What do you think will be the Word of the Year for 2022?

My heart says it should be “Ukraine,” but my head says it will be “inflation.”

If I was a betting man, I’d bet with my head.

To be fair, one of the driving factors of 2022’s inflation is Russia’s illegal and immoral war in Ukraine. Specifically, it is the grain and gas shortages caused by the war and the international sanctions against Russia. However, that’s not the only cause of 2022’s inflation problem. It turns out that inflation’s rise in 2022 is a complicated story, one with many villains and few heroes, which also means it may be sticking around for some time. As a result, it’s an excellent time to analyze the vulnerabilities of a pension without an inflation-fighting mechanism.

Not The Beginner’s Inflation Article You’re Looking For

I’ve already written a lot about how inflation impacts your pension. In fact, I go super deep on the topic in my book, The Golden Albatross: How to Determine If Your Pension is Worth It. If you haven’t read my book, I suggest you purchase a copy or request it from your local library.

I’ve also written many articles on the topic. So much so that if I wrote another explanative article, I’d feel like I was rehashing old content. If you’re not interested in reading a whole book just to learn about inflation and pensions, then here’s a short list:

• • The Pension Series (Part 3): What is Your Pension Worth? — Read this article in total if you want a fundamental understanding of how inflation interacts with a pension.
  • The Pension Series (Part 4): Total Dollar Value (TDV) of Your Pension (Updated) — This article shows you how to calculate the value of a pension impacted by inflation.
  • The Pension Series (Part 7): How to Create Your Own COLA — A cost of living adjustment (COLA) is the primary tool used to dampen inflation’s impact on a pension. However, not every pension comes with one, so this article shows you how to create one.
  • The Pension Series (Part 29): The Golden Albatross vs. Women’s Pensions — This article explains why women should be particularly concerned about how inflation impacts their pensions. Spoiler alert! It’s because they live longer than men on average.

The above links aren’t the only articles where I mention or discuss inflation-pension issues, so use the search bar on this web page if you want more.

Show and Tell

Since writing another explanative article about inflation holds no appeal for me as a writer or value for you as a reader, I took a different tack for this post. I show you what inflation does to an early retirement plan built around a pension without a COLA. I do this courtesy of a reader codenamed Don Quixote, who is considering an early retirement lasting nearly 50 years.

Don asked me to analyze three payment options connected to his pension, which I modelled in Flexible Retirement Planner (FRP). FRP is free retirement planning software that allows users to layer their various fixed retirement income streams alongside their retirement investments. The user can then run simulations using different spending, investment, and tax strategies.

If that sounds daunting, it can be at first. However, running such models is worth it because they offer invaluable insights. In fact, I wrote an entire series of articles on the importance of testing your retirement plan based on how much insight I gained doing it for my retirement. So, don’t let the learning curve stop you. If my hairy, knuckle-dragging, Cro-Magnon-man math-brain can learn to use it, so can yours.

Don’s Situation

Don is a 41-year-old law enforcement officer from a big city on the US east coast with 17 years of service (YOS). When he initially contacted me, he was burnt out and looking to retire immediately instead of working three more years to reach his pension plan’s normal YOS. If Don retires immediately, he will forego a few pension-enriching options but would still receive a generous pension. The only hitch is that he’d have to wait three years before payments start in 2025. Also, the pension only comes with an extremely small COLA, which leaves it vulnerable to inflation’s effects.

It turns out that Don and his wife are good savers. For instance, when his pension plan presented him with the option to contribute extra early in his career, he took it. All totaled, he amassed $308,440 in extra pension contributions and investment earnings. That’s a serious chunk of money! As you’re about to find out, though, that’s not the entirety of his and Mrs Don’s impressive savings and investment achievements.

Interestingly, the pension plan gave Don a few different options regarding the extra $308,440. First, he could take out the entire lump sum and roll over approximately $65k to his (post-tax) Roth IRA and the remaining $243k to his traditional (pre-tax) IRA. At that point, his pension would total ~$75K annually and start in 2025. Second, he could leave the entire lump sum in the pension plan and collect ~$110K annually beginning in 2025. Finally, he could mix the two, take some out, and leave the rest. Don told me that if he chose that option, he would roll over the post-tax $65K to his Roth IRA but leave the remaining $243K in his pension. That would provide a ~$105K pension starting in 2025.

Don’s Details

Don originally asked me what he should do. He understood when I responded that I don’t give out financial advice because I’m not a financial professional. However, I told him I could analyze his pension options by modeling them and then let him decide what to do. He was happy with that.

Unsurprisingly, Don’s main concern was whether he would run out of money by taking one of these early retirement options. I told him it depended on his planned annual retirement spending compared to his other fixed income streams and investment portfolio returns. As a result, he sent me the following information for my modeling.

Estimated annual retirement spending: $84,000

Current savings:

• • Taxable portfolio: $132,000
  • Pre-tax (i.e., 401K, Traditional IRA) savings: $224,000
  • Penalty-free pre-tax (i.e., 457) savings: $475,000
  • Post-tax/Tax-free (i.e., Roth) portfolio: $110,000

Other future retirement income:

• • Mrs Don’s pension (2036): $13,835 
  • Mrs Don’s Social Security (2044): $12,624 
  • Don’s Social Security (2051): $32,424

First Impressions

Upon compiling Don’s information, I noticed two things. First, he has a generous pension coming his way, no matter which option he chooses. Second, he and his wife have saved and invested a lot of money for retirement. As a result, I thought the chances were small that the simulations would find challenges to his early retirement plan. I even told him that in an email.

However, as you’re about to find out, my initial impression was incorrect. Do you want to take a guess as to what factor proved me wrong? I’ll give you a hint … it’s the 2022 word of the year candidate I mentioned earlier. The one that isn’t a country.

Assumptions

When modeling future scenarios, assumptions must be made; otherwise, the model won’t work. That goes for the programmers as well. They couldn’t create a workable piece of modeling software without requiring some assumptions.

Let’s take Don’s longevity as an example. Don estimated he’d live to 90 when the Social Security Administration’s most recent life-span average for 41-year-old males is 38 years (i.e., 79 years old). Of course, we don’t know if Don will live to 79 or 90, but the software requires a number. I chose 90 because modeling longer-lived scenarios is the more conservative route. A long life span means a future retiree must save more so they don’t run out of money towards the end of their retirement. In Don’s case, that translates to modeling 49-year retirement scenarios.

A lot can happen in 49 years, especially around taxes and tax laws. The FRP software allows you to choose two rates, one for investments and one for income. That’s great, but those are averages for the effective rates over the entire scenario. Over Don’s projected 49-year retirement, he will transition from using his 457 and taxable portfolio money for the first three years; to adding his pension, his wife’s pension, and his wife’s SS; then taking required minimum distributions; and finally adding his own SS. How does a modeler boil all those differences into one effective tax rate?

The answer is that you do the best you can, make an educated guess, and then heed the words of the program’s developers:

Rather than viewing the output of the tool as a prediction of what will happen, it’s best to view it as support for making short term decisions about what to do over the next few years. Use the tool to guide these decisions but revisit them often and reevaluate them in light of any new information you have.

Don’s Assumptions

Here are the initial assumptions I used for Don’s three scenarios.

• • Inflation: 3.7% average with a 2.8% standard deviation
  • Tax rates:
    • Investment tax rate: 15%
    • Income tax rate: 12%
  • Investment returns: 9.8% with a standard deviation of 20.8%
  • Spending policy: Stable
    • Stable means Don’s family would spend the inflation-adjusted equivalent of $84K annually, no matter what.
    • This is the worst-case scenario since many families might adjust their spending based on investment returns.
  • COLA type for pension: None
    • In reality, Don gets a complicatedly small 1-3% COLA on the first $18,000 of his pension … starting at age 55.
    • He and I agreed it sucks.
    • There’s also no way for me to model something this complicated in FRP.

Initial Results: Scenario 1

As a reminder, in scenario 1, Don takes out the entire $308,440 lump sum and rolls over $65k to his (post-tax) Roth IRA and the remaining $243k over to his traditional (pre-tax) IRA. As a result, his pension annuity pays out $75K annually but doesn’t start until he turns 44 in 2025. Until then, his family would live off the money in his 457b and taxable accounts.

I’m about to display a screen capture of the results of scenario 1. Of the three scenarios, this one looked the most likely to run into problems. That said, Don and his wife had saved a lot of extra money. Plus, they had her pension and their Social Security checks kicking in later in the scenario. That’s FOUR fixed-income streams, two of which (i.e., Social Security) are inflation-protected. Here’s a screenshot of where I modelled those four streams:

If this scenario failed, I didn’t expect it to fail by much. By failure, I mean the investment money runs out before age 90. Obviously, the pensions and Social Security won’t run out. Still, it wouldn’t necessarily be enough to live on if inflation ran rampant over the entirety of Don’s retirement.

After all that context, you may be surprised to see that the scenario failed far more convincingly than I expected. In the end, it only scored a 62% probability of success. Here’s what it all looked like once I ran the scenario:

A Small Note on Simulations

It’s worth pointing out that FRP doesn’t make one modeling run and then declare success or failure. It actually runs 10,000 simulations per model. Those simulations use randomly generated stock market returns and inflation averages within the parameters of the assumptions I noted (above). As a result, no two simulations are the same. Once done, the software compiles the results on a year-by-year basis and produces a probability of success based on the number of simulations that run out of money compared to the number of simulations that don’t.

Maybe you’re thinking, “62% of the time means the scenario succeeded almost twice as much as it failed. I like those odds!” And, if you were in Las Vegas, you’d be correct. But I would advise caution and recommend you heed what the programmers wrote in the FAQ:

While on average (50% of the time) you may end with a high balance, successful retirement planning is usually based on much better odds for success, typically 90% or greater. 

Initial Results: Scenario 2

Since scenario 1 looked the riskiest when I eyeballed Don’s situation, I wasn’t panicking on behalf of Don when it failed. I was surprised to see it fail by that much, but I didn’t question my model. However, I didn’t have to wait long to start questioning the model because scenario 2 also failed. It, too, was not all that close because it scored a 74% probability of success. Check it out for yourself below:

Remember, scenario 2 included leaving the entire lump sum in the pension plan and collecting ~$110K annually in pension payments starting in 2025. Of all the scenarios, this one had the highest probability of success, given the size of the annual pension value. In other words, scenario 2’s pension annuity was worth $35K more annually than the $75K pension of scenario one. Furthermore, it was backed up by the same additional income streams (i.e., wife’s pension and Social Security) as scenario 1. Check those out here:

Initial Results: Scenario 3

Since I am naturally curious, I decided to run scenario 3 before messing with my model. In scenario 3, Don rolls over the post-tax $65K from his extra pension money into his Roth IRA but leaves the remaining $243K in his pension. That would provide a ~$105K pension starting in 2025. All other income streams remained the same. The results were similar to scenario 2, with a 74% chance of success. Here’s the screenshot of the results:

What in the Hell!?

As already mentioned, the initial results from scenarios 2 and 3 seemed counterintuitive. Therefore, I started double-checking all the inputs to ensure they were correct, and they were. I then went into the deeper parts of FRP to see if all the inputs functioned as designed, and they did. So, the problem wasn’t with the model.

At that point, I started to change the inputs one at a time and re-run the scenarios to see which one affected outcomes the most. I changed the investment returns, tax, inflation, and spending rates, all of which helped me determine causality. In doing so, I found that changing the inflation rate and its standard deviation affected the model the most. The changes didn’t need to be all that large, either.

An Inflation Rate Change

Given that the inflation statistics I used in the original scenarios were assumptions, it’s probably worth discussing why I chose them. I used the average US inflation rate (3.7%) and standard deviation (2.8%) from 1951 onward. I did this because, in 1951, the US Treasury and the Fed signed The Accord, “which ended interest rate controls and freed the Fed to use its monetary tools to control inflation.” Thus, 1951 onwards is seen as the modern era of the US Federal Reserve Bank’s inflation management. So, it seemed like a logical point of origin for the scenarios’ average inflation numbers.

However, in 1982, the Paul Volcker-led US Fed tamed the sky-high inflation left over from the 1970s. This ushered in a gradual trend towards lower average inflation and lower average inflation volatility, making it somewhat distinct from 1951 to 1982. So, I could’ve easily chosen the post-1982 statistics for the models’ inflation assumptions instead. Thus, when I went back and plugged in the average inflation rate (3.0%) and standard deviation (1.3%) since 1982, I got dramatically different results from the simulations.

The screenshots below show the change in inflation numbers. The new numbers increased the probability of success for scenario 1 to 75%, scenario 2 to 91%, and scenario 3 to 90%.

Why is This Happening?

How do minor changes to inflation assumptions impact these simulations so much? There are several answers, the first of which is time. Since the scenarios last 49 years, there’s a lot of time for the inflation rate to compound itself. Consider Don’s initial $84k in retirement spending to put this into perspective. It’s only $84K in that first year. After that, the FRP program increases annual expenditures at the inflation rate generated within the model to maintain purchasing power. This ensures that Don and his family can keep buying the same amount of groceries from the first year to the last.

In other words, $84K in spending at age 41 compounded annually at 3.0% for 49 years turns into $357,522 in spending at age 90. In the meantime, Don’s pension continues to pay either $75k, $110K, or $105K annually, depending on the scenario. It doesn’t adjust for inflation in the scenarios because he has no COLA worth mentioning. That means there’s a widening annual gap between the nominal dollar amount his pension pays, and the real dollar amount he spends in each scenario. In other words, even in the scenario with the largest pension (scenario 2), once Don’s pension starts, it goes from covering all the family’s retirement spending to roughly one-third by the 49th year. That’s a giant swing downward due to inflation!

Filling the Gap

As my two Gap Number articles chronicle, you fill the gap between your retirement spending and your retirement’s fixed-income shortfall with investment withdrawals. As already noted, Don and his wife have more income than just his pension; plus, they have all those investments. However, despite those resources, one-third of the simulations in scenario 1 and a quarter in scenarios 2 and 3 ran out of money.

If you’re familiar with Sequence of Returns Risk (SRR), you can probably guess what happens during the failed runs. Don and his wife’s investment portfolios are taking significant losses early in retirement, most likely when withdrawing the $84K annually over the first three years. As a result, the portfolios can’t recover — even with the extra fixed-income streams that start later in retirement. If you want to know why late-year income streams can’t stop the slide, check out the “Immediacy” section in Part 3 of the Pension Series.

I also suspect those losses in the failed scenarios were exacerbated when the FRP software randomly but simultaneously generated high inflation rates during the down years. The double whammy of higher inflation and lower investment returns can devastate an early retirement plan built on a pension that’s not inflation protected.

Don Can’t Control Inflation, But …

As a result of these simulations, you may think that poor Don is doomed to work three or more years to earn a bigger pension since he can’t control inflation. That’s true; he can’t. But he can control something that inflation also directly impacts. Care to guess what it is?

Beuller? Beuller? Beuller?

If you said “spending”, then you’re definitely paying attention. The truth is that changing the spending input on all of Don’s scenarios was just as powerful as changing the inflation input. This makes sense because, as I just got done discussing, the program automatically increases spending by each year’s inflation rate. As a result, if you cut down on spending, you cut down on inflation’s impact on your spending over the entirety of the scenario.

A Change in Spending

If Don budgeted $75K for retirement year one and adjusted spending annually to account for inflation, at a 3% inflation rate he’d only spend $319,216 at year 49 (i.e., age 90). That’s about a $35K difference in final year spending from the $84K annual budget I mentioned above. More importantly, it’s money saved annually by spending the inflation adjusted equivalent of $9K less in the annual budget.

Thus, it’s no surprise that when I loaded $75K in annual spending into the FRP scenarios and re-ran the simulations, they all returned higher probabilities of success. I didn’t have to change the inflation figures either; I used the 3.7% average and 2.8% volatility from the original attempts. Upon doing so, scenario 1 returned a 78% chance of success, scenario 2 produced a 92% chance of success, and scenario 3 produced a 91% chance of success. Check out the screenshots below:

Simulation vs. Real World

Lowering spending in a simulation is much easier than dropping it in the real world. So, again, you might think that Don and his family may not be able to live on less. Fortunately, when Don sent me all his details, he said that his annual spending was anywhere from $6K to $7K per month or $72K to $84K annually. Being a worst-case planner, I used the $84K number. However, it appeared that Don and his family could live on less if needed. When I reported what I found, here’s what he told me:

I think it is clear that we need to reduce our spending… I figured let me give [you] the MAX we would spend in a month, and $7k is definitely on the higher end, especially without a mortgage payment. I think $75k per year is very feasible for us.

One Final Note of Caution

Well, that’s that, right? Don and his family need to avoid taking the entire lump sum and not spend over $75K (or its inflation-adjusted equivalent) annually. If they do that, he can retire tomorrow, live for 49 years exactly, and everything will be great no matter how big inflation bites, correct?

Yeah, not so much. The simulations help determine the damage inflation can cause to the purchasing power of a no-COLA pension over a long retirement (i.e., from covering all spending to covering only a third). However, they are far from a guarantee of success. That lack of certainty comes partly from the assumptions I discussed earlier. They’re necessary to run the program, but they inject a lot of uncertainty into the model. Will Don live to 90, 78, or 48? No one knows. Will inflation drop to historical norms after 2022? Possibly. So, remember the words of caution I quoted from the developers earlier in this post! Use the simulations as a guide in the early years and re-run them often as you get new data.

The Program Used Also Matters

Furthermore, FRP isn’t the only retirement planning software out there. Many high-power retirement calculators use mechanisms like FRP to generate random stock market returns and inflation. Others use historical trends as a guide. There’s no right or wrong method, just what the user is most comfortable with. So, I don’t want anyone to think that FRP has a magical corner on the market.

In fact, here’s a link to the most comprehensive review of retirement calculators on the internet. You’ll see that FRP is but one of many in the “best” category. Even then, the authors of that post warn you against falling in love with any one calculator.

I just like FRP for the ease with which you can model multiple fixed-income streams alongside investments. In my opinion, it’s ideal for future pensioners. And, no, I don’t have any business arrangements with FRP’s owners.

A Promise Fulfilled

At the beginning of this article, I promised to show you what inflation does to an early retirement plan built around a pension without a COLA. To do that, I ran multiple scenarios on Don’s 49-year retirement plan using two different inflation numbers. In the end, a slightly lower inflation rate and inflation volatility rate drastically improved Don’s forecasted outcomes. But, since you can’t control your nation’s inflation rate, I also showed you how lowering spending achieves the same effect. As I told Don, it’s up to you to use that information wisely.

Speaking of Don, my FRP-based simulations fulfilled their primary purpose. They showed him that scenario 1 courts the most risk due to inflation. Therefore, he now knows to avoid scenario 1 and concentrate on scenario 2 or 3. That assumes he’s looking for the highest chance of success over a long retirement.

Don’s long 49-year retirement may not be in the cards, though!

Last I heard, he’d caught his second wind and intended to finish his entire 20 years. That’s cool, but he’s got my analysis in his back pocket should things turn sour again. That said, working three more years would help a lot. It will significantly raise his pension value and avoid those large withdrawals from his investments over the first three years of retirement.

I don’t need a model to tell him that should work.

Although … based on my last guess … maybe I do.

OK, if I were Don, I’d check every few years … just in case inflation is still rampant.

---

Your Help

Do you enjoy the free content on this site? Help keep The Golden Albatross Blog running by choosing one of the three options below!

1) Buy the website a beer on its Buy Me a Coffee page. All beers “bought” go straight to keeping the site running ad-free.

2) Purchase The Golden Albatross: How to Determine If Your Pension Is Worth It from ChooseFI publishing. All royalties from book sales also go directly towards keeping the site up and running.

3) Re-post links to my articles on your social media feeds, and message your favorite personal finance podcast or radio show to suggest they cover my content.