Lifetime Planner and other Retirement Tools

gtamm

Has anyone tried to reconcile Lifetime Planner Portfolio $ values to another type of retirement planning tool (i.e., Flexible Retirement Planner, Ontrajectory, etc.

My attempts using the same set of assumptions usually result in 15-20% delta in the 20-30 year out range ending balances of the portfolio.

It could be average verse median balances, but would think if would be closer.

Jim_Harman

Quicken's Lifetime Planner assumes fixed rates of return while some other tools use a statistical Monte Carlo analysis. That might account for some of the differences you are seeing.

Chris_QPW

Well all I can say on the subject is that I have been able to go through the detail numbers Quicken uses/generates and they seem correct, but in fact other than a ballpark figure I don't use it because their assumptions don't line up with what I intend to do.

gtamm

I think the Monte Carlo tools are also valuable, but I like LTP for the level of Income/Expense data.

Are there specific assumptions that LTP use that are different than what you would plan to do?

Chris_QPW

Well the biggest weaknesses/differences between what I would do and Quicken are in the tax rate and where money is withdrawn at a given time.  And also the investment returns being the same for all the assets so you have to guess at an average.

You can only put in two tax rates, one for before retirement and one after.  But in reality your tax rate is going to be dependent on your "income".

I certainly do not expect my tax rate to be the same all the way through my retirement.

And that is where it gets important of which accounts you are going to withdraw the money for your expenses at any given time.

To me I break my assets down into three major categories.  Savings, Roth IRAs, and Traditional IRA/401K accounts.

When you are retired or (even semi-retired like I am) or have a business you get to play by the "rick man's rules" provide you aren't just living on Social Security.  In other words you have at least some control over what is called "income" by the IRS.

And as such which account type you might want to take money out of at a given stage in your retirement can vary well change, but Quicken has a fixed set of rules of what account types to take money of and it is a "exhaust all funds in that type before moving on to the other types" kind of rule.  Which is not really that likely.  For instance it hits your taxable accounts first.  It is unlikely you would reduce your taxable accounts to zero before drawing money from say your Traditional IRA/401K.

gtamm

Great insight...Thanks!

luce1234

Why is the lifetime planner still using the old RMD date instead of the new 72years old?

Scooterlam

@gtamm

A few thoughts....

While I don't use this particular tool (pralana), I found the article useful to understand some facets of retirement planner models:  https://pralanaretirementcalculator.com/html/calculator-design.html 

As far as reconciling the differences between results of different planning tools, that will be a challenge.  Some will say, its a waste of time to try to compare tools.  But IMO its a challenge that will pay off in a far better understanding of how these tools work and a far better understanding of financial planning in general.

As other have stated, its a model providing an estimate 30 years or more in the future!   There will be assumption error, modelling error and external risks that influence your results.   It's important to keep your assumptions current, validate the model and keep an eye on external risk factors (eg. inflation - see recent Federal Reserve comment regarding the 2% inflation target)

@Jim_Harman points out to be sure to compare apples to apples when it comes to the calculation methods used in the tools you compare.  If you say you are comparing average to median results, then you just might be comparing a "fixed rate" model like LTP to a monte carlo model - is that the case?

Assuming your are comparing models with like calculation methods and you affirmed that your input assumptions are the same, my next step would be in carefully inspect the result reports / tables for expenses, income, account balances to see where the divergence begins. From there, work backwards to try to figure out how the model works.   You may likely find one or more of your assumptions inconsistent with how the tool expects you to enter it.  But first, make sure you have exhausted the tool owner's documentation!!! 

Off the top of my head.....specific things I found in doing this exercise, that have caused sometimes marked differences in outcomes:

• Income and expense timing differ as to when they are taken (start/end of year, proration)
• Capital gains tax using actual or assumed cost basis
• Average tax estimate assumption on ordinary income vs. actual tax calculation from tool
• Application of SS taxes
• Application of RMDs
• Spouse model assumptions
• Order of account draw-down
• Healthcare inflation taken at a higher rate than living expenses
• Use of basis in tax-deferred accounts (NOT cost basis)
• Tax treatment of various tax-free and tax-deferred account types (IRA, HSA, ROTH etal)
• SS income modeling (age and amount, increases, spousal benefits, survivor benefits....)
• Use of actual asset allocation for historical rate of return assumptions
• Inflation assumptions set in model versus user input
• Result reports in current dollars v. future dollars
• Result reports showing average results vs. median results.
• Monte carlo return assumptions using historical return and inflation data version your own.
• more....

LarPet

Interesting thread. I am also trying to reconcile results against other planners. It would be good to know how the Quicken Planner managers specific calculations. For instance, how is SS income taxed? There is a formula in there somewhere based on something, but I haven't been able to figure it out. And the sequence and amount of withdraw from various accounts. I couldn't agree more with Scooterlam, understanding these details is really important. Just seems doing reverse engineering isn't the best way to learn them. There should be a detailed users guide somewhere. Perhaps there is and I missing it?

Scooterlam

@LarPet Quicken Help and the Quicken Community are your best bets to understand LTP better   Quicken help can be a challenge to navigate, IMO.  

I found that most planning tools don't provide specific formulas but (sometimes) detailed feature description of their capability.  Ex.   The formula for SS taxes is well known and published,  the fact it is or is not (and to what extent),  applied in a particular tool is useful to know.  Perhaps the Excel based planners are more transparent to this level of detail.   

You could create a separate idea post to outline your specific needs on an enhanced user guide for LTP.
  
Also,  please consider voting on the "LTP ideas and bug report thread" and let Quicken know your need for them  to improve LTP:  https://community.quicken.com/discussion/7713110/lifetime-planner-bug-and-idea-list-make-yourself-heard#latest

mshiggins

LarPet said:

Interesting thread. I am also trying to reconcile results against other planners. It would be good to know how the Quicken Planner managers specific calculations. For instance, how is SS income taxed? There is a formula in there somewhere based on something, but I haven't been able to figure it out. And the sequence and amount of withdraw from various accounts. I couldn't agree more with Scooterlam, understanding these details is really important. Just seems doing reverse engineering isn't the best way to learn them. There should be a detailed users guide somewhere. Perhaps there is and I missing it?

There is a lot of detail in the LTP Help. For example:

Chris_QPW

I thought I would look into the details of what the Lifetime Planner is doing as far as taxes on Social Security are concerned, and sure enough it is extremely simplistic.

You have to hope that you fall into "what most people will have happen", and even then I think at best you can only predict "what is happening now", not the future.

Let me demonstrate what it is doing and at least one failure scenario.

Quicken uses this formula (for all cases) to calculate the taxes on Social Security income:
Social Security Income * .85 * retirement tax rate.

I'm going to use married couple for the rest just to make it easier.

First off the "real formula":
taxable amount for below =  other income + 1/2 social security income

Below $32,000, no tax.
between $32,000 and $44,000 tax 50%.
$44,000 and above tax 85%

Quicken is using that last line.

Now for some scenarios.  only is Social Security income or other income comes from non taxable income like a taxable/savings account or Roth IRA.
In other words "other income" from the "real formula" is $0.  Note this is where picking to get your income from a Roth IRA or a Traditional IRA/401K makes a difference, and you might not draw from the same type of accounts Quicken does at any given time.

If 1/2 of you Social Security Income is less than $32,000 ($64,000 total) then you wouldn't owe any tax, but Quicken will still apply its formula above.

And imagine this one, $50,000 social security plus $10,000 from other income like a Traditional IRA/401K.
$25,000 + $10,000 = $35,000

Tax on Social Security = $50,000 * .5 * your tax rate.
(Quicken would do $50,000 * .85 * your retirement tax rate)

My wife and I live quite modestly, have the house paid off, and such, so I can easily see us living on less than $32,000 ("other income" + 1/2 Social Security) at least until we are forced to take money out for RMD.

But it is exactly that point where Quicken's calculations are really going to go wrong for us.  Between when we take Social Security and when we have to do RMD I could set the retirement tax rate to 0%, but that would be the same rate used for after we are forced to take out RMD, which of course will change the tax rate considerably.

And if you were a person falling into the $32,000 to $44,000 and Quicken is still using the .85 when you need .5 it is going to be really hard to pick a tax rate that will get you to that correct amount.

Moral is with the Lifetime Planner you have to hope that you fall into the "normal case", and even then your tax calculation will fit what you are doing, but most likely your tax rate will change when you are force to take out RMD, and there is only one "retirement tax rate".

Scooterlam

Good points and illustrations @Chris_QPW.  I see the same behavior in my test data file as well as my actual data file.  I got curious as well ! 

Would be interesting to hear the rational for the simplified model from Quicken.

I get that some complex models would/should be simplified that result in errors that create either a pessimistic or optimistic results, for a given set of inputs - provided one understand the errors that are introduced.  The question is "does it matter?" or "is it good-enough"? Does the impact on cash flows throughout the plan or ending balance cause the plan to fail, when using the Quicken model vs. the real model?   For sure, it makes the eyes glaze over when considering a cumulative effect of all the assumption and modeling errors introduced in these types of tools!

"All models are wrong; some models are useful." George Box

The answer, of course, is not so easy....So, I tried to get a sense for the differences in the two models (Quicken vs. Real) using a spreadsheet and a few scenarios of AGI, SS benefits, and RMD  I tried to keep things isolated to both model's tax impact over the life of the plan.  I'm sure I've introduced some slop, myself....

As you might expect, at lower income levels, the Quicken model gives a more pessimistic result (that is, you're taxed higher levels than expected).  At higher income levels, the two model results are the same.   And, at some place, there is a break-even point.

In the next few images, I ran a few 30 year scenarios for a MFJ couple.   Used both the Quicken formula and the IRS "real"  formulas.  Assumed that real growth for SS of 0 and real growth of AGI (other income) as 3%.  "real" means growth minus inflation.

Note too that the IRS formula is progressive whereas the Quicken formula results in a "cliff", as previously pointed out.    As an aside, I found it interesting that the 50% and 85% thresholds (and their $ ranges) have not been adjusted since introduced in 1983 and 1993, respectively. That is, there have been no adjustments to the $ ranges over the last 37 and 27 years respectively!  Kind of a built-in tax increase on social security incomes?  My understanding anyway.

FWIW, here is a good online SS benefits calculator for you to run your own numbers.
https://www.fool.com/retirement/2016/06/06/social-security-tax-calculator-are-your-retirement.aspx

Anyway....here are a few scenarios and their results.  Right-click on the image and open in new tab to make it bigger.

1.  High SS, Low AGI, No RMD
Result:   Highest delta between Quicken and real models > $158,000 v. $17,975.  An up to ~$5100 yearly, extra expense due to Quicken's tax model, for up to 30 years, can be dramatic.



2.  Mod. SS, Mod. AGI, No RMD
Result:   Second highest delta between Quicken and real models > $94,860 v. $22,535.  Not so different than Scenario 1.



3.   Scenario 2 + Mod. RMD at age 72+
Result:  Quicken and real models closing > $94,860 v. $ 64,826.  RMDs increase provisional income at 72+ and substantially reduce the differences between models.



4.   Mod. SS, Higher AGI to find break-even point for this scenario.
Result:   Quicken and real models produce same result at this and higher levels          of provisional income.



5.   Scenario 4 + Mod. RMD at age 72+.
Result:   Quicken and real models produce same result as Scenario 4.



Was this exercise good?  For me, yes, learned something about SS benefit taxation, history and this particular modelling behavior in Quicken LTP.  How will I used these results?  No immediate adjustments to my LTP assumptions since, given my scenario, the differences are a "don't care".  BUT,  given the various "tax torpedos" coming my in the next 8 -18 years,  I may need to have a look at the benefits of doing some Roth conversions, for example, while there is time to do so.  YMMV. 

markus1957

To the extent an RMD creates a significant change in tax rate, a workaround for a current retiree is to change the (LTP) retirement year (age) to the RMD year (age) and adjust tax rates and investment return using the before/after retirement options for those parameters.  A quick test showed no negative effect that stood out.

For a simple tool like LTP, the workaround is more efficient than trying to make the LTP more complicated with RMD refinements built in.

Chris_QPW

Scooterlam said:

BUT,  given the various "tax torpedos" coming my in the next 8 -18 years,  I may need to have a look at the benefits of doing some Roth conversions, for example, while there is time to do so.  YMMV. 

Personally I have looked into Roth conversions, and other factors like should I take my Social Security "early", "at full retirement", "or as late as possible" and have come to the conclusion that it isn't something that I'm going worry about.

Let me start by saying that it seems that most/some people's goal seems to be to maximize the amount of money they get.  Mine isn't.  Mine has always just been "have enough money", so my methods certainly will not fit into a lot of people's plans.

I guess if a person is young enough a Roth conversion makes good sense, but not as good as one might think. What a lot of financial planners forget to mention is that you are weighing having more money now to invest vs taxes when you get to RMD age.

It is far from clear in my mind that Roth IRAs always win out.  What's more as you get closer to retirement and are just thinking about doing the conversion to avoid the taxes for when you get to RMD age it gets even more "iffy".  Because the only way it will matter a lot is if you are in a much lower tax bracket now then you would be at RMD age.  But as you take out money for the conversion your tax bracket goes up.  And what you are actually saving is only the taxes on the difference in the tax brackets.

To get that you basically have to an income (as defined by the IRS not what you spend) of about what your Social Security would be or maybe even less.  And really for most people that will only be a few years between taking Social Security and their RMD age.  Bottom line for me I took $10,000 out one year only to get some extra income that year, and have to reverse it.  And let's face it, even taking $20,000 out for say 5 years, and getting just the tax difference is going to be a pretty small number over your lifetime.

And that brings me to the age old question of when to take Social Security.
Well one has to understand how they setup the tables of what you get at any given age.  They set it up to "be fair".  As in for a person living to the average age it wouldn't matter if you took it at 62 1/2 or 70 1/2 you would get the same amount.  Now that isn't exactly correct these days because of the fact that those tables haven't been update since they set them up.  So with people living longer you do tend to get more money if you wait longer.

But then that brings me back to my "have enough money".  Let's face it we are not talking hundreds of thousands of dollars here.  We are talking at most tens of thousands of dollars over your lifetime.  And here is how I see that playing out.  The people that, that amount of money would be critical to, have no choice.  They will be the people that have to take Social Security at 62 1/2 because they need it to live on.

For the people that can wait, it is just "maximizing".

So in the past the engineer in me wanted to "find the best/maximum", but now all I care about is being sure I have enough, and that is a whole lot easier to be sure of and you don't have to make such wild guesses at what it going to happen in 20 to 30 years.

Chris_QPW

BTW on this statement "What a lot of financial planners forget to mention is that you are weighing having more money now to invest vs taxes when you get to RMD age."

All the financial planners (TV ones at least) have seen seem to make the assumption that taxes will be much higher in the future.  After all inflation, governments out of control, right?

Personally over my 63 years, I have not found this to be true.  How much taxes one pays isn't just going up like inflation, it moves more by "political wind" then anything else.  Nothing that I can predict what it will be in 30 years.

And if you can't "guarantee" higher taxes in the future then the Roth tax savings vs being able to invest it now becomes a whole lot more murky.