• 👉 Work With Me
• JP Morgan Research:
  • Three Retirement Spending Surprises (2019)
  • Three New Retirement Spending Surprises (2025)
• NEW RESEARCH: Your Plan Is Overestimating Retirement Costs (By 20%)

Last week, I shared new research showing that for the typical retiree, inflation-adjusted spending declines steadily throughout retirement. And more importantly, that modeling spending the way retirees actually behave — rather than the way our software assumes they behave — could support roughly 20% more starting retirement income.

Well, today, I’m continuing that conversation. Because it turns out that one of the largest banks in the country has been studying the same question from a completely different angle.

Instead of relying on surveys and interviews, JP Morgan analyzed the actual, anonymized transactions of more than 5 million American households. That means they were able to watch what happens to real household spending as people move into and through retirement.

In addition to supporting the conclusions we covered last week, their analysis also captures things that are hard to see in academic data and reveals what they call the three biggest retirement spending surprises: a lifetime spending curve, an unexpected spending surge, and spending volatility hiding beneath the averages.

As you’ll hear, these three patterns can mean the difference between a nest egg that lasts through a full 30-year retirement and the very same portfolio, facing the very same market, running out four years early.

Welcome to another episode of the Stay Wealthy Retirement Show. I’m your host, Taylor Schulte, and every week I cover the most important financial topics to help you stay wealthy in retirement. Ok, onto today’s episode.

3 Retirement Spending Surprises From a Study of 5 Million Households

Before we get into the three surprises, it’s worth sharing what makes this research different from the study we explored last week.

Last week’s findings came from David Blanchett, one of the most widely cited researchers in retirement income planning. As shared, his thoughtful analysis was built on the Health and Retirement Study — a university-run research project that has tracked thousands of older Americans for decades now. It’s a rigorous and well-respected dataset, but at the end of the day, it still relies on people reporting their own spending. And if you’ve ever tried to reconstruct your spending from memory, you know how imperfect that can be.

Today’s research comes from JP Morgan, and it doesn’t have that limitation. As one of the largest banks in the country, JP Morgan was able to analyze the real, anonymized banking and credit card transactions of their own customers — more than 5 million households in the original study, later updated with roughly 280,000 additional households that were followed through time, year after year.

In other words, this isn’t what retirees say they spend — it’s what actually left their accounts, month after month, year after year.

Surprise #1: The Spending Curve — Independent Proof Your Plan Is Overestimating

So, with that important distinction in mind, let’s get into the first surprise: the lifetime spending curve. It picks up right where last week’s episode left off — but instead of repeating the conclusion, it puts it to the test with real transactions. And in doing so, it reveals something the survey data couldn’t show us: the exact age when lifetime spending peaks — and what that means for the number your entire retirement plan is built on.

When JP Morgan mapped household spending across every age group, a clear lifetime arc emerged. Spending rises through our 20s and 30s as careers ramp up, homes get purchased, and kids arrive. It then reaches its lifetime peak between our late 40s and early 50s — when the mortgage, the teenagers, the college tuition, and peak career earnings all pile up at once. And from that peak, it begins a long, gradual decline — through our 50s and 60s, into retirement, and all the way through it.

This, by the way, is where a popular rule of thumb comes from — the one that says you’ll spend somewhere between 70% and 80% of the income you earned during your final working years when you enter retirement.

And directionally, it’s right: retirees really do tend to spend less. But I’ve found that rule of thumb can be confusing because it seems to contradict how people picture their own retirement. If you’re planning to travel more, eat out more, and finally have time for hobbies, how could your spending possibly fall?

The answer is that two things happen at once. Yes, you may spend more on travel and dining early in retirement. But at the same time, several of the largest line items in your budget are disappearing. You stop saving for retirement. The mortgage is often paid off, or close to it, and the kids are launched and off the payroll. So even with more trips on the calendar, total outflows for the typical household still end up lower — and they keep drifting lower from there.

In fact, JP Morgan’s data shows that as retirees age, spending declines in nearly every major category — less on travel, less on dining out, less on transportation, and less on clothes. Only two categories grow as households age. The first, as I’m sure you can guess, is healthcare, and we’ll come back to that toward the end of the episode. The second category was charitable giving, which was refreshing to see.

Now, if you listened to last week’s episode, you might be wondering how this squares with the smile-versus-smirk debate — whether spending curls back up at the very end of life or stays mostly flat.

Well, JP Morgan’s data lands in the same place Blanchett’s did, finding that for most households, the late-life uptick never arrives because declines in other areas more than offset rising healthcare costs.

However, among wealthier households, defined as those with $1 million or more in assets, the researchers did observe spending ticking back up toward the end of life, most likely related to paying for optional, more expensive, private care.

So let’s put a number on the gap. As we covered last week, nearly every plan and piece of planning software assumes your spending rises with inflation — call it 3% per year — for life. But when JP Morgan measured what retirees actually do, the number was far lower. For retired households with between $250,000 and $750,000 in investable assets, total spending grew by just 1.8% per year — roughly a full percentage point below the overall inflation rate. And that’s after accounting for the fact that healthcare gets used more, and inflates faster, as we age.

Now, a one-percentage-point gap might sound small. But compounded over a full retirement, JP Morgan estimates that assuming static spending with full inflation may overstate your actual spending needs by as much as 26% by age 95. In other words, if your planning software says you’ll need $130,000 a year at 95, the real figure — based on how retirees actually spend — may be closer to $100,000.

And this tracks closely to the conclusion from David Blanchett’s paper: that planning around real retiree spending behavior could support roughly 20% more starting retirement income.

So we now have two research teams, working with completely different data, concluding that the standard spending assumption overstates the cost of retirement. By correcting it and adapting the assumptions to your situation and preferences, it changes what you can safely spend, how long you need to work, and even how much stock-market risk your portfolio actually needs to take.

Surprise #2: The Spending Surge on the Way Into Retirement

Ok, so the long arc of retirement spending points down – I think we’ve made that clear. But the second surprise is about a specific stretch where it temporarily surges.

As I’ve discussed in past episodes, people don’t retire the way retirement calculators assume, and JP Morgan came to a similar conclusion. According to their research, 53% of American households don’t retire all at once — one spouse retires before the other, or someone leaves a career but keeps consulting or working part-time while collecting Social Security or a pension. The researchers call these households “partially retired.”

And what they found in the data was that partially retired households experience a genuine surge in spending right around the retirement transition. More specifically, median spending starts climbing about two years before retirement income begins, peaks right around the transition itself, and takes roughly three years to settle back down. And the increase is concentrated in a few specific categories — mostly healthcare, apparel, and food, which probably isn’t surprising.

But the finding that really caught my attention in this research wasn’t the spending, it was the debt. These partially retired households didn’t just spend more, they also carried more credit card debt and held less in cash savings, compared to their fully retired peers with the same income.

In other words, even households thoughtfully easing into retirement and continuing to earn income were more likely to take on credit card debt during what can be one of the most financially complex stages of life.

Now, the study stops short of explaining why this happens, but the surrounding data does offer some strong clues.

As an example, these partially retired households tended to retire later, and the researchers suggest that many may still be working part time precisely because they need to. For households that have felt financially stretched, a Social Security or pension check arriving alongside a paycheck can make previously deferred purchases now feel affordable. There may also be a psychological shift at work. After decades of restraint, that first retirement check can signal that it is finally safe to loosen the purse strings, and once spending rises, it may be difficult to pull it back.

Lastly, it’s important for me to note that the spending surge was strongest in households earning less than $50,000 before retirement and remained visible up to about $150,000. Above that level, it seemed to disappear from the data.

But that doesn’t mean higher-income retirees avoid front-loaded spending altogether. In my experience, retirees at nearly every income and asset level tend to spend more in the early years on travel, family, and home improvements. The difference is that, for more affluent households, this spending is less likely to create financial strain or lead to rising credit card debt.

What matters most is whether the surge is anticipated and built into the retirement plan or arrives as an expensive surprise.

And that’s an important point I want to emphasize. The transition spending surge isn’t a mistake to avoid — it’s simply a phase to plan for. If there are big-ticket items on the horizon — the new roof, the kitchen, or the dream trip — one consideration is to knock some of them out in your final working years, while a paycheck is still covering them. On the other hand, if they’re going to land in early retirement instead, put them in the plan explicitly, with real dollar amounts, so they can be planned around.

To show why this matters so much, J.P. Morgan ran a sobering illustration. Imagine someone retiring in 1966 with a $1 million portfolio invested 60% in stocks and 40% in bonds. They withdraw 5% in the first year, then increase that amount with inflation each year after. And over the next 30 years, the portfolio earns a healthy average annual return of 9.7%. However, the timing of those returns is far from ideal – the market falls 10% in the retiree’s very first year, and the rest of the decade remains difficult. But even with that rocky start, a retiree who sticks to the original spending plan reaches age 95 with roughly $200,000 remaining.

However, add a 30% surge in spending during only the first three years of retirement, and the portfolio runs out of money by age 91. The long-term investment returns are exactly the same – the only thing that changes is when the money is spent in the first few years of retirement. Researchers coined this risk: retirement “dollar-cost ravaging.”

Surprise #3: Retirement Spending Is Bumpy — Especially Early On

And that brings us to the third surprise, because the spending surge is only one kind of retirement bump. In fact, when JP Morgan zoomed in on individual households, they found that hardly anyone’s spending behaves the way a financial plan assumes.

According to the research, six in ten households saw their annual spending temporarily jump or drop by 20% or more in the first few years of retirement, compared to the year just before they retired.

And when the researchers sorted all that movement, every household landed in one of six groups — and the nicknames tell the story. On one end were the “steady eddies,” whose spending stayed within 20% of their pre-retirement level, year after year. On the other end were the “rollercoasters,” whose spending swung both up and down by more than 20% within just a few years. And the four groups in between were all shifters — households whose spending settled meaningfully higher or lower, some permanently, and some for just a year or two.

But out of all six groups, only the steady eddies spend the way most financial plans assume — and they were a clear minority. In fact, at the start of retirement, fewer than two in ten households spent that way.

Now, you may recall from last week’s episode, when I explained that year-to-year spending swings tend to widen at older ages, driven largely by those unpredictable healthcare costs. Well, JPMorgan’s data adds another important piece to that picture by showing that spending volatility is also especially common at the beginning of retirement. The share of households experiencing large swings is highest around retirement and declines somewhat as retirees settle in, but it never disappears.

Even among retirees between ages 75 and 80, roughly half were still classified as volatile spenders, and that is before most long-term care expenses typically begin after age 80.

And honestly, this shouldn’t be surprising. Life doesn’t send you expenses in neat, equal, annual installments — the roof needs replacing, a daughter gets married, you finally take that once-in-a-lifetime trip. And then next year, none of those things happen, so spending drops right back down.

We see this with clients all the time, and it’s important to emphasize that lumpy spending is not a problem by itself. If the averages work out over time, a well-built plan can absorb it. But here’s the thing: when your spending is hard to predict, so are your withdrawals. Imagine you enter retirement planning to spend $100,000 a year, and in year two — between the kitchen remodel and the dream African safari trip— you actually spend $200,000. That extra $100,000 has to come from somewhere, and if it comes from selling investments in the middle of a market downturn because you didn’t build the potential costs into your plan, you’ve just converted a temporary decline into a permanent loss.

That’s the same sequence-of-returns risk we just saw in the 1966 example —and volatile spending makes it worse, because it can force larger withdrawals at exactly the wrong time.

The good news is that the response to this finding is straightforward: liquidity. If you’ve been a long-time listener of this show, you already know where I’m going — this is exactly why we help clients maintain a war chest of cash and safe, shorter-term bonds covering the next two to three years of expenses. When spending surprises show up — and the data says they will, for most of us — the money comes from the war chest, not from selling investments at potentially an inopportune time.

Before We Wrap Up: The One Category That Breaks the Rule

Now, before we wrap up, I want to talk about the exception to today’s episode — not just because it’s a big one, but also because it connects directly to the healthcare “tail” we discussed last week.

Healthcare is the one major spending category that consistently moves in the opposite direction. It takes up a larger share of the budget as we age, and its prices have a long history of rising faster than everything else. For planning purposes, JP Morgan recommends assuming healthcare inflation of roughly 6% per year — about triple what most plans assume for everything else.

More specifically, the study estimates that the most comprehensive Medicare coverage available today costs about $6,860 per person per year, or $13,720 for a couple. If that cost rises by 6% annually, the same couple would be paying roughly $24,500 per year one decade from now. In other words, they would be paying nearly twice as much for the exact same coverage, just ten years later.

And that’s before we get to long-term care, which is not included in any of those spending curves. According to the data JP Morgan cites, roughly three-quarters of women and two-thirds of men age 65 and older are projected to need some form of long-term care during their lifetimes. Now, not everyone pays for that care out of pocket — many rely on family and friends, and some qualify for Medicaid. But among households that do pay, the range is enormous: According to JP Morgan, more than a quarter spend less than $30,000 in total… while nearly one in three spend more than $300,000.

That wide range is the same “tail” we unpacked last week, just measured in a different dataset. And the lesson is the same: this isn’t a risk you can average away. It deserves its own explicit line in your plan — quantified and funded — so it doesn’t sit in the back of your mind as a vague fear that keeps you from enjoying your go-go years.

Notably, JP Morgan closes with specific guidance for planning professionals: inflate Medicare-related costs at 6% per year, and consider reducing the assumed inflation rate on nearly everything else by a full percentage point.

Now, at our firm, we follow the first half of that guidance to the letter — healthcare and Medicare-related costs get inflated at 6% in every plan we build. However, on the second half, we intentionally part ways with the research and do not dial down the inflation assumption on everything else. Instead, we keep and use the long-term historical average as our baseline.

We knowingly use an assumption the data clearly documents is too high because no one can predict every surprise a 30+ year retirement will serve up, and that extra pressure is one way we account for the unknowns.

Think of it as a built-in stress test. If your plan is strong and healthy while assuming 6% medical inflation, full historical inflation on everything else, and lower-than-expected investment returns, it’s clearing a higher bar than history or any of the research from these past two weeks says it will likely face. And knowing that is where real confidence and peace of mind come from.

The key is that it’s a deliberate choice, not a hidden default we missed in the planning softare. A plan that overstates costs because nobody ever questioned the software’s settings can push you to work longer and spend less without realizing it. A plan that holds itself to a tougher standard on purpose — with the cushion out in the open — gives you a margin of safety you can actually see.

The Bottom Line

Over the last two weeks, we’ve examined retirement spending through two very different lenses: a decades-long academic study of thousands of households and the real bank transactions of millions more. Both point to the same conclusion: That retirement spending does not rise smoothly with inflation. It often surges around retirement, remains uneven in the early years, and then gradually declines across nearly every category except healthcare.

And what this conclusion means is that the standard assumption used in many retirement plans can miss in both directions. It may overstate spending later in life while potentially overlooking the larger expenses and volatility that often occur early on.

The solution is to build a plan around how retirement actually unfolds. Use realistic spending assumptions, prepare for the initial surge, maintain enough liquidity to navigate uneven years, and plan separately for healthcare and long-term care.

Get those pieces right, and the research points to an encouraging conclusion: You may be able to spend more than your plan suggests during the years when you are most likely to enjoy it.

Thank you, as always, for listening. To view JPMorgan’s research and the other resources from today’s episode, just head over to youstaywealthy.com/296.

Disclaimer

This podcast is for informational and entertainment purposes only, and should not be relied upon as a basis for investment decisions. This podcast is not engaged in rendering legal, financial, or other professional services.