Nominal returns systematically overstate what investors actually earn, and the corrective math is unambiguous: since 1928 the S&P 500 has compounded at roughly 10% per year in nominal terms but only about 7% in real terms after ~3% average inflation — a gap that must be computed with the exact Fisher formula, (1+nominal)/(1+inflation)−1, not the common subtraction shortcut. Everything below traces to a primary or authoritative source current to mid-2026 and is presented so it can be cited directly in a long-form educational article.

TL;DR

Nominal misleads; use real. The S&P 500's ~10.02% long-run nominal compound return (Damodaran/NYU Stern, 1928–2025) becomes roughly 6.7–7.0% real after ~3% average inflation. The correct conversion is r = (1+nominal)/(1+inflation)−1; the popular "nominal minus inflation" shortcut introduces a small but compounding error.
Every headline figure is sourced. CPI from BLS/FRED, real yields from FRED/U.S. Treasury, withdrawal research from Bengen (1994) and Morningstar (2025 edition), the Fisher equation from Fisher's 1930 Theory of Interest, and real-vs-nominal institutional reporting from Yale's published spending policy.
The four documented planning killers are inflation, fees, taxes, and sequence-of-returns risk — each addressed by regulators (SEC Investor.gov, CFPB) and asset managers (Vanguard, Fidelity, Morningstar).

Key Findings

1. Compound interest fundamentals

Standard formula: A = P(1 + r/n)^(nt), where A = final amount, P = principal, r = annual nominal rate (as a decimal), n = compounding periods per year, and t = number of years.
Continuous compounding: A = Pe^(rt), where e ≈ 2.71828.
Rule of 72 — origin: The number 72 was first referenced by Italian mathematician and Franciscan friar Luca Pacioli in his 1494 work Summa de arithmetica, geometria, proportioni et proportionalità. Pacioli noted the usefulness of 72 for estimating doubling time without deriving it; the modern rule-of-thumb form was popularized later.
Accuracy and the Rule of 69.3 / Rule of 70: The mathematically exact doubling constant is the natural log of 2, ln(2) ≈ 0.693, so the "Rule of 69.3" is the most accurate for continuous compounding. Per Wikipedia's derivation, the Rule of 72 is most accurate for periodic compounding around 8%, and the Rule of 70 is most accurate around 2%. The number 72 is preferred for mental math because it is evenly divisible by 2, 3, 4, 6, 8, 9, and 12. Example: at a 12% continuously-compounded return, 69.3 ÷ 12 = 5.78 years, which closely matches the exact calculation.
Compounding-frequency effect (worked example): For a 5% nominal rate on $10,000 over one year — annual → $10,500.00; monthly → ≈$10,511.62; daily → ≈$10,512.67; continuous → $10,000 × e^0.05 = $10,512.71. The entire gap between annual and continuous compounding at 5% is only about $12.71 per $10,000 in one year. The effect grows with higher rates and longer horizons: over 30 years at 8%, $10,000 grows to about $100,626 with annual compounding versus about $110,232 continuously — a roughly $9,600 difference.

2. Nominal vs. real returns / the Fisher equation

Attribution: Named after American economist Irving Fisher, who formalized the relationship in his 1930 book The Theory of Interest.
Exact formula: (1 + i) = (1 + r)(1 + π), where i = nominal rate, r = real rate, and π = inflation. Solving for the real rate: r = (1 + i)/(1 + π) − 1.
Common approximation: r ≈ i − π, which is accurate only when rates and inflation are low.
Worked example: With a 7% nominal return and 4% inflation, the approximation gives 3.00%, while the exact formula gives (1.07/1.04) − 1 = 2.88% — a 0.12-point overstatement per year that compounds meaningfully over decades. At higher inflation the error widens: nominal 10%, inflation 8% → approximation 2.0% vs. exact (1.10/1.08 − 1) = 1.85%.

3. Historical U.S. CPI-U inflation (BLS, annual average)

Annual average CPI-U inflation: 2006: 3.2%; 2007: 2.8%; 2008: 3.8%; 2009: −0.4% (deflation); 2010: 1.6%; 2011: 3.2%; 2012: 2.1%; 2013: 1.5%; 2014: 1.6%; 2015: 0.1%; 2016: 1.3%; 2017: 2.1%; 2018: 2.4%; 2019: 1.8%; 2020: 1.2%; 2021: 4.7%; 2022: 8.0%; 2023: 4.1%; 2024: 2.9%; 2025: 2.6% (average).

The 2021–2023 spike: The 12-month CPI-U rate peaked at 9.1% in June 2022, the highest reading since 1981. Year-end (December) figures were: Dec 2021 = 7.0%, Dec 2022 = 6.5%, Dec 2023 = 3.4%.
Most recent (mid-2026): CPI-U rose 3.5% for the 12 months ending June 2026 (BLS release July 14, 2026). Monthly 2026 year-over-year prints: Jan 2.4%, Feb 2.4%, Mar 3.3%, Apr 3.8%, May 4.2%, Jun 3.5%. The 2026 pickup was driven substantially by an energy-price spike tied to Middle East conflict. October 2025 CPI data is unavailable because of a 2025 lapse in appropriations (government shutdown) — a genuine data gap in the official series.

4. Historical stock market returns — nominal vs. real (Damodaran/NYU Stern)

The canonical free long-run dataset is Aswath Damodaran's "Historical Returns on Stocks, Bonds and Bills: United States" (NYU Stern), updated January 5, 2026. Note a source-labeling quirk: the page title still reads "1928-2024," but the underlying data now runs through 2025.

S&P 500 geometric (compound annual) nominal return, dividends reinvested: 9.94% for 1928–2024; 10.02% for 1928–2025.
S&P 500 arithmetic average nominal return: 11.79% for 1928–2024 (matching Damodaran's own stated figure); 11.86% for 1928–2025.
10-year Treasury bond geometric nominal return: ≈4.50% (1928–2024), ≈4.54% (1928–2025). 3-month T-bill geometric nominal: ≈3.36%–3.37%.
Value of $100 invested at the start of 1928 (end of 2025): stocks $1,157,598.95; 10-year Treasury bonds $7,752.88; T-bills $2,578.30. (End of 2024: stocks $982,817.82.)
Real returns: The Damodaran nominal page carries no inflation column; inflation averaged approximately 3% per year over the period, implying real geometric returns of roughly stocks ≈ 6.7–7.0%, bonds ≈ 1.5%, bills ≈ 0.35%. Shiller-CPI-based calculators (officialdata.org) put the 1926–2026 S&P 500 real return at about 7.16%–7.30% per year.
Marketing vs. reality: The widely cited "10%" is a nominal, dividend-reinvested figure. Forward-looking forecasts are far lower because of elevated valuations. Vanguard's 2026 Economic and Market Outlook (December 10, 2025) states verbatim: "Our muted U.S. stock return forecast of 4%–5% average returns over the next 5-to-10 years is nearly single-handedly driven by our risk-return assessment of large-cap technology companies." Its published VCMM 10-year range (as of March 31, 2026) is 3.5%–5.5% nominal for US equities — below the range for US bonds. Per Morningstar's "Experts Forecast Stock and Bond Returns: 2026 Edition," Vanguard is most pessimistic on US growth equities, expecting 10-year returns of 2.3%–4.3% for US growth versus 5.8%–7.8% for value. Fidelity's Planning & Guidance Center assumes 2.5% inflation, and (per the same Morningstar survey) forecasts 5.8% nominal / 3.2% real for US equities over 20 years.

5. Fixed income and cash in real terms (mid-2026)

10-year TIPS real yield: 2.35% on July 16, 2026 (FRED series DFII10); ≈2.31% on July 20, 2026 (Trading Economics). A May 2026 10-year TIPS auction (CUSIP 91282CPU9) produced a real yield of 2.169%.
Nominal 10-year Treasury: 4.57% on July 16, 2026 (FRED DGS10), implying a breakeven inflation rate of roughly 2.2%.
High-yield savings APYs (July 2026): top nationally available rates ran roughly 4.10%–4.50% APY (Bankrate best up to 4.15%; NerdWallet's Newtek 4.20%; Fortune tracked up to 4.50%; Yahoo Finance cited 4.10% at Bask Bank and CIT Bank on July 20, 2026). The FDIC national average savings rate was 0.38%. The Fed funds target range was held at 3.50%–3.75% through mid-2026.
In real terms: With June 2026 CPI at 3.5%, a 4.15% HYSA delivers a real return of only about +0.6% (exact: 1.0415/1.035 − 1 ≈ 0.63%), while the 0.38% national average is deeply negative in real terms (≈ −3.0%).
Negative real rate history: 10-year TIPS real yields hit a record low near −1.18% in August 2021; the Fed's tightening (funds rate from near zero in March 2022 to 5.25%–5.50% by July 2023) swung real yields to above +2% by late 2023, a move exceeding 300 basis points. Series I savings bonds hit a record composite rate of 9.62% (annualized) for bonds issued May–October 2022; the November 2025–April 2026 composite rate is 4.03%.

6. Retirement withdrawal math

The Bengen 4% rule: William Bengen, "Determining Withdrawal Rates Using Historical Data," Journal of Financial Planning, October 1994. Assumptions: a 30-year horizon, a tax-advantaged account, a 50% US large-cap stock (S&P 500) / 50% intermediate-term US Treasury portfolio, annual rebalancing, and a first-year withdrawal set as a percentage of the portfolio that is then adjusted for inflation each subsequent year. Bengen's finding was that a ~4.15% initial "SAFEMAX" rate had never failed over any 30-year historical period; the worst-case starting year was 1966, on the eve of stagflation. Bengen has since revised his default upward — in his August 2025 book A Richer Retirement: Supercharging the 4% Rule to Spend More and Enjoy More, and as he told CNBC (September 3, 2025), "the maximum safe withdrawal rate is 4.7%," a figure he calls the "Universal Safemax," reflecting broader diversification.
Trinity Study (1998): Cooley, Hubbard & Walz, "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable," AAII Journal, February 1998, corroborated Bengen using rolling historical periods and "portfolio success rates." A 4% inflation-adjusted withdrawal on a 50/50 portfolio over 30 years showed a ~100% historical success rate, while a 5% rate fell to ~68%.
Morningstar's updated research: "The State of Retirement Income: 2025 Edition" (released December 3, 2025) sets a base-case safe starting withdrawal rate of 3.9% for 2026 — stated verbatim as "the highest starting safe withdrawal rate for retirees seeking a consistent level of inflation-adjusted spending... assuming a 90% probability... at the end of an assumed 30-year retirement period," at a 20–50% equity weighting. The base-case history: 3.3% (2021), 3.8% (2022), 4.0% (2023), 3.7% (2024 and 2025 report), 3.9% (2026). Flexible strategies (guardrails, TIPS ladders, forgoing inflation raises after down years) can lift the starting rate to as high as 5.7%.
Inflation-adjustment mechanics: The withdrawal is set in year one and increased each subsequent year by the actual inflation rate. Morningstar's worked example: a $33,000 first-year withdrawal (on a $1M portfolio) becomes $35,145 after 6.5% annualized 2022 inflation, then $36,340 after 3.4% 2023 inflation.

7. Common calculation mistakes (regulators and researchers)

Ignoring fees: SEC Investor.gov's investor bulletin "How Fees and Expenses Affect Your Investment Portfolio" shows a $100,000 portfolio growing 4% annually over 20 years ending at approximately $208,000 with a 0.25% annual fee, ~$198,000 with 0.50%, and ~$179,000 with 1.00%. The difference between the 0.25% and 1.00% fee is roughly $29,000 — about 14% of the ending balance.
Ignoring inflation: The CFPB and industry sources stress projecting future purchasing power, not just nominal balances, and using real rates of return in projections. A dollar's purchasing power erodes continuously; a fixed nominal income stream steadily buys less.
Ignoring taxes: Fidelity's head of wealth planning, David Peterson, names inflation, taxes, and fees as the "3 main drags on performance."
Sequence-of-returns risk: Poor returns early in retirement, combined with ongoing withdrawals, permanently impair a portfolio because shares sold at depressed prices cannot participate in the recovery (MIT Sloan, Charles Schwab, Fidelity). Fidelity's illustration runs the same 30-year return set in reverse order with $50,000 annual withdrawals (6.8% average return, 13% volatility): early losses can cause irreversible depletion that later-arriving identical losses would not.

8. Real-world "real vs. nominal" reporting (Yale endowment)

Yale University targets spending of 5.25% of endowment value annually, explicitly described as the amount "projected to be sustainable given reasonable long-term growth expectations in the endowment's value, after adjusting for inflation." Its spending rule: 80% of the prior year's spending plus 20% of (5.25% × the prior year's beginning market value), with the sum adjusted for inflation, constrained to be at least 4.0% and no more than 6.5% of the endowment's inflation-adjusted market value at the start of the prior year.
Per Yale's 2025 provost budget FAQ, Yale estimates a long-term nominal return of roughly 8.25%, sets aside ~3% for inflation to preserve real purchasing power, and spends the remaining 5.25% (currently about $2.2 billion per year). Reported returns: FY2024 = 5.7%; FY2025 = 11.1%; the endowment was $44.1 billion as of June 30, 2025, with a 10-year annualized return of about 9.4%. This is a clean, citable example of an institution explicitly separating nominal return into an inflation-preservation component and a real spending component.

9. European reference points (Eurostat / ECB)

Eurozone HICP inflation was 2.8% in June 2026 (Eurostat), down from 3.2% in May 2026, which had been the highest since September 2023; the ECB's target is 2.0%. Germany's June 2026 rate was 2.4%. The May–June 2026 elevation was tied to an energy-price shock from Middle East conflict.
The ECB Survey of Professional Forecasters (Q2 2026) expected headline HICP inflation of 2.7% in 2026, 2.1% in 2027, and 2.0% longer-term. (As of January 1, 2026, the euro area expanded to 21 countries with Bulgaria's entry.)

Details

Compound interest — why it matters

Compound interest is interest earned on both principal and previously accumulated interest (SEC Investor.gov). The SEC operates one of the most-used free compound interest calculators on the web (investor.gov/financial-tools-calculators). The frequency comparison in Key Finding 1 shows diminishing marginal benefit as compounding becomes more frequent — the leap from annual to monthly captures most of the available benefit, and daily-to-continuous is negligible — but the rate and time horizon are what truly drive compounding, which is why the difference between a 7% and 10% assumption over 40 years dwarfs any frequency effect.

The real-return trap

The Fisher relationship is the single most important correction in a personal-finance article. Because most people mentally subtract inflation from nominal returns, they overstate real growth — modestly at low inflation, materially at high inflation. Over a 30-year horizon, a recurring 0.12–0.15 percentage-point annual error compounds into thousands of dollars of misestimated purchasing power. The precise formula, (1+nominal)/(1+inflation) − 1, should always be used for multi-year planning.

The inflation spike in context

The 2021–2023 episode was the sharpest since the early 1980s: annual average inflation rose from 1.2% (2020) to 4.7% (2021) to 8.0% (2022), peaking at 9.1% year-over-year in June 2022, before falling to a 4.1% average in 2023 and 2.9% in 2024. This period is the clearest modern demonstration of why nominal returns mislead: cash and conventional bonds delivered deeply negative real returns even as their nominal values held.

Equities: the nominal-vs-real wedge

The gap between the ~10% nominal and ~7% real S&P 500 return is entirely inflation, and over long horizons the wedge is enormous. $100 growing at 10% nominal for 40 years reaches about $4,526, but at 7% real it reaches only about $1,497 in today's purchasing power. Forward-looking forecasts from Vanguard, Fidelity, and BlackRock are notably lower than the historical average because of high current valuations — an important caveat: planners should not simply extrapolate the 1928–2025 average forward.

Fixed income and cash

Mid-2026 is an unusual environment: real yields on TIPS (~2.2–2.35%) are among the most attractive in over a decade, and top high-yield savings accounts (~4.1–4.5%) roughly keep pace with or slightly beat current inflation. This is a sharp reversal from 2021–2022, when real yields were record-negative and savers lost purchasing power despite positive nominal balances. TIPS and I Bonds are the two instruments that guarantee a positive real return by construction, because their principal or rate adjusts with CPI.

Recommendations

Always convert to real terms. Use r = (1+nominal)/(1+inflation) − 1 for every multi-year projection. Benchmark equity planning at roughly 6–7% real historically (but weight toward the lower forward-looking forecasts given 2026 valuations), bonds at ~1–2% real, and cash near 0% real.
Subtract fees explicitly. Model the expense ratio as a direct reduction to your assumed return. Per SEC figures, a 1.00% annual fee versus 0.25% costs about 14% of the ending balance over 20 years — the difference between an index fund and an expensive active fund can quietly consume a large fraction of lifetime returns.
Use a conservative, inflation-adjusted withdrawal rate. Start near Morningstar's 3.9% (2026 base case) or Bengen's 4%–4.7% range, increase the dollar amount by actual inflation each year, and build flexibility — a cash buffer, guardrails, or a TIPS ladder — to defend against sequence-of-returns risk in the first decade of retirement.
Hold explicit inflation protection. TIPS and Series I Bonds guarantee a real return; at mid-2026, 10-year TIPS real yields near 2.2% are historically attractive relative to the negative real yields of 2021–2022.
Benchmarks that would change this advice: If 10-year TIPS real yields fall back below roughly 0%, shift emphasis away from cash and nominal bonds toward equities and I Bonds. If CPI re-accelerates above roughly 4% on a sustained basis, widen inflation cushions in projections and favor floating-rate or inflation-linked instruments. If Morningstar's or Bengen's published safe rate moves materially, re-anchor withdrawal assumptions accordingly.

Caveats

Forecasts are not facts. Forward-looking return projections from Vanguard, Fidelity, and BlackRock are hypothetical and use conditional language ("expect," "could"); actual outcomes vary widely and are not guarantees. Present them as projections, not established results.
Derived vs. printed figures. The S&P 500 geometric returns (9.94% / 10.02%) were computed from Damodaran's published "$100 invested" end-values; the arithmetic mean (11.79%) matches Damodaran's own stated figure. The ~7% real equity figure is derived using ~3% average CPI and is not printed on Damodaran's nominal page.
A real data gap: October 2025 U.S. CPI is missing due to a government-shutdown appropriations lapse.
Savings rates are volatile. High-yield APYs change daily and vary by institution and balance tier; the mid-July 2026 figures cited here will drift.
Personal inflation differs from CPI. Older households and those with heavy healthcare spending often experience higher inflation than headline CPI; the BLS maintains an experimental CPI-E for those aged 62+.
Source-labeling discrepancy: Damodaran's page title still reads "1928-2024" even though the data runs through 2025 (updated January 5, 2026).

Personal Finance Math: Compound Interest, Real Returns, and Inflation-Adjusted Growth — Research Dossier