Bitcoin Purchasing Power: DCA Returns After CPI
A nominal gain and a purchasing-power gain answer different questions. To compare a Bitcoin DCA result after inflation, convert both the final holding and each contribution into dollars from the same base date. That avoids treating deposits made years apart as if each dollar had identical purchasing power when contributed.
On this page
If you measure bitcoin purchasing power instead of nominal dollars, the answer changes by start date. In the CPI-adjusted DCA runs below, monthly $100 Bitcoin buys from 2018, 2021, and 2022 still ended with positive real profit through September 1, 2026. But shorter recent windows did not: a January 2024 start was slightly negative after CPI, and a January 2025 start was clearly negative in start-date dollars.
This article discusses our own simulator, BTC DCA Engine. The operator and publication process are described on the About BTC DCA Engine page. The numbers here are historical calculations, not financial advice, and not a forecast.
The exact result: nominal dollars versus purchasing power
The experiment is deliberately narrow: invest $100 once per month into Bitcoin, beginning on the listed start date and ending on September 1, 2026. The end date is capped at the last available CPI observation in the calculation brief, even though Bitcoin coverage extends later. That avoids extrapolating inflation.
The inflation adjustment uses BLS CPI-U all items, U.S. city average, seasonally adjusted, series CPIAUCSL, retrieved via FRED. FRED identifies CPIAUCSL as the “Consumer Price Index for All Urban Consumers: All Items in U.S. City Average,” monthly, seasonally adjusted, with September 2026 at 336.298. The BLS explains that CPI measures the average change over time in prices paid by consumers for a representative basket of consumer goods and services, and that CPI does not necessarily match any one household’s own price experience.
| Start date | Buys | Nominal contributed | Nominal BTC value | Real contributions in start-date dollars | Real BTC value in start-date dollars | Real profit |
|---|---|---|---|---|---|---|
| 2018-01-01 | 105 | $10,500.00 | $50,176.44 | $9,168.72 | $37,130.34 | $27,961.61 |
| 2021-01-01 | 69 | $6,900.00 | $12,556.31 | $5,971.89 | $9,807.90 | $3,836.02 |
| 2022-01-01 | 57 | $5,700.00 | $10,424.56 | $5,171.50 | $8,758.27 | $3,586.76 |
| 2024-01-01 | 33 | $3,300.00 | $3,441.88 | $3,178.32 | $3,169.64 | -$8.68 |
| 2025-01-01 | 21 | $2,100.00 | $1,909.92 | $2,053.20 | $1,811.46 | -$241.73 |
The table has two denominators on purpose. “Nominal contributed” is the plain sum of scheduled $100 buys. “Real contributions” converts each contribution into the purchasing power of the start date. That second denominator matters because a $100 buy made years after the start date is not treated as if it had the same purchasing power as $100 at the beginning.
What the experiments show
The 2018 window is the strongest result in this set. It made 105 monthly buys, contributed $10,500 nominally, and ended with Bitcoin worth $50,176.44 nominally. After CPI adjustment, the same ending Bitcoin was worth $37,130.34 in January 1, 2018 dollars. The inflation-adjusted contribution base was $9,168.72, producing $27,961.61 of real profit.
That result says two things at once. First, the nominal gain was not just inflation illusion: the ending Bitcoin value remained far above the CPI-adjusted contribution base. Second, the real-value framing is meaningfully lower than the nominal headline. The nominal ending value was $50,176.44, but the purchasing-power-adjusted value in start-date dollars was $37,130.34.
The 2021 start is a more modest but still positive case. Across 69 monthly buys, the investor contributed $6,900 nominally and ended at $12,556.31 nominally. In January 1, 2021 dollars, the contribution base was $5,971.89 and the ending value was $9,807.90, leaving $3,836.02 in real profit. This window shows why looking only at the plain account balance can overstate what the balance means: $12,556.31 sounds like the ending amount, but the purchasing-power comparison for this experiment is $9,807.90 versus $5,971.89.
The 2022 start is similar in direction. The run made 57 buys, contributed $5,700 nominally, and ended at $10,424.56 nominally. In January 1, 2022 dollars, real contributions were $5,171.50 and real ending Bitcoin was $8,758.27. The real profit was $3,586.76. The nominal result and real result both point positive, but the real calculation gives the cleaner answer to the purchasing-power question.
The shorter windows are the limitation example. Starting January 1, 2024 produced a nominal gain: $3,300 contributed and $3,441.88 ending value. But after CPI, that same run was slightly negative. Real contributions were $3,178.32 in January 2024 dollars, real ending value was $3,169.64, and real profit was -$8.68. In other words, the nominal account was up, but purchasing power was not.
The January 1, 2025 start was negative both ways. It made 21 buys, contributed $2,100 nominally, and ended at $1,909.92 nominally. In January 2025 dollars, real contributions were $2,053.20 and real ending value was $1,811.46, for -$241.73 of real profit. That is the clearest warning in the table: CPI adjustment does not rescue a weak nominal window, and recent DCA windows can still lose purchasing power.
For readers comparing this to drawdown-focused work, our separate article on how long Bitcoin DCA can stay underwater addresses a different question: duration below contribution basis rather than CPI-adjusted ending purchasing power.
Reproducible method
The method is mechanical:
- Use monthly $100 Bitcoin buys on the executed buy dates shown in the ledger.
- Use Bitcoin historical pricing from the simulator’s data files and CPIAUCSL from the macro seed.
- Cap the ending date at September 1, 2026 because that is the latest CPI observation used here; do not extrapolate CPI beyond it.
- Use BLS CPI-U all items, U.S. city average, seasonally adjusted, via FRED.
- Linearly interpolate between monthly CPI index observations when a buy date falls between observations. This is a retrospective convention using revised data, not point-in-time release data.
- Convert ending Bitcoin value into start-date dollars: nominal ending BTC value × CPI(start) / CPI(end).
- Convert each executed $100 buy into start-date dollars: $100 × CPI(start) / CPI(each executed buy), then sum those adjusted contributions.
- Real profit equals real ending Bitcoin value minus real contributions.
- Exclude fees and taxes.
For the exact scheduling and engine conventions, see the core engine methodology. For data coverage, refresh cadence, and source notes, see the BTC DCA Engine data sources. You can also test closed-window assumptions in the Bitcoin DCA simulator, but do not treat a synthetic calculation as a real account return.
Interpretation and limits
The CPI-adjusted runs answer a narrower question than “was Bitcoin a good investment?” They answer: under these exact monthly $100 schedules, how much start-date purchasing power did the ending Bitcoin represent after adjusting both the ending value and every contribution with CPI?
CPI is an index, not an investable asset. It is also not a household-specific inflation rate. BLS notes that CPI is based on average price changes for a representative basket and that a national average seldom mirrors a particular consumer’s experience. FRED’s CPIAUCSL notes also describe it as an index of a basket of goods and services paid by urban consumers and warn that CPI is not applicable to all consumers or suitable for determining relative living costs.
That matters here. A retiree, renter, homeowner, student, or family with high medical costs may experience price changes differently from CPI-U. Taxes, trading fees, bid-ask spreads, custody costs, lost keys, and account rules are also outside this calculation. The result is useful as historical arithmetic, not as a statement that a person actually could have captured the same after-tax purchasing power.
FAQ
Why use start-date dollars?
Start-date dollars make each window internally consistent. A 2018 run is expressed in January 1, 2018 purchasing power; a 2025 run is expressed in January 1, 2025 purchasing power. That avoids mixing nominal dollars from different dates as if they were equivalent.
Why does the 2024 run show a nominal gain but real loss?
Because the nominal ending value was only slightly above nominal contributions. After adjusting contributions and ending value for CPI through September 1, 2026, the ending Bitcoin value was $3,169.64 in January 2024 dollars versus $3,178.32 of real contributions.
Are fees, taxes, or personal inflation included?
No. The calculation excludes fees and taxes and uses CPI-U as a broad index. It does not model a brokerage account, tax lot accounting, household spending, or advice for any individual.
How this was made
Every number above — in the sentences and in the charts — is read from one frozen snapshot of the data, so a claim and the figure beneath it cannot disagree, and a nightly refresh cannot move a published figure. Every feed, its first reading and its refresh cadence are on the sources page; the rounding rules are in the knowledge base. Where another tool is named, the same plan was run through that tool's own interface and recorded with the date it was checked. Published by BTC DCA Engine; how these notes are made, and who is accountable for them, is on the about page.