A retiree with $750,000 in savings who follows the 4 percent rule will pull out $30,000 in year one, then bump that amount each year to keep pace with inflation. The rule traces back to financial planner William Bengen, who published his research in the Journal of Financial Planning in October 1994. His method was straightforward: withdraw 4 percent of your portfolio in year one, adjust that dollar amount annually for inflation using the Consumer Price Index, and the money should survive three decades of retirement. It became the most widely cited rule of thumb in personal finance.
On paper, the math is supposed to keep the money flowing for 30 years. In practice, a growing number of financial planners say the formula is too rigid for the economy retirees are actually living in, and that following it blindly could mean running short or leaving a fortune unspent. Even Bengen himself has revised his recommendation upward, suggesting in more recent work that 4.5 to 4.7 percent is closer to the historically safe rate. The deeper criticism, though, is not about whether 4 percent is the right number but about whether any fixed starting percentage with an automatic inflation escalator is the right structure.
But as of mid-2026, the criticism is no longer just about whether 4 percent is the right number. Advisors and researchers are challenging the entire structure: a fixed starting percentage with an automatic inflation escalator that ignores how retirees actually spend, how markets actually move, and how inflation actually hits older households.
Why the original formula is losing credibility
The 4 percent rule adjusts withdrawals using the CPI-U, the broad Consumer Price Index for All Urban Consumers published by the U.S. Bureau of Labor Statistics. That index tracks price changes across a wide basket of goods and services, from used cars to college tuition, weighted to reflect spending by the general urban population. It was never designed to mirror the budget of a 72-year-old couple in Tampa spending heavily on Medicare supplements, property insurance, and prescription drugs.
The BLS itself has acknowledged this gap, at least indirectly. Since 1982, the agency has published an experimental measure called the CPI-E (Consumer Price Index for Americans 62 years and older), which reweights the standard basket to reflect older households’ spending patterns. The CPI-E has historically run higher than the CPI-U on average, by roughly 0.2 to 0.4 percentage points per year depending on the period studied, largely because medical care and shelter eat a bigger share of retiree budgets. That gap has narrowed in recent decades, but the underlying point that older households face a different inflation profile than the urban population the CPI-U measures has held up.
Social Security cost-of-living adjustments add another layer of mismatch. Those raises are tied to the CPI-W, which tracks prices for urban wage earners and clerical workers. A retiree relying on both portfolio withdrawals and Social Security checks is exposed to two different inflation measures, and neither one is calibrated to their actual household costs.
Then there is the spending curve itself. Longitudinal data from the RAND Health and Retirement Study, supported by the National Institute on Aging and the Social Security Administration, shows that retiree spending does not climb in a smooth, steady line. Households tend to spend more in their 60s, pull back through their 70s as travel and discretionary activity slow, and then face rising costs again in their 80s when healthcare and long-term care needs escalate. Researchers, including Morningstar’s David Blanchett, have described this pattern as the ‘retirement spending smile.’ A rigid annual CPI bump ignores all of that variation, potentially leaving retirees underspending when they are healthy enough to enjoy it and overspending when their real costs have dropped.
What planners are recommending instead
No single replacement has emerged as the new consensus. But several alternative frameworks have gained traction among credentialed advisors and in peer-reviewed research. Each trades the 4 percent rule’s simplicity for something more responsive.
Guardrails strategies
Originally developed by financial planner Jonathan Guyton and computer scientist William Klinger in research published in the Journal of Financial Planning in 2006, guardrails set upper and lower boundaries around a target withdrawal rate. If strong market returns push the portfolio’s effective withdrawal rate below the lower guardrail, the retiree gets a raise. If a downturn pushes the rate above the upper guardrail, spending gets cut.
The key advantage is adaptability. Retirees are not locked into the same dollar amount regardless of what markets do. Vanguard’s retirement research group has published its own version of a dynamic spending framework based on similar principles, reinforcing the approach’s credibility among institutional researchers.
Dynamic percentage withdrawals
Instead of fixing the dollar amount in year one and adjusting for inflation, some planners recommend recalculating the withdrawal as a fresh percentage of the current portfolio balance each year. A retiree might take 4.5 percent of whatever the portfolio is worth every January.
To put that in dollars: on a $750,000 portfolio, that is $33,750. If the portfolio drops to $650,000 after a bad year, the withdrawal falls to $29,250. If it grows to $850,000, the check rises to $38,250. The approach naturally forces spending down after losses and allows more after gains. The tradeoff is income volatility, and retirees need to be comfortable with that fluctuation or pair it with a stable income floor from Social Security or a pension.
Bucket strategies
This method divides a portfolio into time-based segments. A short-term bucket, covering one to three years of expenses, is held in cash or short-term bonds. A medium-term bucket, covering three to ten years, goes into balanced funds. A long-term bucket stays in equities. Retirees draw from the short-term bucket first, refilling it periodically from the others.
The psychological benefit is real: knowing that near-term expenses are covered in safe assets can prevent panic selling during a downturn. The approach does not eliminate sequence-of-returns risk, but it restructures how retirees experience it, turning an abstract portfolio loss into something that feels more manageable.
Morningstar’s annual withdrawal rate research
Each year, Morningstar’s retirement research team publishes an updated estimate of what it considers a safe starting withdrawal rate, factoring in current bond yields, equity valuations, and inflation expectations. In its most recent annual study, published in December 2025, the team estimated a starting safe rate of 3.9 percent for a balanced portfolio over a 30-year horizon, up slightly from 3.7 percent the year before, but still below the classic 4 percent. The Morningstar researchers have emphasized that flexibility, specifically the willingness to reduce withdrawals modestly in bad years, can raise the starting rate meaningfully.
That gap between 3.9 percent and 4 percent is modest on its own (about $750 a year on a $750,000 portfolio), but the broader point of Morningstar’s work is the year-to-year movement itself. The “safe” rate has bounced between 3.3 percent and 4 percent across five annual studies, which is the actual case for treating any single number as a starting point rather than a rule.
What the federal government does and does not say
No federal agency prescribes a specific withdrawal rate for retirees. The Department of Labor enforces rules under ERISA that govern how employer-sponsored retirement plans operate, but those rules address plan administration, fiduciary duties, and required minimum distributions, not how fast an individual should spend down personal savings.
The IRS sets required minimum distribution schedules for tax-deferred accounts, but RMDs are a tax compliance mechanism, not a spending guide. A retiree’s RMD might force a withdrawal far larger or smaller than what any planning rule would suggest, depending on the account balance and the retiree’s age.
The BLS publishes an online inflation calculator that lets anyone see how purchasing power has shifted over time, a useful tool for retirees trying to understand what their dollars are actually worth. But the agency does not offer retirement planning advice or endorse any withdrawal methodology.
That regulatory gap matters. It means retirees are left to choose among competing planning frameworks, none of which carries an official stamp of approval. It also means the quality of advice varies enormously depending on whether a planner is a fee-only fiduciary, legally required to act in the client’s interest, or a commission-based salesperson with a different set of incentives.
Where the evidence still has gaps
The sharpest unanswered question is how retirees actually responded to the inflation surge of 2021 through 2023, when CPI-U peaked above 9%. The RAND HRS dataset, the most authoritative longitudinal source on retiree financial behavior, releases data with a lag. A complete picture of how households adjusted withdrawals, cut spending, or shifted asset allocations during that period is still emerging. Researchers drawing conclusions about retiree behavior during the inflation spike are often working from cross-sectional surveys or modeling, not from confirmed longitudinal tracking of the same households over time.
The CPI-E has its own limitations. It uses the same price data as the CPI-U and simply reweights the spending categories; it does not collect separate price quotes from pharmacies, medical providers, or retailers that older Americans disproportionately use. Any claim about exactly how much standard inflation measures overstate or understate retiree cost increases should be treated as an informed estimate, not a precise figure.
Practical steps for people planning retirement in 2026
The 4% rule is not worthless. It remains a reasonable starting point for rough planning, and its simplicity is a genuine virtue for people who are never going to build a spreadsheet. But treating it as a permanent, unchangeable spending prescription is where the risk concentrates.
Three steps are worth taking now:
Stress-test for bad timing. Retiring into a bear market in the first two years, when sequence-of-returns risk is highest, can permanently impair a portfolio even if long-run returns are fine. Run scenarios that assume a 20 percent to 30 percent drop in year one and see what happens to the plan at year 20.
Build in flexibility. Whether that means formal guardrails, a dynamic percentage, or simply a personal commitment to trim discretionary spending after a down year, the ability to adjust is the single most powerful tool a retiree has. Delaying Social Security to age 70, when monthly benefits are roughly 77 percent higher than at 62, can also create a larger guaranteed income floor that reduces the pressure on portfolio withdrawals.
Track actual household spending, not just a CPI number. A retiree whose biggest expenses are a paid-off house and Medicare premiums faces a very different inflation profile than one carrying a mortgage and paying for a spouse’s long-term care. Knowing your own numbers makes every withdrawal decision sharper.
The 4 percent rule gave a generation of retirees a simple, memorable answer to a genuinely hard question. The problem is not that the answer was wrong in 1994. It is that retirement planning in 2026 demands something more responsive, more personalized, and more honest about what the data can and cannot tell us.
