Grumpus The Confessor
I have a confession to make. I put off writing this post for a while. When I first started my blog, I had always intended to demonstrate how to test your retirement plan. I wanted to do this by using a high powered retirement calculator. Doing so would complement what I consider the biggest strength of my website: the series of practical “How To” retirement plan articles in the Planning section. However, I needed to tackle some other topics first. I wanted to walk financially novice readers up to a point where they could understand the subject matter of this article. Yet, I essentially hit that point weeks ago, and still, I delayed.
Part of that delay was due to the complexity of what I intended to describe. It’s hard to write effectively about the steps needed to test your retirement plan. A technically savvy blogger would simply post a video of how to do this, but that is beyond my capability at the moment. As a point of reference, I was happy enough when I figured out how to embed a spreadsheet into this post. Maybe someday I will circle back and create a video once I obtain the skills, and find the time.
Another cause for my delay was pure procrastination. Procrastination is one of my fatal flaws. Or it would be if I ever got around to the business of dying. I might one day, but not today.
That revelation aside, in order to write this post I needed to update my retirement plan with my 2016 spending data. I also needed to re-test all my scenarios using my preferred high powered calculator, Flexible Retirement Planner (FRP). It’s not an overly arduous task, simply tedious. Since my 2016 spending didn’t change significantly, I knew the probabilities of success for my retirement plan’s scenarios wouldn’t change much. However, it needed to be done, and this post was meant to be the impetus.
The more personal reason for my delay was the intended use of my financial numbers in the article. I know a lot of the bloggers in the Financial Independence (FI) space like to post Quarterly Summaries of their expenses, portfolio, and net worth. Most bloggers are probably doing this with good intentions. They want to prove that they live and invest the same way they advocate. However, it tends to turn me off. At best those posts seem like a challenge, a metaphorical throw of the gauntlet, meant to exhort their readers to something better. At worst it smacks of bragging and may explain why some bloggers who do this face significant blow-back, scrutiny, and judgment from their audience.
I don’t want any of my three non-related, yet highly devoted, readers to consider my finances as a yardstick. Personal finances depend too much upon individual circumstances to make a comparison of one person’s situation to another useful. It is not my intent to brag about how much I earn (which is a subject of public record as a military officer) or save. Nor is it my intent to subject myself to the scrutiny of what I perceive to be some of the worst elements the internet offers.
More to the point, I never proclaimed to be anything other than a moderate to good saver. I do not consider myself, or my family, overly frugal. We definitely are not minimalists. Our savings rate, which typically ranges in between 25 and 35% annually, rarely tops 40%. That is a statistic which puts us in the middle to back of the pack when it comes to FI bloggers.
Despite this, and possibly against my better judgment, I decided to use my own projected retirement expenses to demonstrate how to test your retirement plan. I did this mostly because it proved too hard to recreate realistic spending with made up numbers. Thus, I need to add a caveat prior to moving onto the demonstration portion of this article.
The expenses in the attached spreadsheet (below) were based on my family’s spending from 2014 to 2016. We were stationed in Europe the entire time. Cost of living in Europe was high. So much so, that we received a cost of living allowance. This skewed a lot of our spending numbers upwards. As a result when calculating chances of success in our preferred retirement location (Southern California) I had to invoke a 12% cost of living reduction (I found the value online) for some of my spending categories. Doing so adds a level of unreliability I would prefer not to have in my calculations. Yet, I have no choice. Going back further in our spending history raises realism issues as well. On the plus side, you get to see how to work Geo Arbitration calculations into testing your retirement plan.
Prior to moving onto the calculations, it’s worth noting this article is a continuation of a previous post. Thus, the “Part 2” in the title. If you did not read Part 1, you may want to. It was my attempt at a primer. I want your mindset correct before testing your retirement plan using a high power retirement calculator. On the other hand, if you are familiar with using high powered retirement calculators, and their limitations, you can probably get by without reading Part 1.
Regardless of whether or not you chose to read Part 1, there are some other things you must do. Tracking your expenses for a certain length of time is a must. If you haven’t read my article on how to track your money, start there. Without accurate insight into where your money goes, and how much you plan to spend in retirement, any retirement plan is at best a guess. The less accurate your spending data is, the worse that guess becomes. At some point, it turns into what we in the military call a SWAG — Shitty Wild Ass Guess. Don’t SWAG your retirement plan’s numbers, we don’t want you running out of money at age 70.
The second thing you need is a draft retirement plan. Again, I wrote a two part post on how to build a retirement plan using the GRO2W model. If you’ve no idea how to start building a plan, start there. If you don’t like the GRO2W model, there are plenty of other methods on the Interwebs. I’ve no doubt that you can find something to suit your style. Just make sure your numbers are as accurate as possible. Again, no SWAGing.
Finally, you will need a high powered retirement calculator. If you have a favorite high power calculator, use it. For anyone in a Golden Albatross situation (i.e. working towards a pension), make sure whichever calculator you use has the ability to add fixed income streams like pensions, annuities, and Social Security at the appropriate age. I prefer FRP, but there are plenty of others worth your while. I do not, however, suggest a simple calculator. It will not provide you the fidelity you need for the calculations we are about to embark upon.
If some of you are lost at the moment, probably best to go back to Part 1 of this series, and read the homework. I assigned the task of reading through a series of articles about retirement calculators written by my friend Darrow Kirkpatrick of Can I Retire Yet?. Darrow made retirement calculator reviews one of the cornerstones of his website and his book. I’ve yet to see anyone come close to the breadth and depth of his analysis on the advantages and drawbacks of the calculators available today.
Let’s stop there to give you a chance to collect those items. Much like a university science class, we are about to head into a laboratory setting. Make sure you got the tools listed above, and then follow me inside.
Grumpus the Testor
Please download the spreadsheet linked above. Make sure you run it through a virus scanner. It was clean when I uploaded it, but it’s better to be safe than infected. With that done, open the spreadsheet and take a look at the different tabs. As you scroll through the first two the tabs, you will notice two things. I’ve calculated both my fixed retirement income and my estimated retirement expenses. These bits of information are crucial to testing a retirement plan. I do not need to calculate my investment income because my calculator will do that for me. Yours should too. If it doesn’t, get a new calculator.
I broke out my fixed income according to two different scenarios identified in my GRO2W plan. In one scenario I serve in the military for 21 years, and in the other, I serve for 23 years. 21 years earns me all my benefits including the ability to transfer my GI Bill to my kids. 23 years of service provides an extra cushion of retirement income, which I may or may not need, depending on my retirement scenario.
I also account for Social Security kicking in. Whether or not Social Security will be available for Gen Xers and Millennials, is a much-debated topic in financial circles. Some people choose to ignore it in their retirement planning. That’s too conservative for my purposes. Although, I doubt it will be there in its full capacity when I reach disbursement age. Based on my research, I’ve planned for about 75% of my earnings to be available. To provide extra cushion I also planned for it to kick in at age 70, vice 67 (my current estimated age of availability according to the website). You should use a calculation that accurately represents your likelihood of receiving Social Security.
Now to the spending tab. A quick scan reveals various spending categories based on a three year rolling average of my family’s spending. Some categories are sub-totaled separately from others. I did that for two reasons. First, as previously noted, I had to account for the cost of living (COL) difference in my annual retirement spending from Europe (where I was stationed) and Southern California (my intended retirement location). Thus, I segregated my expenses subject to location inflation and reduced them by the COL difference. That percentage happened to be 12% according to the website NUMBEO. There are other websites that do the same thing, but I like NUMBEO’s comparison tool. This is essentially how you Geo Arbitrage your retirement plan as well if you are so inclined.
My second reason for subtotaling my spending deals with the number of different scenarios I chose to test. I chose those based on options identified during my GRO2W retirement planning. In one scenario, I test buying a house outright upon retirement. Therefore we would have no mortgage payments. In another scenario, I test the notion of keeping our current condo with tenants in it, while renting a house ourselves. In my other two scenarios, I test selling our condo and becoming renters for life; as well as selling our condo and buying another home with a mortgage.
I tested four scenarios in all. To make matters more complicated I chose to test them using my two different proposed fixed income streams of 21 and 23 years of military service. Finally, I chose to test both the full fat European based spending average and the reduced Southern California COL average. Doing so allowed me to judge my spending cushion. All in all, I made four calculations per tab and had four tabs worth of calculations. That made for a total of 16 different calculations. Don’t expect to run as many different scenarios as I did. I just like to explore my options.
Time to crunch all those numbers. Let’s look at which categories most high powered retirement calculators look for. A good retirement calculator possesses the ability to compute figures for the remainder of your wealth accumulation stage (i.e. your working years), as well as your wealth preservation stage (i.e. your retirement). Thus, a good calculator should ask for your current age; annual savings and investment amounts; portfolio balance; investment vehicle allocation (i.e. tax deferred, Roth, and taxable); and how long you intend to work. A good calculator will also ask for your intended annual spending. It should also allow input of your fixed income streams at their various age of disbursement.
Most calculators worth their salt will require an inflation input. You may be able to click a button for historical inflation averages. It’s your decision if you want to use that feature. Many, if not most, calculators will also ask for your expected portfolio returns. You may or may not have an option to allow the calculator set your expected average rate of return by using historical averages, or generic risk descriptions (i.e. aggressive, conservative). I would caution against its use. Since most calculators don’t know your investment allocation (i.e. stocks, bonds, gold, cash, etc) or investing philosophy, it is best left to you to estimate.
And make no mistake, it is an estimate. If you are not comfortable estimating your portfolio’s rate of return, I would advise you pause your testing and do some research on the issue. This is one of the most important values you will input in order to test your retirement plan. I can’t overstate the value in understanding the issue and having a level of confidence in the value you provide. This may be a topic I delve into more in future posts.
At this point, you need to input all your various numbers into your calculator and run your simulations. You may want to run different scenarios like me based on different income streams, savings rate, or nest egg totals. Or you may only have one scenario to test at the moment. Either way, this is where the rubber meets the road, so again, make sure you understand the effects your various inputs have on your probabilities for success.
I would suggest you at least experiment and see what different inflation and portfolio returns do to your plan’s chances of success. It may not seem like it, but one slight change can make a huge difference in outcomes. Ultimately, I chose a 3% inflation rate (with a 1% standard deviation). That estimate is in line with historical averages. I chose a conservative 5% return for investments (with a 12% standard deviation). That makes for an effective return of 2% plus or minus the effects of standard deviation.
Overall, that is a fairly conservative scenario to choose as my base model. However, I would rather model a low return environment to ensure my plan can work under stressful conditions. If real life returns turn out to be higher, or the inflation rate turns out to be lower, all the better for my plan’s real world chances of success.
As previously mentioned, my preferred calculator is FRP. It allows you to input all the above categories and more. Specifically, it allows for some flexible spending options, the ability to include lump sums income/payments, and the ability to designate spending order of priority for your investment accounts. I’ve yet to dream up a scenario FRP could not handle. Of course, as Darrow Kirkpatrick points out, FRP has drawbacks. I specifically dislike its use of the Monte Carlo scenario over historical stock market data. Overall though, for a free calculator, it works extremely well. There are others, and again I would refer you to Darrow Kirkpatrick’s reviews to find one that best suits your needs. Finally, for any novices to financial calculators, the output will ultimately look something like this:
So what does all this work tell us? Well, if you look at my spreadsheet again, you see that only 7 of my 16 calculations scored above a 90% probability of success. A 90% probability of success, is my cut off point for considering an option viable. 5 of the 7 viable options were in the 23 years of military service pay band (not ideal). Only one scenario showed a high probability of success under all four pay and spending options. That was purchasing a house outright upon retirement. If I wanted to improve the number of viable options, without working longer, I would need to look at other cities and states with a lower COL.
However, Southern California is where we are happiest. As a result, my and Mrs. Grumpus’s primary plan is to purchase a house outright upon my retirement from the military. Doing so will allow us to live a lower cost lifestyle. It will place our annual expenses well within the ability of my retirement pay (from 21 to 23 years of service) and our investments (at a 5% or higher average rate of return) to cover. We’ve yet to decide on whether either of us will work after military retirement. However, it is nice to know the option to never work again is there if needed.
These calculations also tell us that we should be engaged in cash accumulation. Thus, I am socking the money we save monthly into high APR savings accounts and 18 – 36 month CDs. The only retirement investments we make are our annual Roth IRA contributions. Anything else locks too much money away or puts it at needless risk. We need the cash to be there in three to five years, not riding the highs and lows of the stock or bond markets. The only question now is how long it will take to accumulate the cash we need to buy a house outright.
I should add that our plan includes selling our current condo in Southern California. Which means we have some principle already built up to assist in purchasing our new house. Even with that, there are still a lot of variables in play. However, they are far fewer in number than prior to testing our retirement plan. Hopefully, the story will prove the same when you test your retirement plan.
Alas, I leave you with this reminder. These calculations are only probabilities of potential future outcomes. The future, by definition, is unknowable. Do not place too much faith in any one number or calculation — no matter how good you believe the calculator to be. Understand the biases and drawbacks that underlie the calculator you choose. Continue to engage in routine tests of your plan as circumstances in your life change. Much like the Pirate’s Code and the concept of parlay in Pirates of the Caribbean, consider the results of your calculations more like guidelines, and less like rules. Do this, and I have no doubt your retirement plan will be a success. Good Luck!
***Grumpus Maximus is an Amazon associate. See Disclosures for more details***