Bengen on Inflation: The Math Behind Retirement's Greatest Risk

On the May 19, 2026 episode of Morningstar's The Long View, Bill Bengen, the planner who derived the 4% rule in his 1994 Journal of Financial Planning paper, told Christine Benz and Amy Arnott that “inflation, in my opinion, is the greatest enemy of retirees.” That is a strong claim from the researcher whose original finding still anchors retirement planning thirty years later. It is also one most DIY investors only half understand.

The retirees who fail in Bengen's historical dataset are not the ones who watched prices drift up at 2.5% a year. They are the ones whose first decade combined an inflation surprise with falling asset prices, while their withdrawal formula raised spending every year regardless of what the portfolio was doing. The 4% rule is, in its bones, an inflation-adjusted spending rule. When inflation misbehaves at the wrong time, the rule itself becomes the problem.

Two distinctions sharpen the Bengen mechanism. Expected inflation, the kind that markets and contracts already price in, is mostly a planning input. Unexpected inflation, the kind that exceeds what nominal yields and budgets assumed, forces the withdrawal raise higher than the portfolio can support. And the failures concentrate in cohorts where unexpected inflation arrives in the first decade of retirement, before the portfolio has time to recover from any market drawdown.

What Bengen's 1994 Paper Actually Said

The 4% rule is shorthand for a more specific finding. Using Ibbotson Associates' Stocks, Bonds, Bills and Inflation: 1992 Yearbook data going back to 1926, Bengen tested every 30-year retirement window. For a portfolio of 50% S&P 500 stocks and 50% intermediate-term government bonds, he asked: what is the highest first-year withdrawal rate, adjusted upward each subsequent year for CPI, that would have survived every historical 30-year window?

His conclusion, verbatim from the 1994 paper, was that “a first-year withdrawal of 4 percent, followed by inflation-adjusted withdrawals in subsequent years, should be safe. In no past case has it caused a portfolio to be exhausted before 33 years, and in most cases it will lead to portfolio lives of 50 years or longer.” The worst case in that dataset was the retiree who began on January 1, 1966: their portfolio lasted 33 years at 4%, the minimum across all starting years. A 4.25% first-year withdrawal, by contrast, could exhaust a portfolio in as little as 28 years.

The 4.15% “SAFEMAX” figure that is often quoted comes from Bengen's later refinement, not the original 1994 paper. In a 2026 update tied to his book A Richer Retirement, Bengen further widened the portfolio to include small-cap and international equity and moved the headline worst-case starting year from 1966 to October 1968, with a new safe rate of 4.7%. Different starting portfolio, different worst-case year, different headline number. The mechanism, however, did not change.

That mechanism is what Bengen himself flagged in the 1994 paper. Of the three major historical stress events in his dataset (the Depression, the 1973-1974 recession, and the post-2000 bear), he wrote: “The ‘Big Bang’ of the 1973-74 recession was the most devastating because it occurred during a period of high inflation. Not only did investors suffer large paper losses in their portfolios, but the purchasing power of what remained was reduced substantially.” His broader takeaway, verbatim: “it is not a deflationary period like the Depression that is to be truly feared, but rather an inflationary period that wreaks havoc on purchasing power as well as portfolio values.” Thirty-two years later he restated the same idea on Morningstar: inflation is the greatest enemy of retirees.

The mechanic that matters here is the formula itself. The retiree does not withdraw 4% of the portfolio each year. They withdraw 4% in year one, then increase that dollar amount by CPI every subsequent year, regardless of what the portfolio did. Walk through what happens with a $1M starting portfolio and 8% inflation:

If markets cooperated and the portfolio is still worth $1M, year 2 spending is 4.32% of the balance. Survivable. If markets fell to $800,000, that same $43,200 is now 5.4% of the remaining balance. Repeat that ratchet for several years of inflation while the portfolio struggles to recover and the effective withdrawal rate spirals. Bengen put it plainly in 1994: “the deleterious impact of the 1973-1974 period can be seen to reach back to retirement portfolios whose withdrawals begin many years earlier, as much as 20 or more years earlier.” The bear market hurt cohorts that had been retired since the early 1950s, two decades into what looked like a comfortable plan.

The 1994 result also depended on the duration of the bond sleeve. Bengen used intermediate-term Treasuries, which were less savaged by rising rates than longer-duration nominal Treasuries. The interactive calculator below uses 10-year Treasuries (the publicly available Damodaran series), which in the 1966-1981 rising-rate regime delivered materially worse total returns than 5-year intermediates. The duration choice in the bond sleeve is itself part of the inflation story, and is often the most underrated retirement decision.

For the broader framework around withdrawal rates and how modern research moved beyond a single static number, see the Summitward guide on safe withdrawal rates.

Expected vs. Unexpected Inflation

Inflation as a planning variable splits cleanly in two.

Expected inflation is what markets, contracts, retirement plans, and policymakers already build in. It is the difference between nominal Treasury yields and TIPS yields (the breakeven inflation rate). It is the assumption behind Social Security cost-of-living adjustments, IRS tax-bracket indexing, wage-bargaining COLAs, and the inflation row in any retirement projection. The Cleveland Fed publishes a monthly Inflation Expectations model that combines CPI data, Treasury yields, inflation swaps, T-bill yields, and survey forecasts to estimate expected inflation across horizons from one year to thirty.

Unexpected inflation is the surprise: realized inflation higher than what was priced into nominal yields and budgeted into plans. That is the dangerous part for retirees, because fixed nominal income streams (pensions without COLAs, nominal bond coupons, cash balances) lose purchasing power while inflation-adjusted spending needs rise. Pension annuities, long nominal bonds, and CDs are all bets against unexpected inflation. The 1973-1981 surge was an unexpected-inflation event by any measure: 10-year breakeven inflation expectations in 1972 were nowhere near the 9-13% realized rates of the late 1970s.

The market impact also depends on the source of the inflation. A 2025 Federal Reserve Board paper categorizes inflation regimes as “good” (inflation co-moves with growth, often driven by demand) or “bad” (supply shocks, stagflation). In the good regime, equities can absorb higher inflation expectations because corporate earnings grow with prices. In the bad regime (the 1970s, 2022), inflation news suppresses equity multiples, widens credit spreads, and raises the correlation between stocks and long bonds. Retirees feel the bad-inflation regime sharply because both halves of a 60/40 portfolio can fall together.

How Inflation Hits Each Asset Class

Different parts of a portfolio respond differently to inflation. Knowing which is which is what lets a retiree assemble a plan that can take an inflation shock without forcing a panic move.

Nominal bonds. Expected inflation is already priced into the yield. Unexpected inflation hurts in two ways: the fixed coupon buys less in real terms, and when rates rise to catch up with inflation, bond prices fall. The pain is concentrated in longer-duration bonds. A 30-year Treasury at 4% loses roughly 17% of its market value if yields rise by 100 basis points, while a 2-year Treasury loses about 2%. Long nominal bonds are not an inflation hedge.

TIPS. Treasury Inflation-Protected Securities are the cleanest public-market instrument for matching inflation-linked spending. Per TreasuryDirect, principal adjusts with CPI, semi-annual interest is paid on the adjusted principal, and at maturity the holder receives the greater of the inflation-adjusted principal or the original principal. TIPS are not risk-free if sold before maturity: their market prices fall when real yields rise. They also create annual taxable phantom income from inflation adjustments in taxable accounts, which is why most planners hold them in IRAs. For a practical implementation of TIPS as inflation-protected retirement income, see the Summitward guide on building a TIPS ladder.

Stocks. Equities are long-run claims on corporate earnings, which tend to grow with inflation over decades. They are not, however, a reliable short-run inflation hedge. High and unexpected inflation compresses valuation multiples, raises discount rates, and pressures profit margins. The 1968 cohort learned this the hard way: real S&P 500 returns from 1968 through 1982 were negative for the full fourteen years. Over 20-30 year horizons, stocks have outpaced inflation. Over the next five, that is not promised.

Commodities. Commodity prices can respond strongly to inflation surprises, particularly supply shocks (oil 1973, oil 2022, food spikes). Vanguard's research has described commodities as having a high “inflation beta” that hedges beyond the dollar invested, but at the cost of high standalone volatility and no real claim on long-run economic growth. Commodities are a tactical diversifier and a short-run hedge, not a retirement income plan.

Real estate and fixed-rate mortgages. Real estate is a partial hedge because rents and replacement costs adjust over time, but residential property is rate-sensitive, illiquid, and idiosyncratic. The more underrated household hedge is a long-term fixed-rate mortgage: the debt is nominal, so inflation erodes the real cost of the principal balance over decades. The trap is overconcentration in one home in one market. The Summitward guide on treating a 30-year mortgage as an inflation hedge walks through the conditions where this works.

For a wider menu of explicit inflation hedges (I Bonds, TIPS, commodities, gold, fixed-rate debt) and an after-tax comparison across them, see the guide on the best inflation hedges.

How Inflation Compounds Sequence-of-Returns Risk

Inflation late in retirement is manageable. Inflation early in retirement, when the portfolio is fully exposed and the withdrawal formula is rigid, can permanently impair the plan. This is the inflation-flavored version of sequence-of-returns risk. Bengen flagged it directly in 1994: the destructive ingredient in the worst-case 1966 cohort was not just the bear market of 1973-1974, but that the bear arrived alongside double-digit CPI prints that ratcheted the withdrawal dollar amount higher year after year. He concluded that “an inflationary period that wreaks havoc on purchasing power as well as portfolio values” was the case to be feared, not a deflationary one.

The mechanic is straightforward. In any given year the rule increases the withdrawal by CPI. If portfolio returns happen to be positive, the higher withdrawal is absorbed. If portfolio returns are negative, the higher withdrawal compounds the damage. After three or four years of an inflation shock during a bear market, the effective withdrawal rate (nominal dollars divided by current portfolio) can climb from 4% to 6% or 8%, at which point recovery becomes mathematically very hard.

The calculator below makes this visible. It runs every historical 30-year retirement window from 1928 onward using actual S&P 500 total returns, 10-year Treasury bond total returns, and CPI inflation. It compares the Bengen-style rigid rule against a simple flexibility rule: skip the inflation raise after any year the portfolio fell. Try the 1966 preset (Bengen's original worst case), then 1969 (stagflation peak), then 1982 (the easy case). Toggle the flexibility off and on to see how much it changes.

Three patterns repeat across the historical record. First, starting years with high inflation in the first decade are dramatically worse than starting years with high inflation in the second or third decade, holding average returns roughly constant. Second, the simple skip-the-raise rule converts most of the failed starts (1966, 1968, 1969) into survivals. Third, 10-year Treasury bonds (which this calculator uses, following the Damodaran dataset) produce harsher outcomes than the 5-year intermediate bonds Bengen originally used. That last point is itself part of the inflation story: longer-duration nominal bonds amplify inflation damage. For the deeper treatment of how return-order destroys portfolios independent of inflation, see the Summitward guide on sequence-of-returns risk.

Your Personal Inflation Is Not CPI

The CPI-U headline that everyone quotes is the average price change for an urban consumer's basket. Retirees do not buy the average basket. They spend more on healthcare and housing than the working-age population, and less on transportation, education, and apparel. Healthcare in particular has compounded well above headline CPI for decades.

The Bureau of Labor Statistics publishes an experimental Consumer Price Index for the Elderly, R-CPI-E, covering households where the reference person is age 62 or older. Per the BLS R-CPI-E page, the series is explicitly experimental, has acknowledged methodological limitations (the expenditure weights are not as rigorously sampled as CPI-U), and is not an official replacement index. Over long periods R-CPI-E has run roughly 0.2 percentage points per year higher than CPI-U. That sounds small until you compound it across a 25-year retirement, where a sustained gap of 1% (a plausible upper bound for retirees with above-average healthcare spending) compounds to roughly 28% more cumulative price increase than the headline number suggests.

Social Security COLAs use CPI-W (urban wage earners), not CPI-U, and not R-CPI-E. The mismatch is part of why Social Security replacement rates erode in real terms for retirees with healthcare-heavy spending baskets, even though the program is nominally inflation-indexed.

The Principal Inflation Risks to Plan For

• Purchasing-power risk. Cash, fixed pensions, nominal bonds, and any fixed-dollar income stream lose real value. The longer the horizon, the larger the cumulative loss.
• Inflation sequence risk. Inflation shocks early in retirement, combined with falling asset prices, can permanently impair the portfolio in a way late-retirement inflation cannot.
• Duration risk. Long nominal bonds amplify inflation damage. A 30-year Treasury can drop 25% in market value when long-term inflation expectations reset 200 basis points higher.
• Stock-bond correlation risk. In bad-inflation regimes, stocks and nominal bonds can fall together. The 2022 calendar year saw the worst combined return for U.S. 60/40 portfolios in decades, with both halves down double digits. The diversification benefit thins out exactly when retirees need it most.
• Personal inflation risk. Headline CPI may understate the spending growth a specific retiree faces, particularly on healthcare and long-term care.
• Tax drag. Most federal income tax brackets and the standard deduction are inflation-indexed (the IRS publishes adjustments each fall, including its 2026 inflation adjustments), but not every feature is. Long-term capital gains brackets are indexed, the Social Security taxation thresholds are not, and IRMAA Medicare premium thresholds are indexed but adjust on a lag. TIPS in taxable accounts also generate phantom income from inflation accruals.
• Longevity risk amplifies all of the above. A 70-year-old may face 20 to 30 years of cumulative price increases. A 55-year-old early retiree may face 40 or more. At 3% inflation, prices double in 24 years; at 5%, they double in 14.

What Actually Helps

Bengen's answer in the Morningstar interview was not to sell stocks and hide in cash. His updated research (the 2025 book A Richer Retirement) raised the starting safe rate to 4.7% with a more diversified portfolio that included small-cap and international equity exposure. Morningstar's independent 2026 State of Retirement Income research, by Amy Arnott, Christine Benz, and Jason Kephart, arrived at a 3.9% base-case starting rate for retirees seeking constant inflation-adjusted spending with a 90% success probability over 30 years. The same research found that flexible withdrawal strategies could push the sustainable starting rate as high as 5.7%.

The shared signal across the literature: rigidity is the problem, not the withdrawal rate. The following framework holds up regardless of which starting number you pick.

1. Plan in real dollars. Express the target spending budget in today's purchasing power. Model expected inflation in the projection; stress-test against unexpected-inflation paths separately. A nominal-dollar plan feels precise and is quietly wrong.

2. Match essential spending with inflation-aware income. The portion of retirement spending that is non-negotiable (housing, food, healthcare, baseline utilities) should be funded from sources that adjust with inflation: Social Security (CPI-W indexed), a TIPS ladder sized to the liability, I Bonds where the per-person purchase limits allow, and any true inflation-linked pension. Discretionary spending (travel, hobbies, gifts) can absorb more market volatility.

3. Use TIPS to match liabilities, not as a return engine. A TIPS ladder is for known real spending needs. It is not a growth allocation. Expecting a TIPS ladder to also outperform equities defeats its purpose.

4. Keep equities for long-term real growth. The historical evidence supports equity allocation for multi-decade retirements precisely because long-term real returns are the only realistic source of inflation-beating growth. Selling stocks during an inflation panic locks in damage.

5. Avoid heavy long-duration nominal bond exposure for near-term spending. If a retiree needs the bond sleeve to cover spending in years one through ten, longer durations are the wrong tool. Intermediate-term Treasuries or short-duration TIPS are more defensible.

6. Use a flexibility rule. The simplest version, skipping the inflation raise after any year the portfolio fell, converts most of Bengen's failed starts into successes. The Morningstar research evaluated eight flexible methods (guardrails, percent-of-portfolio, RMD-style, Guyton-Klinger) and most produced higher lifetime spending than a rigid rule even at lower starting rates. Pick one method, write it down before retirement, and follow it.

7. Delay Social Security if it fits the broader plan. Each year of delay from 62 to 70 increases the monthly benefit by 6-8% in real terms, and the benefit itself is CPI-W indexed. The Summitward guide on claiming Social Security walks through the math on when delaying pays off.

Who Should Care Most

Inflation-sequence risk concentrates on a specific population.

Most relevant for: investors within ten years of retirement; early retirees and FIRE practitioners with 40-plus year horizons; retirees funding essentials primarily from portfolio withdrawals rather than from Social Security or a true inflation-linked pension; households with heavy nominal long-bond allocations; high earners deciding how to split between TIPS, I Bonds, nominal Treasuries, equities, and cash.

Less urgent for: younger accumulators whose primary asset is future wage income, which tends to adjust with inflation on a multi-year lag; retirees whose essential spending is fully covered by Social Security and a true inflation-linked pension; investors who have not yet built any spending plan at all (those need a plan first, then the inflation overlay).

Key Takeaways

• Bengen's 1994 worst-case retiree started in 1966. At a 4% inflation-adjusted withdrawal rate, that cohort's portfolio lasted 33 years; 4.25% could fail in 28 years. The destructive ingredient was bad equity returns combined with the 1973-1981 inflation surge, not either factor alone. (Bengen's 2026 update widens the portfolio and shifts the worst case to October 1968 with a safe rate of 4.7%.)
• Expected inflation lives in market prices; unexpected inflation breaks plans. Treasuries and TIPS embed the expectation. Surprise inflation is the part retirees need to hedge or absorb.
• Long-duration nominal bonds amplify inflation damage. The duration choice in the bond sleeve can matter more for retirees than the equity-vs-bond split.
• A flexibility rule is the highest-leverage change. Skipping the inflation raise after a down portfolio year saves most failed historical starting windows.
• TIPS, I Bonds, and Social Security are liability-matching tools, not return engines. Use them to floor essential spending; keep equity for long-term real growth.

Frequently Asked Questions

Is the 4% rule still safe in 2026 given today's valuations and bond yields?

Morningstar's 2026 research arrived at 3.9% as a base-case starting rate for fixed real spending with 90% success probability over 30 years, using forward-looking return and inflation assumptions and portfolios with 30-50% equities. That is slightly more conservative than 4%, and meaningfully more conservative than Bengen's 2025 updated 4.7% figure (which uses a more diversified portfolio). The honest answer is that the starting number depends on the equity allocation, the bond duration, the flexibility rule, and the time horizon. Spend less time agonizing over 3.9 vs. 4.0 vs. 4.7 and more time institutionalizing a flexibility rule.

Should I just put everything in TIPS and live off the real coupon?

For a finite real liability over 30 years, a TIPS ladder is excellent. For an indefinite-horizon plan with bequest goals or long-tail longevity risk, an all-TIPS portfolio gives up the real growth that long-horizon equity exposure provides. The Summitward guide on living off Treasury interest works through how much capital a coupon-only plan actually requires and where it breaks.

How much should I worry about a 1970s-style scenario today?

The 1973-1981 inflation surge had specific causes: oil shocks, the end of the Bretton Woods gold-dollar peg, accommodative monetary policy, and wage-price spirals enforced by stronger union bargaining power. The institutional setup of 2026 is different on each of those dimensions, and the Federal Reserve now has both an explicit 2% inflation target and a credibility track record. None of that means inflation cannot return. It does mean that betting the plan on a literal 1970s rerun is less calibrated than building a plan that survives a generic bad inflation regime: skip-raise flexibility, TIPS for essential spending, Social Security delay, and moderate bond duration.

Why does this calculator show worse outcomes for 1966 than Bengen's 1994 paper reported?

Bengen used Ibbotson SBBI intermediate-term U.S. government bonds (roughly five-year duration) in his bond sleeve. His 1994 paper found that a 1966 retiree at 4% inflation-adjusted withdrawals had a portfolio that lasted 33 years. This calculator uses 10-year Treasury total returns (the Damodaran historical series), which carry more duration risk. In the 1966-1981 rising-rate regime, 10-year Treasuries delivered materially worse total returns than 5-year intermediates, so the same retiree at the same withdrawal rate deplete earlier with a 10Y sleeve. That gap is precisely the inflation-duration story this guide is about. If your real-world plan holds long-duration nominal bonds, your calculator outcome is the more realistic one. If you hold intermediate-term bonds, your true outcome falls between the two.

Related Guides

• Safe Withdrawal Rate: Why the 4% Rule Isn't Enough covers the broader withdrawal-strategy landscape and Bengen's place in it.
• Sequence of Returns Risk explains the mechanic that inflation amplifies.
• The Four Deep Risks of Investing positions inflation alongside deflation, confiscation, and devastation in Bernstein's framework.
• The Best Inflation Hedges Are Boring compares TIPS, I Bonds, fixed-rate debt, and gold on after-tax real yield.
• How to Build a TIPS Ladder turns the liability-matching idea into a concrete buying plan.
• Stocks Usually Win. “Usually” Is Not a Financial Plan. covers the 1968-2009 historical backdrop where equity premiums collapsed.
• When to Claim Social Security walks through the case for delaying inflation-indexed income.
• The Treasury Bond Fantasy stress-tests the all-Treasury retirement plan.

Sources

1. Bengen, W. P. (1994). “Determining Withdrawal Rates Using Historical Data.” Journal of Financial Planning, October 1994.
2. Morningstar The Long View, May 19, 2026: “Bill Bengen: ‘Inflation Is the Greatest Enemy of Retirees.’” morningstar.com/retirement/bill-bengen-inflation-is-greatest-enemy-retirees.
3. Arnott, A., Benz, C., and Kephart, J. (2025). The State of Retirement Income: 2026. Morningstar. morningstar.com/retirement/whats-safe-retirement-withdrawal-rate-2026.
4. Pfau, W. “The 4% Rule and the Search for a Safe Withdrawal Rate.” Retirement Researcher. retirementresearcher.com.
5. Federal Reserve Bank of Cleveland. “Inflation Expectations.” clevelandfed.org/indicators-and-data/inflation-expectations.
6. TreasuryDirect. “Treasury Inflation Protected Securities (TIPS).” treasurydirect.gov/marketable-securities/tips.
7. U.S. Bureau of Labor Statistics. “R-CPI-E: Experimental Consumer Price Index for Americans 62 Years of Age and Older.” bls.gov/cpi/research-series/r-cpi-e-home.htm.
8. Internal Revenue Service. “IRS releases tax inflation adjustments for tax year 2026.” irs.gov.
9. Damodaran, A. “Historical Returns on Stocks, Bonds and Bills: 1928 to current.” NYU Stern. pages.stern.nyu.edu/~adamodar (S&P 500 and 10-year Treasury total return series used in the calculator).
10. Federal Reserve Economic Data. “Consumer Price Index for All Urban Consumers, U.S. City Average (CPIAUCNS).” fred.stlouisfed.org/series/CPIAUCNS (CPI series used in the calculator).