66-Year-Old Couple With $2.2M Thought They Were Frugal Until They Audited Their Spending

Even disciplined retirees can lose track of where the money goes. Consider a couple in their mid-60s who pulls 18 months of credit card statements, bank downloads, and a spreadsheet. They believe they live within their means. The audit tells a different story: $112,000 in spending they did not plan for and largely cannot recall. Restaurant tabs crept up 28%. Travel rose 44%. Home improvement, the quiet budget killer, climbed 60%. None of those categories felt extravagant in the moment, which is precisely how lifestyle creep works.

The couple had started retirement at a 4.5% withdrawal rate on a $2.2 million portfolio, pulling $99,000 a year. The audit revealed an additional $75,000 annualized in unplanned spending, pushing their effective draw to roughly 5.0%.

The original 4% rule, first published by William Bengen in the October 1994 issue of the Journal of Financial Planning, was built using rolling historical returns and designed to survive the worst 30-year sequences in U.S. market history. Bengen’s own calculations initially produced a rate of 4.15%, which was later rounded down to 4% and became the shorthand everyone remembers. In his 2025 book “A Richer Retirement,” published by Wiley, he revised that ceiling upward to 4.7%. He calls this his “Universal Safemax”: the historical worst-case safe rate across a broader portfolio of seven asset classes, including small-cap and mid-cap equities, configured as roughly 55% stocks, 40% bonds, and 5% cash. Importantly, that 4.7% figure represents the floor, not the typical outcome. Bengen’s data show the average Safemax across all retirees in his study ran to approximately 7.1%, meaning most historical retirees could have withdrawn far more, provided they started at a favorable point in the market cycle. Moving from 4.5% to 5.0% might sound like a rounding error, but in Bengen’s original framework and subsequent Trinity-style studies, that half-point shift reduces the 30-year portfolio success rate by roughly 10 percentage points. The damage compounds because dollars pulled early in a flat or declining market never get the chance to recover, a problem known as sequence risk.

Before laying blame on the economy, this couple needs to separate genuine price increases from behavioral drift. The May 2026 PCE report showed headline PCE inflation at 4.1% year-over-year, the highest reading since April 2023, driven largely by energy costs tied to the conflict in Iran. Core PCE, which strips out food and energy, ran at 3.5% for May. The June 2026 PCE report offered some relief: headline PCE eased to 3.7% year-over-year as a temporary Iran ceasefire pulled energy prices lower, while core PCE ticked down to 3.3%. July 2026 data from the Bureau of Economic Analysis held at the same pace, with headline PCE again at 3.7% and core PCE at 3.3%, even as the ceasefire broke down and fighting resumed, suggesting the energy reprieve may be short-lived. Headline CPI for May 2026 came in at 4.2% year-over-year, also a three-year high, with energy accounting for more than 60% of the monthly increase according to the Bureau of Labor Statistics.

A 28% jump in restaurant spending against 3% to 4% services inflation reflects more meals out, well beyond what menu price increases alone can explain. A 60% climb in home improvement against a single-digit price backdrop points to scope expansion rather than cost-of-living pressure. Both categories are behavioral choices hiding inside a line item that looks like “inflation.”

The Strategy That Holds: Guardrails Plus Visibility

Two changes, used together, address this scenario directly.

1. Monthly category-level tracking with a discretionary/essential split. Most retirees track a single number: total spending. Splitting the budget into essentials (housing, healthcare, insurance, groceries, utilities) and discretionary (restaurants, travel, hobbies, gifts, home projects) makes drift far easier to catch early. Restaurant, travel, and home improvement are the categories most vulnerable to lifestyle creep at every wealth level.
2. Guyton-Klinger guardrails on the withdrawal rate itself. Jonathan Guyton and William Klinger formalized this framework in March 2006 in the Journal of Financial Planning. The strategy sets upper and lower bands around a starting withdrawal rate. The capital preservation rule is triggered when the current withdrawal rate rises more than 20% above the initial rate, prompting a 10% cut in dollar withdrawals. The prosperity rule works in reverse, allowing a 10% raise when the withdrawal rate drops more than 20% below the initial rate. For this couple, guardrails would have flagged the drift toward 5.0% as a trigger for a spending reset well before any need to sell investments at a bad time. Critics, including researchers writing on Kitces.com, have noted that the rules can produce steep cumulative income cuts in extended bear markets, so guardrails work best as an early-warning system rather than a mechanical formula followed in isolation.

What to Do First

The single most useful next step is the one the couple already started: pull 12 to 18 months of transactions and tag every line as essential or discretionary. Without that split, no withdrawal rule works, because there is no way to tell which spending is structural and which is optional.

A retiree at this asset level can probably absorb one expensive stretch without lasting damage. What erodes a plan is the same drift compounding for a decade without correction. Catching it at 18 months, with $2.2 million still on the balance sheet and both partners at 66, is early enough to course-correct without altering the broader shape of retirement. The real advantage of running the audit now is that the couple still has time and portfolio mass to let adjustments work gradually, rather than being forced into abrupt cuts later.

One final note: revisit the withdrawal rate annually rather than monthly. Short-term market noise is a poor basis for long-term lifestyle decisions, and annual reviews give the numbers time to mean something.

Editor’s note: May 2026 core PCE was corrected to 3.5% (from 3.4%) per the BEA official release. July 2026 PCE data from the Bureau of Economic Analysis was added, showing headline PCE held at 3.7% and core PCE at 3.3% even after the Iran ceasefire broke down and fighting resumed in July. The Bengen section was updated to note that the 4.7% Universal Safemax is a worst-case floor and that the average Safemax across all retirees in his study ran to approximately 7.1%, and to clarify that his seven-asset-class portfolio is configured as roughly 55% stocks, 40% bonds, and 5% cash. The Guyton-Klinger publication date was specified as March 2006.

Contact [email protected] for any questions or corrections.