How CPI Moves Bitcoin, Gold, the Dollar, Yields, and Stocks
Published 1 September 2026 · Evidence cutoff 31 May 2025 · 14 minute read
The short answer: markets do not trade the inflation level in isolation. They trade the difference between the CPI release and what was already expected, then translate that surprise into a new expected path for Federal Reserve policy. Treasury yields usually form the bridge to the dollar, gold, Bitcoin, and the S&P 500.
The common chain is: actual versus consensus → composition and persistence → expected Fed response → short and real yields → dollar and liquidity → asset prices. It is a useful starting framework, not a prediction rule. Growth information, positioning, revisions, another release, or a rush for liquidity can break the chain.
What the CPI release actually contains
The US Bureau of Labor Statistics publishes the Consumer Price Index on scheduled dates, normally at 8:30 a.m. Eastern time. Use the official BLS release calendar rather than copying a future date into an evergreen guide.
The headline CPI-U covers the prices paid by urban consumers across the market basket. “Core” CPI excludes food and energy. Analysts watch core because those categories can be volatile, but households still pay for them and an energy shock can influence later inflation. The release presents seasonally adjusted monthly changes, unadjusted annual changes, index levels, and category detail.
For the immediate market response, four figures receive particular attention: headline month over month, core month over month, headline year over year, and core year over year. Monthly figures offer a more recent momentum reading. Annual figures are intuitive but can rise or fall because an unusual month leaves the comparison base.
CPI is not the Federal Reserve's formal inflation target. The Fed's longer-run goal is 2% inflation measured by the price index for personal consumption expenditures. CPI still matters because it arrives earlier, contains detailed price information, and changes expectations for PCE inflation and policy.
Surprise, not level
A release is “hot” or “cool” only relative to a timestamped expectation. If annual inflation is high but exactly matches consensus, much of that information may already be in prices. If inflation is low but one tenth above a tightly clustered forecast, yields can still rise.
Save the median consensus before the release and calculate surprise = actual − expected for each measure. Do not silently substitute the prior reading for consensus. The previous number describes direction; the forecast describes what the market had a chance to price.
The transmission sequence
- The figures arrive. Automated systems compare headline and core monthly and annual readings with expectations within seconds.
- Short-rate expectations reprice. A broad, persistent-looking upside surprise often raises expected policy rates and the two-year Treasury yield; a downside surprise often does the opposite.
- The yield move is decomposed. Nominal yields can change because of real yields, inflation compensation, or both. This distinction is important for gold and valuation-sensitive assets.
- The dollar responds. Higher relative US rate expectations commonly support the dollar; lower expectations commonly weaken it. The relationship changes across regimes and currency pairs.
- Equities discount rates and cash flows. Higher yields reduce the present value of distant earnings, but the release can also update expectations for demand, margins, and nominal revenue.
- Gold weighs inflation against carry. A hot report may strengthen the inflation-hedge story while higher real yields and a stronger dollar pull the other way.
- Bitcoin absorbs the liquidity signal. Evidence since late 2020 suggests Bitcoin has become more sensitive to US monetary and CPI news and often behaves like a risky asset around releases.
- The first move is challenged. Components, positioning, Federal Reserve communication, and other news can reverse the reaction over the next hour or day.
Common first-order tendencies
Every cell below is a common tendency, not a forecast.
| Release | Yields / USD | Gold | Bitcoin | S&P 500 |
|---|---|---|---|---|
| Hot headline and core, broad-based | Yields and USD often rise | Often pressured if real yields and USD rise | Often pressured by tighter-liquidity expectations | Often pressured by higher discount rates |
| Cool headline and core, broad-based | Yields and USD often fall | Often supported by lower real yields/USD | Often supported by easier-liquidity expectations | Often supported unless growth fear dominates |
| Hot headline, cool core; energy-led | Mixed; persistence determines the rate response | Inflation story and yield channel can conflict | Mixed; oil-driven risk-off can matter | Margins and growth may outweigh rates |
| Cool headline, sticky core/services | Initial relief may fade | Two-way | Two-way; whipsaw risk | Initial rally can reverse |
| Exactly consensus | Components and positioning decide | Often small/two-way | Leverage or crypto news may dominate | Components and revisions decide |
| Cool because demand is collapsing | Yields may fall; USD response is ambiguous | May gain as a haven, but can be sold for cash | Can fall with risk assets | Can fall on earnings fears |
Why Treasury yields usually lead the explanation
The two-year Treasury is relatively sensitive to the expected policy path. Longer yields reflect that path plus longer-run real-rate, inflation, term-premium, and growth expectations. Federal Reserve research using intraday Treasury and inflation-protected security data finds that price-index news affects inflation compensation, while macro news also moves real rates.
A historical New York Fed event study estimated that a 0.1 percentage-point core-CPI surprise was associated in its sample with higher two- and ten-year yields after 30 minutes. That coefficient is a historical average, not a rule for the next release. BIS research shows the two-year yield's sensitivity to core CPI surprises has varied over time.
The US dollar: relative rates, not inflation alone
A hotter report often lifts the dollar when it raises expected US rates relative to rates abroad. Yet an inflation-expectation shock is not identical to a tighter-policy shock. Federal Reserve research distinguishes a shock that raises inflation compensation and nominal yields while weakening the dollar from tighter-than-expected policy, which tends to raise yields and the dollar while lowering stocks.
Always define “the dollar.” DXY is a weighted index; EUR/USD and USD/JPY have different conventions and foreign-policy drivers. Older Fed research found no systematic inflation-news response in exchange rates, while later high-frequency studies found macro surprises can be incorporated within minutes. Regime and measurement matter.
Gold: inflation hedge versus real yields
Gold produces no contractual cash flow. When real yields rise, the opportunity cost of holding it can increase; when the dollar strengthens, gold becomes more expensive in other currencies. Those channels explain why a hot CPI release can initially hurt gold even when long-run inflation protection is part of the investment thesis.
The academic record is mixed. A 2000 intraday study found gold futures responded strongly to CPI releases. A 2010 study found surprise CPI changes affected bond yields as predicted, but did not affect gold spot prices in its sample. Follow the observed real-yield and dollar response rather than forcing every move into the phrase “inflation hedge.” For a different shock framework, see Bitcoin vs gold during geopolitical crises.
Bitcoin: scarce asset or liquidity-sensitive risk?
Bitcoin's issuance rules support a long-run scarcity thesis, but a monthly CPI announcement tests a much shorter mechanism. Research by Bauer and Morales found that Bitcoin did not systematically respond to US monetary news for much of its history, then became more sensitive after 2020 and reacted qualitatively like other risky assets around monetary and CPI releases.
Other results disagree. Pinchuk estimates a negative announcement-day Bitcoin response to inflation surprises. Rodriguez and Colombo find a positive dynamic response to CPI shocks in their full monthly sample, but not to core PCE surprises; the result changes across periods and disappears after COVID in their split. The responsible conclusion is that Bitcoin's inflation response is context-specific. Our S&P 500 vs Bitcoin guide explains why their structural risk differs even when event-day direction matches.
The S&P 500: discount rates versus earnings
Higher expected rates can lower the present value of future corporate cash flows, placing pressure on equities. Companies and sectors with more distant expected profits may be particularly rate-sensitive. But CPI can also contain growth information: stronger nominal demand can support revenue, while input-cost inflation can squeeze margins.
A cool release is therefore not automatically bullish. If investors interpret it as evidence of collapsing demand, lower yields and lower stocks can coexist. Similarly, hot inflation may accompany resilient growth. The discount-rate and cash-flow channels can point in opposite directions.
Why the textbook reaction fails
- Composition: energy-led headline inflation may be treated differently from broad core-services pressure.
- Attention: Federal Reserve research finds CPI reactions became much stronger in the 2021–2023 inflation surge and were larger when pre-release attention was high.
- Positioning: a hot number can trigger only a small move if traders were already prepared for worse.
- Simultaneous releases: another statistic published at 8:30 can contribute to the same price move.
- Liquidity and leverage: thin books, forced liquidations, spreads, and option hedging can amplify or reverse moves.
- Time horizon: the five-minute policy response can differ from the closing or monthly inflation-hedge result.
Before the release: research checklist
- Confirm date and 8:30 a.m. Eastern release time on BLS; convert daylight saving time correctly.
- Save a timestamped consensus for headline/core monthly and annual figures, plus prior readings.
- Write down the market's main question: shelter, services, energy, goods, or another component.
- Record two- and ten-year nominal yields, available real yields and breakevens, a declared dollar measure, gold benchmark, Bitcoin reference rate, S&P futures or cash, volatility, and rate expectations.
- Note every other release at the same timestamp and whether the equity cash market is closed.
- Predefine windows such as −5/+5 minutes, −5/+30 minutes, previous close/same close, and +1 day.
- Review volatility, spreads, and volume before interpreting an announcement print.
After the release: analysis checklist
- Calculate actual minus saved consensus for all four headline/core measures.
- Read the detailed BLS tables. Separate energy noise from persistent-looking shelter or services.
- Check revisions and seasonal-adjustment notes, especially around the January-data release.
- Observe the two-year yield and short-rate expectations first, then the ten-year yield.
- Where data allow, separate the nominal-yield move into real yield and inflation compensation.
- Check whether dollar, gold, Bitcoin, and equities agree with the policy-path interpretation.
- If they disagree, document competing explanations instead of assigning false certainty.
- Measure every pre-declared window, spread, maximum move, and subsequent reversal.
- Archive the BLS release, consensus snapshot, timestamps, benchmarks, and methodology.
A reproducible event-study method
Use synchronized intraday data and declare instruments. State whether the S&P series is cash or futures, whether gold is the LBMA benchmark, spot, futures, or an ETF, which Bitcoin exchange or reference rate is used, and whether USD means DXY or a currency pair. Bitcoin trades continuously while the equity cash session is normally closed at release time.
Estimate the surprise from a preserved consensus source. Analyze headline and core separately, include simultaneous-release controls, and report more than endpoint return: maximum adverse and favorable movement, spread, volume, and reversal. Divide the sample into regimes rather than assuming a coefficient from low inflation applies during high attention. Avoid overfitting a handful of memorable releases; see backtests that survive live crypto markets.
BitBank's model documentation and risk tools can support broader research, but a forecast is an uncertain input—not an instruction to trade a high-volatility release.
Methodology and sources
This is an evergreen evidence synthesis with an evidence cutoff of 31 May 2025. It intentionally contains no current CPI values, market consensus, live yields or prices, next-release date, or directional recommendation.
- BLS CPI release schedule — official dates and times.
- BLS CPI questions and answers — scope and construction.
- BLS common CPI misconceptions — core CPI and rental equivalence.
- BLS rent and owners' equivalent rent FAQ — shelter methodology.
- BLS seasonal adjustment — recalculation and revision policy.
- Federal Reserve longer-run goals — the PCE-based inflation goal.
- New York Fed, How Economic News Moves Markets — historical event responses.
- Federal Reserve, high-frequency news and yields — real yields and inflation compensation.
- Federal Reserve, macro sources of rate moves — inflation, growth, and policy-shock patterns.
- Federal Reserve, investor attention and CPI reactions — regime-dependent sensitivity.
- Federal Reserve, high-frequency FX macro news — rapid price incorporation.
- BIS Quarterly Review, September 2024 — changing two-year-yield sensitivity.
- Bauer and Morales (2023), Monetary policy and Bitcoin — changing Bitcoin sensitivity.
- Rodriguez and Colombo (2024), Is bitcoin an inflation hedge? — context-dependent evidence.
- Pinchuk, Bitcoin Does Not Hedge Inflation — announcement-day counterevidence.
- Christie-David et al. (2000) — intraday gold and macro releases.
- Gold prices, cost of carry, and expected inflation (2010) — gold counterevidence.