Articles · Behavioral finance
Money illusion: why investors confuse nominal and real returns
What money illusion is, the Shafir-Diamond-Tversky experiments, the Modigliani-Cohn idea, and the S&P 500 and Treasury yields in real terms since 1985.
On this page
- What money illusion is
- The experiments
- Money illusion and the stock market: Modigliani and Cohn
- Real data: the SP 500 since 1985
- Real data: the 10-year Treasury yield
- Taxes are levied on nominal gains
- Seeing past the illusion
- How Finabase helps
- Frequently asked questions
- Summary
- Check your understanding
- Exercise
Money illusion is the tendency to think in nominal dollars, the numbers printed on statements and price tags, rather than in real dollars, what that money can buy. A 5% raise feels better than a 2% raise even when prices rose 4% in the first case and not at all in the second. A savings account paying 5% feels generous even when inflation is 6%. The economist Irving Fisher named the idea in 1928; behavioral researchers have since shown how persistent it is.
For investors the illusion matters in three places: judging returns, comparing stocks with bonds, and paying tax. This article explains the experiments that measured it, the hypothesis that it once distorted the whole stock market, and what the S&P 500 and Treasury yields look like once inflation is taken out, using data since 1985.
What money illusion is
Every price is quoted in money, and money itself loses value when prices rise. The real return on an investment is roughly its nominal return minus inflation; the exact relationship is
A 7% nominal return with 3% inflation is a real return of 1.07 ÷ 1.03 − 1 ≈ 3.9%. The mechanics, and long-run real returns by asset class, are in Inflation and real returns: what an investment earns after prices. The illusion is not ignorance of the formula. Most people can explain inflation if asked. The problem is that the nominal number is the one in front of them, and it is the one they feel.
The experiments
Eldar Shafir, Peter Diamond and Amos Tversky published a set of survey experiments in 1997. In the best known, two people, Ann and Barbara, graduate a year apart and take similar jobs at the same starting salary of $30,000. In Ann's first year there is no inflation and she gets a 2% raise. In Barbara's first year inflation is 4% and she gets a 5% raise. In real terms Ann is better off: a 2% gain against roughly 1%.
The answers depended on the question. Asked who was doing better in economic terms, most respondents picked Ann. Asked who was happier, most picked Barbara. People knew the real answer when the question pointed them at it, and felt the nominal one otherwise. Other problems in the study found the same pattern in judgments about house sales and contracts. A 2021 replication in the Journal of Economic Psychology reproduced the main results.
The same habit shows up in markets. Investors anchor on the price they paid in dollars, which is part of why they hesitate to sell at a nominal loss even when inflation has already eroded the holding (see The disposition effect: selling winners and holding losers). Workers resist nominal pay cuts far more than they resist the real cuts that inflation imposes silently.
Money illusion and the stock market: Modigliani and Cohn
In 1979, in the middle of a decade of high inflation and low stock prices, Franco Modigliani and Richard Cohn argued that the whole market was making a money illusion error. Their reasoning, simplified:
- When inflation rises, nominal interest rates rise with it, because lenders demand compensation.
- A company's future profits also rise with inflation in nominal terms, at least for businesses that can raise prices.
- Investors, however, discount profits at the higher nominal interest rate while projecting profits that grow only at the real rate. That double-counts inflation and makes stocks look less valuable than they are.
If they were right, stocks would be undervalued when inflation is high and overvalued when it is low. Price-to-earnings ratios were indeed very low in the late 1970s and early 1980s. Later work found support for the idea: John Campbell and Tuomo Vuolteenaho (2004) and Randolph Cohen, Christopher Polk and Vuolteenaho (2005) showed that the stock market's pricing errors moved with inflation in the direction Modigliani and Cohn predicted.
A common rule of thumb, the so-called Fed model, compares the stock market's earnings yield with the 10-year Treasury yield. Clifford Asness and others have criticized it for exactly this reason: it compares a real quantity (earnings yield, which rises with inflation over time) with a nominal one (a bond yield). How rates and stock prices actually interact is covered in How interest rates affect stocks: discount rates and equity duration.
Real data: the S&P 500 since 1985
From its close of 181.18 at the end of February 1985 to 7,686.14 at the end of August 2026, the S&P 500 price index rose about 42.4 times, or about 9.5% a year. Over the same period the consumer price index rose from 106.3 to 334.1, a factor of 3.14. In constant dollars the index rose about 13.5 times, or 6.5% a year. Dividends, which this price index leaves out, would add a couple of percentage points a year to both lines but would not change the gap between them.
The real line also changes the story of the 2000s. In nominal terms the S&P 500 recovered its 2000 peak by 2007. In real terms, measured at these three-month closes, it did not get back above its August 2000 level until early 2015, so an investor who bought near the 2000 peak waited about fourteen years for a real gain on the price index alone. A nominal chart hides that.
Real data: the 10-year Treasury yield
Bond yields make the illusion easy to see. Subtracting the past year's inflation from the 10-year Treasury yield gives a rough, backward-looking real yield:
| Month | 10-year yield | CPI inflation (12 months) | Yield minus inflation |
|---|---|---|---|
| January 1985 | 11.38% | 3.5% | about +7.9% |
| July 2020 | 0.62% | 1.0% | about −0.4% |
| March 2022 | 2.13% | 8.6% | about −6.4% |
| September 2026 | 4.99% | 3.7% | about +1.3% |
A saver who bought a 10-year Treasury in 1985 locked in a high real yield; one who bought in 2020 or early 2022 locked in a negative one. Many investors in 2022 nonetheless judged bonds by their nominal prices and yields. The forward-looking real yield on inflation-protected Treasuries is covered in TIPS and breakeven inflation: how inflation-protected bonds work, and the concept in Real interest rates: nominal rates minus inflation, explained.
Taxes are levied on nominal gains
The US tax system mostly ignores inflation, which turns money illusion into a real cost. Interest is taxed on the full nominal amount, and capital gains are taxed on the difference between the nominal sale price and the nominal purchase price.
The same applies to a stock held for decades: part of the "gain" taxed at sale is just inflation. The rules themselves are in Capital gains tax on stocks: 2026 rates, holding periods and basis and How interest income is taxed: savings, CDs, T-bills and money funds.
Seeing past the illusion
- Convert to real. When reviewing a return, subtract the inflation over the same period. Over long horizons, chart in constant dollars.
- Compare like with like. Compare stocks' earnings yields with real bond yields, and nominal bond yields with nominal growth.
- Think in purchasing power. A goal such as "enough to retire" is a real amount. Projecting it in nominal dollars without inflation understates what is needed.
- Mind the after-tax real return. For taxable bonds and cash, the after-tax real return can be close to zero even when the nominal yield looks attractive.
How Finabase helps
Finabase charts prices in nominal dollars, like every market data service, so the conversion is yours to do, and it is quick. The full history of SPY on a chart gives the nominal gain over any period; dividing by the change in the consumer price index over the same dates gives the real one. A comparison chart such as SPY against TIP and BIL puts stocks beside inflation-protected Treasuries and Treasury bills, and a portfolio shows your nominal gains, which you can then deflate.
Frequently asked questions
What is money illusion in simple terms?
It is thinking about money in face-value dollars instead of what the dollars can buy. Someone with money illusion feels richer after a 5% raise with 6% inflation, even though their purchasing power fell.
What is the Modigliani-Cohn hypothesis?
It is the 1979 argument that investors value stocks by discounting real profit growth at nominal interest rates, which makes stocks look too cheap when inflation is high and too expensive when it is low. Later research found evidence consistent with it.
How much has the S&P 500 returned after inflation?
From February 1985 to August 2026 the S&P 500 price index rose about 42-fold, about 9.5% a year. After CPI inflation it rose about 13.5-fold, about 6.5% a year. Dividends would add to both.
What is the difference between nominal and real returns?
A nominal return is the percentage change in dollars. A real return adjusts for inflation: (1 + nominal) ÷ (1 + inflation) − 1. A 7% nominal return with 3% inflation is about a 3.9% real return.
Summary
Money illusion is the pull of the nominal number. Experiments show that people know the difference between nominal and real when prompted but react to the nominal figure otherwise. Modigliani and Cohn argued that the same error once misprised the entire stock market, and later research supports them. The data make the stakes concrete: the S&P 500's 42-fold nominal rise since 1985 is about 13.5-fold in real terms, its 2000 peak took until 2015 to regain after inflation, and Treasury buyers in 2020–2022 accepted strongly negative real yields. Converting returns to real, after-tax terms is the simplest defense.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
Ann gets a 2% raise with no inflation; Barbara gets a 5% raise with 4% inflation. Who is better off in real terms?
- Barbara, because 5% is more than 2%
- Ann: her real raise is about 2%, Barbara’s about 1%
- They are exactly equal
Show the answer
B. Ann: her real raise is about 2%, Barbara’s about 1% Real change ≈ nominal change minus inflation. In the experiment most people knew this when asked about economic terms but said Barbara would be happier.
What did Modigliani and Cohn argue about stocks and inflation?
- Inflation has no effect on stock valuations
- Investors discount real profit growth at nominal interest rates, so stocks look too cheap when inflation is high and too expensive when it is low
- Stocks always beat inflation in the short run
Show the answer
B. Investors discount real profit growth at nominal interest rates, so stocks look too cheap when inflation is high and too expensive when it is low Mixing a nominal discount rate with real growth double-counts inflation. Later research found pricing errors moving in the direction they predicted.
From February 1985 to August 2026 the S&P 500 price index rose about 42-fold and consumer prices about 3.14-fold. Roughly how much did it rise in real terms?
- About 39-fold
- About 13.5-fold
- About 3-fold
Show the answer
B. About 13.5-fold Divide, do not subtract: 42.4 ÷ 3.14 ≈ 13.5. That is about 6.5% a year after inflation instead of 9.5%.
A bond yields 5%, inflation is 3% and your tax rate is 24%. What is your approximate after-tax real return?
- 2.0%
- 0.8%
- 3.8%
Show the answer
B. 0.8% Tax is charged on the full nominal 5%: 5% × 0.76 = 3.8% after tax, minus 3% inflation ≈ 0.8%.
Exercise
One task to do in Finabase. Nothing is graded or saved.
Open SPY’s chart with the full range. Note the closing price at the start of a year at least ten years ago and the latest close, and compute the nominal gain. Then look up the change in the consumer price index (CPI-U) over the same dates on the BLS or FRED website and divide: (1 + nominal gain) ÷ (1 + CPI change) − 1.
Open SPY’s full chartWhat you should find
The real gain is always noticeably smaller than the nominal one over a decade, typically by a quarter to a third or more depending on inflation in the period. If you subtracted the CPI change instead of dividing, your answer will be off over long periods.
Sources and further reading
- Fisher — The Money Illusion, Adelphi, 1928
- Shafir, Diamond and Tversky — Money illusion, Quarterly Journal of Economics 112(2), 1997, pp. 341–374
- Revisiting money illusion: replication and extension of Shafir, Diamond, and Tversky (1997), Journal of Economic Psychology 83, 2021
- Modigliani and Cohn — Inflation, rational valuation and the market, Financial Analysts Journal 35(2), 1979, pp. 24–44
- Campbell and Vuolteenaho — Inflation illusion and stock prices, American Economic Review 94(2), 2004
- Cohen, Polk and Vuolteenaho — Money illusion in the stock market: the Modigliani-Cohn hypothesis, Quarterly Journal of Economics 120(2), 2005
- Asness — Fight the Fed model, Journal of Portfolio Management 30(1), 2003
- U.S. Bureau of Labor Statistics — CPI for All Urban Consumers (CPI-U), seasonally adjusted, via FRED series CPIAUCSL
- Federal Reserve Board — 10-year Treasury constant maturity rate, monthly, via FRED series GS10
- Finabase chart data — S&P 500 index (^GSPC), closes at three-month intervals, February 1985 to August 2026, price only
