A 4% account can gain before tax and lose afterward
On 11 September 2026, the U.S. Bureau of Labor Statistics released its August Consumer Price Index. The all-items CPI-U was 334.980 in August 2026, up from 323.976 in August 2025. That ratio, 334.980 ÷ 323.976, is 1.03396548, or 3.3965% over twelve months; BLS reports the rounded headline as 3.4%. A $10,000 account that hypothetically earned 4.0% APY for exactly that year would finish at $10,400 before tax. Measured in August 2025 buying power, $10,400 ÷ 1.03396548 is about $10,058.36. The account gained about $58.36 of real buying power before tax.
Now make a separate, explicit tax assumption: all $400 of interest is taxable and its incremental federal tax effect is 22%, with no state tax or other interaction. Paying $88 tax leaves $10,312. That amount buys what about $9,973.25 bought a year earlier. The same example has a 0.584% pre-tax real gain and a 0.267% after-tax real loss. It is a worked scenario, not an actual advertised account rate, a tax estimate for every saver or a forecast of next year’s inflation.
Use the two CPI index levels, not just a subtraction
The BLS archived August 2026 release supplies both index levels and the 3.4% rounded twelve-month figure. The index ratio is the multiplier by which a representative consumer basket’s price changed. To express an ending dollar balance in starting-month dollars, divide it by that ratio. For a one-year nominal return n and index change p, the exact real-return calculation is (1 + n) ÷ (1 + p) − 1. “Nominal return minus inflation” is a convenient small-rate approximation, not an exact identity.
| Hypothetical position, August 2025–26 | Nominal ending dollars | August 2025 buying power |
|---|---|---|
| No interest on $10,000 | $10,000.00 | $9,671.50 |
| 3% APY on $10,000 | $10,300.00 | $9,961.65 |
| 4% APY on $10,000 | $10,400.00 | $10,058.36 |
Same $10,000 start, nominal versus August 2025 buying power
At 0%, the paper balance stays $10,000 but its real value falls 3.285%; it does not fall exactly 3.4% because converting current dollars back to old dollars uses the reciprocal of the price multiplier. At 3% APY, the balance rises by $300 but still loses roughly $38.35 in starting-dollar buying power. At 4%, it beats this historical CPI interval before tax by only about $58.36. None of these rows says a bank paid those rates during the interval. Use the APY that actually applied to your account, including any rate changes and balance tiers.
The SEC’s real-return glossary describes real return as earnings after accounting for taxes and inflation. The precise denominator matters near break-even: using only 4.0% minus the rounded 3.4% would suggest 0.6%, whereas the CPI-index calculation gives 0.5836% before tax. That difference is small here but prevents a claim of mathematical precision based on a rounded press-release rate.
The tax break-even is about 15.1% in this one example
The IRS says most bank-account interest available to withdraw is taxable income and should be reported even if a Form 1099-INT does not arrive. This does not mean every saver owes 22% of interest. Actual federal and state results depend on taxable income, filing status, account wrapper, deductions and credits. The 22% rate here is only an illustrative marginal tax on the $400 interest.
Holding the same historical CPI ratio and $10,000 opening balance, the nominal ending amount after an assumed interest-tax rate t is $10,400 − $400 × t. Divide that by 1.03396548 to get August 2025 dollars. At a 12% assumed tax rate, the balance has about $10,011.94 of real value. At 22%, it has $9,973.25; at 32%, $9,934.57. Those are sensitivity tests, not actual tax brackets or personal tax computations.
Real ending value of a 4% APY example after assumed interest tax
The tax rate that exactly erases the pre-tax real gain solves $10,400 − $400 × t = $10,000 × 1.03396548. The result is about 15.09%. Above that rate in this simplified model, a 4% APY turns into a real loss; below it, the account still inches ahead. If the account has monthly fees, the break-even tax rate is lower. If the interest is tax-free under a genuine legal provision, it may be higher. Do not insert a guessed rate into a personal return or treat the 15.09% threshold as a general rule for all inflation periods.
Deposit safety and buying-power safety are different questions
The FDIC’s insured-deposit guide says eligible balances at insured banks are covered against bank failure, including principal and accrued interest, within the applicable insurance limits. Insurance does not index a deposit to the consumer price level. A fully insured $10,000 deposit can therefore be safe from bank default while buying less after a period of inflation. On the other hand, a deposit yielding more than measured inflation can gain purchasing power, as the pre-tax 4% row did in this real BLS interval. “Cash always loses” is as inaccurate as “insured means inflation-proof.”
The purpose of the money matters. The Consumer Financial Protection Bureau says emergency savings should be safe and accessible, so a near-term reserve has value even when its real yield is modest or negative. The SEC’s asset-allocation guidance likewise ties an investment mix to time horizon and risk tolerance. A family expecting a repair bill next month should not move that reserve into volatile shares merely to chase a historical CPI reading. A long-term investor can evaluate diversified assets, but their prices can fall just when cash is needed.
Treasury inflation-protected products address a different risk with their own restrictions. TreasuryDirect explains that TIPS principal adjusts with CPI-U and pays at least original principal at maturity; selling before maturity exposes a holder to market-price changes. The Treasury’s Series I bond page lists a 4.26% composite rate for bonds issued May–October 2026 for their first six months. That rate is not locked for the bond’s lifetime: its inflation component resets every six months. Treasury’s cashing rules say I bonds generally cannot be redeemed for the first year and sacrifice three months’ interest when redeemed before five years. Neither is an instant-access replacement for an emergency account.
Make the measurement fit your own money
The CPI-U is a broad urban-consumer average. The BLS explains that a household spending unusually heavily on rent, medicines or fuel can experience price changes different from the national basket. The August 2025–August 2026 index comparison is a backward-looking U.S. average. It should not be applied to a different country, used as a precise household inflation rate, or projected unchanged through 2027.
For your account, note the balance at the beginning and end of a period, any deposits and withdrawals, interest credited, fees and applicable tax. A deposit of new savings during the year is not investment return; isolate interest from cash you added. Match your interest period to the same months of CPI data, then convert nominal ending value to starting-month dollars. A variable-rate account may require monthly subperiods rather than a single advertised APY. If the goal is a future rent payment, compare the product with that specific date and required access.
The practical question is not “Which asset always beats inflation?” It is whether a particular balance, after its actual interest, fees and tax, still buys the goods and services you need when you need them. The August 2026 case shows the answer can change from positive to negative with an explicit tax assumption. Recompute with current numbers rather than keeping the example’s 4%, 22% or 3.4% forever.
