Breakeven Inflation Is Not a CPI Forecast: Read the Yield Gap Before Trading It

A fall in ten-year breakeven inflation does not establish that investors cut their inflation forecast by the same amount. The measure subtracts a comparable inflation-linked Treasury yield from a nominal Treasury yield. Its appeal is immediacy: it updates with market prices. Its weakness is that the gap prices more than expected inflation. For anyone using the signal to interpret bonds, gold or currencies, the important task is to identify which leg moved and what else changed—not simply label every decline “disinflation”. All figures below are hypothetical.

The same 20-basis-point fall can tell different stories

Start with a nominal yield of 4.40% and a comparable TIPS yield of 1.90%, giving a quoted spread of 2.50%. Each of the following changes takes that spread to 2.30%. Yet the movement in the underlying securities is different.

The table identifies price arithmetic, not causes. A nominal-bond rally could reflect demand for liquid assets; a TIPS sell-off could reflect real-rate repricing or trading conditions. Neither explanation follows automatically from the spread. Even if an inflation release triggered the move, it may have changed expected policy rates and risk compensation at the same time.

What TIPS actually index—and what they do not protect

Treasury Inflation-Protected Securities adjust principal with their reference consumer-price index. The coupon rate stays fixed, while the cash coupon changes with adjusted principal. Suppose original principal is 1,000, the applicable index ratio is 1.06 and the annual coupon is 1.5%. Adjusted principal is 1,060; the semiannual coupon is 1,060 × 0.015 ÷ 2 = 7.95, rather than 7.50 at an index ratio of one.

The ratio represents the cumulative change in the contract’s reference index, not the latest monthly inflation rate. TIPS use non-seasonally-adjusted CPI-U with an indexation lag; payment calculations require the correct security and date-specific ratio. This is not the same object as core CPI, the PCE price index or a household’s personal spending basket.

At maturity, the principal payment cannot fall below original principal. That floor does not guarantee recovery of a secondary-market purchase price of, say, 1,070, nor does it prevent the bond’s market price from falling before maturity. A fund holding TIPS also does not inherit a guarantee that each investor’s purchase price or fund unit will be repaid at par.

A yield spread is a shorthand, not an exact cash-flow comparison

A stripped-down one-year example shows the distinction. Invest 1,000 either at a nominal annual return of 4.5% or at a real annual return of 2%, with inflation added multiplicatively. Ignore taxes, fees, indexation lag, any deflation floor and interim coupons. The nominal investment finishes at 1,045. The indexed investment finishes at 1,020 × (1 + realised inflation).

Equal final wealth requires inflation = 1.045 ÷ 1.02 − 1 = 2.45098%. Simply subtracting 2% from 4.5% gives 2.50%, a useful approximation but not the exact threshold in this annual-compounding example.

This is an idealised zero-coupon comparison, not an exact valuation of a pair of traded coupon bonds. Their purchase prices, accrued amounts, payment schedules, reinvestment, indexation and maturity floor must be modelled. Beating the nominal alternative over a matched horizon is also different from having a positive mark-to-market return over the next month.

Separate expected inflation from compensation for risk and liquidity

A useful conceptual decomposition is: breakeven ≈ expected inflation + inflation risk premium − relative TIPS liquidity premium. Here the liquidity term means an extra yield required for holding TIPS relative to a comparable nominal bond. Its sign in the expression matters: a higher required TIPS yield narrows the observed gap, other components unchanged.

For example, expected inflation of 2.10%, an inflation risk premium of 0.50% and a liquidity premium of 0.10% produce 2.50%. Raise only the liquidity component to 0.30% and the result falls to 2.30%, without changing the inflation expectation. This is a possible decomposition, not an estimate of today’s market.

Inflation risk compensation is not always positive, and neither premium is displayed as a separately tradable quote. Estimates depend on models and data. Two observed yields cannot uniquely identify all components. Surveys can provide a cross-check, but they have their own respondents, dates and horizons; they are not a magic subtraction that reveals a precise premium.

Inflation protection still leaves real-rate price risk

If a hypothetical TIPS position has modified duration of seven with respect to its real yield, a 50-basis-point increase in that yield gives an approximate price effect of −7 × 0.005 = −3.5%, before convexity, carry and inflation accrual. Positive indexation can coexist with a falling quoted price. Treat this as a sensitivity estimate, not a total-return forecast.

The distinction between a yield change and its cash impact is developed in bond duration and DV01 calculations. For an inflation-linked position, specify which yield is being shocked. A nominal-yield duration is not automatically the same exposure as real-yield duration, and identical maturities do not guarantee identical cash-flow weights.

A ten-year breakeven also concerns a long horizon, not next month’s CPI or the inflation rate specifically in year ten. A near-term energy shock can affect short and long horizons differently. Do not infer a particular gold or exchange-rate move from breakeven alone: real yields, policy expectations, dollar funding and risk demand can change simultaneously.

A practical check before calling the move an inflation signal

First match currency, issuer, maturity, observation time and curve convention. Do not combine an intraday nominal quote with yesterday’s real yield, or mix a par-yield series with a fitted zero-coupon curve. Subtracting trailing CPI from a nominal yield constructs a different measure; it is not the TIPS breakeven.

Next separate the nominal and real legs, then inspect bid–ask spreads, available trading depth and relative supply pressures. Compare inflation compensation with surveys or model estimates on compatible horizons and price-index definitions. For shorter maturities, examine known index accrual, seasonality and the maturity floor before interpreting annualised yield changes.

Finally decide what claim the evidence supports. If the gap moves while liquidity deteriorates and survey expectations barely change, a pure inflation-expectations story is weak. If several independent measures move together, that story gains support, but the exact contribution remains uncertain. Breakeven is a market price to interrogate, not a CPI number that has already been forecast with certainty.