If the feature is required, it would still be possible to declare a date where a different asset allocation should be reached. The migration between the initial asset allocation and target asset allocation could be automatically smoothed out.

A3: All good points!  The design will be further refined in PRC2017.

Q1: I am puzzled as to why when the baseline and what-if parameters are set identically the analysis remains different. I cannot find an explanation in your help files and have to think it represents a bug.

See screen shot below.

q and a pic 2

A1: There are at least two possible explanations.  One is that you didn't perform a scenario analysis via the Basic Analysis page with no intervening changes prior to exploring the effects of changes on the Sensitivities page. Another is that, by design, the Senstivities page doesn't include the full set of parameters that are available for establishing the baseline. An example of this is the two separate periods for rates of return, only one of which is available for testing sensitivities. 

User follow-up: Eliminating the second period of ROR allows the sensitivities to come together at baseline.

Q: I don’t understand why retirement age isn’t one of the fields in the sensitivity analysis.  Is there some other easy way to see what changing the retirement age does to the plan?  If so, I haven’t found it yet.  Or, maybe your thought is that you model that with the three different scenarios, but then I have to keep everything else constant and go to three different screens to see the differences.  Anyway, I think you get what I’m looking to do, and can point me in the right direction if there’s an easier way I don’t understand.  The only way I know right now is to adjust the pre-retirement income on a different page and then come back to the graph, but if you’re trying to watch how things change as you adjust parameters, it’s awful hard if you’re jumping back and forth between sheets.

A: Retirement age isn't included in the sensitivity analysis because there are too many corresponding assumptions that you'd be expecting the tool to make that would be wrong 90% of the time.  Here's a list of examples:

    - PRC allows you to define many different income streams, and they start and stop based on age and not based on whether you're retired or not.  This allows for maximum flexibility but makes it impossible for the tool to make income projection adjustments simply based on your retirement age.

    - Pension amounts are typically a function of years of service and average income level; PRC has no mechanism for adjusting that based on you retiring in one year versus some other year (the reason is that there's just too much variation in pension plans for me to try to model all those formulas).

    - What if you're planning on moving from a high state income tax state to Texas with no state income tax when you retire?  PRC allows for you to specify different state tax rates, but it's not specifically tied to your retirement date.  If I did tie it to retirement date, would I tie it to yours and your wife's?

    - Expense streams are tied to specific start and stop years for maximum flexibility, and this is independent of retirement age.  Maybe you're planning on downsizing your home when you retire, or maybe you plan on downsizing two years after you retire, or maybe you're planning on trading your home for a sailboat and sailing around the world three years after you retire.  Whichever the case, PRC has no way to correlate your expenses with your retirement date.

    - I think by now you get the point.  If income and expenses were simply lumped into two simple categories, such as pre- and post-retirement, it would be easy to do what you're suggesting.  But that's not the case.

Q: If I understand the way the sensitive analysis works, it would be really nice if you could tweak things in there, and then when you’re happy with the “plan” you could tell it to apply those changes to your base plan.  Hopefully that made sense … and I’m assuming there’s no way to do that, right?

A: I understand what you're asking in regard to applying the sensitivity page settings to the baseline plan and to confirm, no, there's no way to do that.  It's purely intended as a what-if tool and it only includes some of the major parameters that I thought a user might like to do what-if's with rather than the total set of input parameters.  Finally, the ROR parameters are in an aggregate form rather than much more detailed form uses to derive ROR on the financial assets pages.  Therefore, it wouldn't be feasible to apply at least some of the sensitivity page settings to the baseline plan.

Nominal vs. Real Rates of Return

Q1: On the Asset Allocation tab in Financial Assets, I see a Tax-deferred Aggregate Real ROR of 3.05%, yet in the Accounts tab of Projections, I see the column that has Nominal ROR on Tax-Deferred Savings has 6.1% … Exactly double, which is either a coincidence, or I did something wrong in the data entry.  Any thoughts?

A1: The nominal ROR shown on the projections pages is the real ROR from the financial assets pages plus inflation.

Q2: Not sure I understand that.  It would seem like inflation would decrease your return rate, not increase it.

A2: Maybe it’ll seem more correct if I explain it this way:

 The nominal ROR is the actual return you see on your money.  If you start with 100,000 and earn a nominal rate of 6%, you’ll have 106,000.  But if we have a 3% inflation rate, then the REAL rate of return is the 6% nominal rate MINUS inflation, yielding a real rate of just 3%.  So, you can look at two ways:

real ROR = nominal ROR – inflation

or

nominal ROR = real ROR + inflation

Q: On the Projections/Accounts page it shows the ROR as 6.8% for the different savings accounts.  However, on the Financial Assets page for these accounts, it shows something lower than this.  Therefore, I am confused where it is pulling this 6.8% value from as it seems high in this environment.

A: The ROR on the projections pages in NOMINAL ROR whereas you're looking at REAL ROR on the financial assets pages.  Here's the basic formula that relates them:

Nominal ROR = Real ROR + Inflation, which can be restated as Real ROR = Nominal ROR - Inflation

Healthcare Expenses

Q: Healthcare expenses in today's dollars in your example data does not look right.

A: I’m assuming your concern is that the cost is specified as $36,000 in today’s dollars but shows up as over $100,000 in today’s dollars on the projection?  The explanation for that is that we’ve specified additional healthcare inflation of 2% above general inflation on the Home page.

Q: Both my wife and I will retire in a few months.  I am 62 and she is 56.  We will both have federal pensions and be eligible for health insurance after retirement. Am I correct in assuming that until I turn 65, the calculator will use the Period 3 rates?  Then I enter our premiums in Period 4 for our medical insurance and my Part B Medicare, in today's dollars?  Likewise for Period 5, but Part B  and health care insurance premiums for both of us? Put another way, does your calculator adjust the annual costs for each period based on when each of us turns 65?

A: Yes, you’re exactly correct.

Q: On the Income screen, how do I account that about 4% of my gross goes towards paying for health insurance and other pretax benefits? 

a.       Should I place this percentage on the “% of pre-tax income contributed to a defined benefit plan” (which doesn’t seem quite right per the instructions, and pop-up comment). 

b.       Or should I reduce my gross by this amount?

c.       Other way to account for this?

A: You enter this on the Healthcare Expenses page.

Property Expenses

Q: I am renting my old primary residence out now and have modeled that. We are retiring, going to travel for 5 years, I have that modeled. But we wish to return to the primary residence for 2 years, to reestablish primary residency, before selling, to avoid capital gains that would incur if sold as a rental. How would I enter that?

A: Sounds like a really good plan, but PRC does not model this case.  I assume you've got the property modeled as rental property to take advantage of the depreciation deduction, which I would say is appropriate; however, there's really no way to convert it back to residential property, terminate the rental income and start paying the expenses yourself and then avoid the capital gains.  So, my best thought is to keep it modeled as rental property and simulate its sale when you move back in to terminate the rental income and depreciation deduction, and set up a separate personal property purchase in the same year so that you can model your expenses for taxes, insurance, utilities, etc and the appreciation on the property in the two years you live there.  THEN, you can set up a one-time Other Income on the Income page to offset the capital gains taxes that PRC will generate at the time the rental property is "sold".  You can specify that this income is non-taxable so as not to further complicate your taxes.  The thing is, though, you'll have to play around some to figure out the correct magnitude of this income.  Also, since you're not REALLY selling anything when you convert the house from rental to residence, you'll need to be careful with the numbers so that the simulated "sale" and the simulated "purchase" are zero-sum.

Q: I'd like to be able to model implications of switching from owning a house to renting an apartment. The problem is I don't see a way to fully do that by scenario. I can, by scenario, indicate when I would sell. But the only way I see to enter the new rent expense is in Discretionary Expense, which isn't tied to scenarios. For example, let's say I input selling the house in 2020 on the Personal Property Scenario 1 table. I can enter a Discretionary Expense, namely rent, starting in 2020, but it's effective across all scenarios. I'll bet a lot of people need to consider this when planning retirement. Any tips?

A: What I would recommend is doing all of this on the Property expense page, with two separate properties.  Sell your current home in 2020 and acquisition of the apartment in the same year (that won't cause double expenses because the transactions will occur on Jan 1) at zero cost.  This way, you can include all of the other expenses associated with the apartment such as utilities, rental insurance or whatever, including rent.

Q1: I own a condo boat slip which I now rent. I had it as “personal property” (by the way, I think houses and the like should be “real property” not “Personal Prop”) but, since I am renting it now, I switched it over to Rental Prop in your model. Actually, I suppose none of these facts matter for the possible issue I am going to describe………...

The issue is that when I run the Expense Projection using “Today’s $”, it appears the final income figure in the year of the sale is shown in “Future $”—it seems too high in any event.  Said another way, the annual net income stays the same on the Expense Projection but the final income figure in the year of sale is substantially higher than the Current Market Value of the property/slip and also much higher than the Current Market Value of the slip plus one year of net income. When I change it to Future $, all of the net income numbers on the report are substantially higher.

So the basic question - When using Today’s $ in the Expense Projection report,  shouldn’t the “Net Expenses For Rental Property” in the year of the sale equal Current Market Value plus one year of net income (Rental Income less operating costs) in current dollars?   

A1: Be aware that the Rental Property Summary Table (on the rental property expense page) is always shown in terms of future $, so it will not agree with the Expense Projection page when it's being shown in today's $. "Net Expenses for Rental Property" will not include any net income from the year of the sale because it's always assumed to occur on January 1.  So, it'll be equal to the total sales commission minus the equity in the property. 

Q2: Perhaps I need a short lesson in the “Today’s $”  and “Future $” as applied in the model.   The “Net Expense for Rental Property” (see Screenshot 4 above) is an equal dollar amount for every future year until the year it is sold in the Expense Projection when the setting is Today’s $. That makes sense to me. However, on this report when using the Today’s $ setting, I would expect to see the proceeds in the year of the sale equal to the value of the property today ($70,000), less the sales commission. This is not the case. It is $118,085 which is apparently the future value of property less commission?? But that does not make sense either because when I run the report with the Future $ setting (see Screenshot 5 above), all of the cash flow numbers grow over time… which makes sense….except for the final figure which is now $213,275. If $118,085 is the Future $, then what is the $213,275 figure?

A2: The issue is that you've specified a Real Appreciation Rate of 3% on the slip, which means it's appreciating at 3% above the inflation rate (please refer to the comment in this cell for details).  If you'll change this to zero, you'll get the numbers you're expecting.

Rollovers

Q1: My task is to try to find the best drawdown strategy while keeping taxes as low as can be achieved while maintaining income of about $100,000 for expenses.  We have a skewed asset base in that 90% of our treasure is in IRAs so potentially facing a huge tax bill with RMD.  Naturally Roth Conversion is something I want to look at.  We are 64 and 63 presently with one worker and I will soon be on Medicare.  Both of us were high earners so our SS benefit at FRA is going to be $31,200.  I was trying to model different Roth strategies and the graph display seems to indicate an advantage but the box to the right of the graph says your strategy is worse than….and the number is in green I think.  I thought the top line which is in green was the Roth conversion and it looked like better choice???Is the box always reading your strategy is worse and the color indicates + or -?  Also since we have 5 years before RMD start can we do a model where we drawdown IRAs for living expenses before and specify the % of drawdown?  I tried the SS optimizer and tried with us both claiming at 70 and me 68, husband 1/2 spouse for 3 years then switch to own….did not seem to make much difference but the output was this bar graph which I did not fully comprehend.  Do you have any insight into what Roth Strategy I should try to model?  Basic info is that we have 305K in taxable but 4.65 million in our combined IRAs and 337K in Roths.  We have a home worth 600,000 and a rental worth 520,000 which we want to pass on to my daughter.  So I know we will be paying a lot of taxes…and we are happy that we have saved so much but would like to leave something to my daughter while enjoying our retirement years.I have used the Fidelity retirement analysis tool and am comfortable with our budget and also Monte Carlo analysis has indicated that we are well funded….can your program help me with crunching out some better drawdown scenarios ?

A1: Regarding the graphs on the Rollover page, the green line is your baseline (the one created when you initiate an analysis from the Basic Analysis page) and the red line is the what-if plan associated with the selected rollover parameters.  The box containing the dollar difference between your baseline plan and the what-if plan will be highlighted in green if the what-if plan is better than the baseline, or in red if it's worse.  So, if the red line is beneath the green line (at end of life), I'd expect to see a red background in the box; if the red line is above the green line (at end of life), I'd expect to see a green background in the box.

 Yes, you can specify smooth withdrawals from your tax-deferred accounts prior to age 70, and these will be treated like RMD's in that they're deposited in your taxable accounts and included in your taxable income. You do that from the Special Controls page. Whatever parameters you enter there will be overriden by RMD's starting at age 70.

I'm puzzled by your reference to a bar graph on the SS optimizer page.  It should only contain a big square with your possible SS start ages across the top and your husband's possible SS start ages down the left side.  Within that square, there should be one dark green square to indicate the optimum start ages for each of you and potentially several lighter green squares that indicate sub-optimum start ages that are almost as good.

I don't think I can really comment on an appropriate Roth strategy for you.  My guess is that you will be better off to just pay the RMD's, but I'd recommend playing with the various parameters (like spreading the rollovers over quite a few years) on the Rollovers page and see what it tells you (like you've already started trying to do).  Another thing I'd recommend that you evaluate is the prioritization of withdrawals, which you also do via the Special Controls page.

PRC can definitely help you establish the best drawdown strategy, but it doesn't do it automatically.  The design intent was to give you the tools to investigate a large variety of approaches while helping you compare the alternatives.  So, one approach you might try is to set up all three scenarios with your basic information as you've described below, and then vary some of the parameters between them and compare long term results.  For example, different withdrawal priorities, smooth withdrawals prior to age 70 on one of them, different Roth conversion amounts and durations, and so on.

Q2: Ok…that helps…I was playing with the drawdown scenarios and it was quite amazing the difference and really neat to be able to just change number of years or percentages. I almost did not believe some of the numbers, so you are right go back and review all your inputs. When I do find a Roth Conversion that I like can it be included in my analysis?  Right now if I am doing the Rollover Page scenes they are just hypotheticals and not being applied to my final models except for the analysis on that page…right?  Is there a way to apply a Roth Conversion, perhaps it is in the Special Controls page.  I will revisit the Social Security page, it is displayed as you say but I think that SS will not be as big a factor as our RMD.  Yes, we are very thankful to have saved and invested well.  I will work on optimizing our retirement and passing some on to our children, hopefully more intelligently with the aid of your program. Thanks for the feedback.

A2: Rollovers are totally controlled via the Rollover page.  Whenever, for a given scenario, the "switch" is in the ON setting, the specified parameters are actually part of that scenario right then and there (as you can verify by looking at the various tabular projection pages); however, they won't be included in your Monte Carlo and historical analysis results until you go back to the Basic Analysis page and reinitiate the analysis for each scenario.  That's why you can observe that the red and green lines diverge on the Rollover page whenever you change the rollover parameter.  But, they'll become totally converged again whenever you run an analysis.

Q: So I was noticing the “OFF” setting in the user manual…so if I set Scenario 1 to OFF and run analysis…then do I turn ON the scenario 1 when I go back to the Rollover page and dial in my various trials on the scenario 1 toggles, or do I do it on scenario 2, etc?  If I find one I like then I leave the switch ON and analyze and it will dial in for subsequent analysis

A: Yes, that's correct.  Simple rule: For any given scenario, if you want your rollover settings to be considered when running an analysis, set the switch to ON; if not, set it to OFF.   All scenarios work the same way and they're mostly independent of each other, as are the respective rollover parameters.

Funded Ratio

Q: I am working with a scenario where I have very significant assets at end of life yet show to be severely underfunded by the funded ratio metric.  I can match all the calculations so think it is just a question of definition.  The numerator does not included the earnings power of assets while the denominator includes the income tax consequences of those earnings.  This leads to the situation where the funded ratio actually improves if I reduce returns (no change to numerator but denominator gets smaller due to lower income taxes).  Was wondering if the funded ratio should maybe include the earnings/capital appreciation on assets?

A: I've come to believe the funded ratio is a questionable concept and generates more questions than it answers.  You're right that growth of savings after retirement are not included. 

Q: Is the funded ratio value based on the fixed rate assumptions?  

a. In the future would it be possible for you to also compute it based on the Monte Carlo and historic data?

b. I am a little confused by this ratio, though, because for the three scenarios, it shows the ratios as 0.84, 0.86, 0.91 which implies there are not enough funds to finance my scenarios.  However, when I look at the graphs, and reports, it shows I don’t run out of money.  In addition, the scatter diagrams show success rates of 100% with money left over.  Therefore, I am confused by the Funded Ratio being less than 100%.

A: I intend to eliminate the funded ratio in the 2017 version because I've come to regret having built it.  It creates more confusion than anything else and belongs on very simple calculators rather than full-blown tools like PRC/Gold.  It is currently based on fixed rate assumptions and can be misleading because it does NOT take into account any growth of your accounts during retirement.  It's intended as a retirement readiness gauge, I think.

Mac-Unique Issues

Q1: I recently purchased your software and it has been running great for the last few days.  However today when I tried to enter a % financed under Personal Property I ran into a problem.  Anytime I tried to put in a number, the entry changed to “xx% of the Value You Are Financing”, actually populating the cell with text.  It would remain in the cell as a text entry and then cause a Run Time error problem. See screen shot below.  I also tried to download a clean copy and tried to change the name cell and got the same problem.  I am running on Excel for Mac 2016.

q and pic 3

A: That is very strange behavior that I can almost replicate on my Mac system running Excel for Mac 2011.  What I'm seeing is that the "% of the Value You Are Financing" text appears whenever I'm entering a value in that cell, but it disappears when I hit enter and doesn't actually get inserted into the cell.  Therefore, it's odd but doesn't cause any problems.  It seems to do the same thing with some of the other rows in that column, but never on any other columns. This does not occur on Windows systems and I'm at a loss to explain it, and it's never been reported before.  Also, I tried Googling this but didn't find anything.

Are you able to work around this by either 1) selecting and deleting the text portion of the cell data or 2) dragging a percentage value from another cell into the troubling cell?

User follow-up: I tried what you said - dragging a value from another cell.  That seemed to work. 

Use of PRC for Long Term Planning

Q: I'm only a year away from retirement, thus I'm interested in using the tool throughout my retirement to do check and adjust type decisions. How best do you recommend doing this? I can fudge dates but have you or any of your clients suggested best ways to do that type of functionality?

A: PRC is absolutely up to the task of supporting updates of your situation and adjustments to your plan throughout your retirement. 

- with its support for multiple scenarios, you can define, explore and compare different options (including timing of events)

- PRC generally undergoes one major update each year and that update is released in December; you’ll have the opportunity to do that upgrade and then have the tool automatically import all of your data from the prior version; then you can make annual adjustments to your old data such as changing the start year, your age, your account balances and anything else you wish to change about your plan going forward

- should you choose not to do the upgrade, you can still change the start date, your age, your account balances and anything else about your plan on an annual basis using the copy of PRC you’re using today.  It’s smart enough to index the tax tables based on the specified starting year so it’ll still be fairly accurate even if doesn’t actually use the latest tax tables (which will only exist in the new version).

Testing of PRC

Q: The program is very sophisticated and complex (as opposed to complicated). How can you test it to ensure that the results are valid?

A: PRC exposes lots of details for users to examine (aggregate rates of return, loan amortizations, income and expense profiles, tax calculation details, year-by-year account  balances, etc).  Therefore, given time and adequate math skills (and a few formulas at the end of the User's Manual) a user can generally see what it's doing for fixed rate calculations (which is what appears on the various projections pages).  Once you convince yourself on that, you can possibly convince yourself of Monte Carlo and historical analysis validity by general comparison to fixed rate results; however, this requires a fairly good understanding of what's going on in these types of analysis and careful study of the data used in those analyses (for example, the historic data for each asset class as shown/selected on the Historical Data page under Financial Assets).  For example, you'd generally expect Monte Carlo results to "bracket" fixed rate results but you might be surprised to discover that 50th percentile results are almost always worse than fixed rate results.

Here's a broad description of how PRC gets tested before it gets released:

First, I personally do 100% of the development.  I look for results that don't seem to make sense and then investigate anything that seems questionable.  Then I go through several standard test cases and compare numeric results to those from the previous version (which we now assume is correct, though there is room for error there, too).  I investigate any and all differences.

After I'm feeling pretty good about it, I turn it over to a couple of other engineers (who were originally customers).  One of them just goes through every screen and beats it up, looking for issues.  The other guy independently implements many of the formulas used to create the projections, particularly including the historical analysis algorithm.  Then, he compares his results to those from PRC.  Together, we investigate any and all differences and come to a resolution (sometimes he's wrong, sometimes PRC is wrong).

Verifying the Monte Carlo analysis process is the most difficult because it involves not only the normal projection algorithms but also market volatility simulation and boiling down the results from 500 test scenarios into something useful to a user.  This is done in a piecemeal manner.  First, the projection algorithms are tested by overriding the market volatility simulation inputs with fixed rate data and verifying that they produce the exact same results shown on the various projection pages (which are quantitatively verified by the other engineer).  Then, the market volatility simulation is tested independently, which involves testing to ensure the data actually does contain the mean and standard deviations specified on the Financial Assets pages.  Then, we put it all together and do a sanity check on the whole dataset created by the 500 test scenario executions as well as the translation of that data into the user-specified percentiles.  Finally, we run through a variety of user test cases with different percentile settings and make sure the results presented to the user appear to make sense.  This includes comparison to fixed rate and historical analysis results.  We investigate and resolve anything that doesn't seem correct.

Bottom line: The testing is fairly rigorous and is a combination of quantitative and qualitative investigations by three engineers.  One last thing that I do from time to time is to compare PRC results to other calculators (including ESPlanner) using a set of common test cases and then investigate significant differences, which involves having to figure out how someone else's calculator works.  Usually, I come away from this exercise more convinced than ever at the quality of PRC compared to the competition.  But does all of this prove PRC is 100% correct?  Absolutely not.  So, I use it myself on a regular basis for a variety of purposes and always investigate anything that looks questionable.  I also carefully investigate all questions sent in by customers, which does occasionally reveal errors.  When those are discovered, fixes are implemented and released immediately along with corresponding updates to the change log on the website.

 

Pralana Consulting LLC, Plano, TX

© 2017 Pralana Consulting.  All Rights Reserved.

PRALANA RETIREMENT CALCULATOR | The Mercedes-Benz of Retirement Calculators

 

New Visitor? Start Here.

ADVANCED PERSONAL FINANCIAL MODELING  |  MONTE CARLO & HISTORICAL ANALYSIS  |   INTERACTIVE WHAT-IF ANALYSIS