Inflation, real rates and breakevens

Everything in M4 so far has discussed rates as one number. But a 5% yield means something completely different when inflation is 2% than when it is 8%, and markets price that distinction explicitly.

The Fisher relation

   nominal yield  ≈  real yield  +  expected inflation

   5% nominal with 2% inflation  →  ~3% real — you gain purchasing power
   5% nominal with 8% inflation  →  ~−3% real — you lose it, while
                                     "earning" 5%

The real rate is the one that governs behaviour. It is what a borrower actually pays in goods, what a saver actually earns, and what belongs in the discount rate of M4.3 when the cashflows themselves grow with inflation. A negative real rate means lenders are paying for the privilege, which is why it reliably produces borrowing, leverage and asset-price inflation.

Markets quote both

Governments issue index-linked bonds — TIPS in the US, linkers in the UK — whose principal and coupons rise with a published inflation index. Their yield is therefore a real yield, directly observable.

Put the two bonds side by side and the difference is the market's inflation compensation:

   nominal 10-year yield   −   real (TIPS) 10-year yield   =   BREAKEVEN

   today (2026-08-07):
        nominal   4.65%
        real      2.40%
        breakeven 2.25%

Called the breakeven because it is the inflation rate at which the two bonds deliver the same return — above it, the linker wins; below it, the nominal.

One honest caveat, immediately. In FRED the breakeven series is constructed as nominal minus real, so checking that the three numbers add up validates nothing — it is a definition, not a finding. The content is in the decomposition, not in the identity.

A second caveat with real teeth: the breakeven is not pure expected inflation. It also contains an inflation risk premium (compensation for the risk that inflation surprises) and a liquidity premium (TIPS trade far less than nominal Treasuries, so their yield carries an illiquidity concession — exactly the Amihud–Mendelson effect from M1.3). Both push the measured breakeven around independently of what anyone expects. Treat it as a useful market-implied indicator, not a survey of beliefs.

2022, decomposed

M4.1 showed a 10-year bond losing 17% as yields rose 225bp. The obvious story is that inflation hit 9% and the bond market repriced inflation. Split the move and that story collapses:

                     nominal      real     breakeven
   ──────────────────────────────────────────────────
   start 2022          1.63%     −0.97%       2.60%
   end 2022            3.88%     +1.58%       2.30%
   ──────────────────────────────────────────────────
   change             +225bp     +255bp       −30bp

The real yield supplied more than the entire move, and breakevens fell.

Sit with that, because it is genuinely counterintuitive. In the year of the largest inflation shock in four decades, the bond market's implied expectation of future inflation went down. What repriced was the real cost of money — from −0.97%, deeply negative, to +1.58%.

The interpretation is coherent once stated: the market never believed 9% inflation would persist. It believed the central bank would raise rates far enough to stop it, and priced exactly that — a much higher real rate, and inflation returning to target. 2022 was a credibility story, not an expectations story. The bond market was, in effect, marking to market its confidence in M4.2's institution.

And it explains M4.3's equity carnage precisely. A real rate moving from −1% to +1.6% is the discount rate in the DCF rising by 255bp, which crushes long-duration cashflows. It was not fear of inflation that de-rated growth equities; it was the end of free money.

Why negative real rates were such a big deal

Look again at that starting point: −0.97%. Lending to the US government for a decade and accepting a guaranteed loss of purchasing power.

That was not a market failure; it was policy working as designed. QE (M4.2) plus a zero policy rate pushed real yields below zero to force capital out of safe assets and into risk. It did exactly that — the entire 2010s regime of TINA, long-duration growth outperformance, and asset-price inflation follows from a persistently negative real rate.

Which frames the whole period as one variable:

   real rates deeply NEGATIVE     real rates POSITIVE
   ────────────────────────────────────────────────────────────
   2010s                          post-2022
   growth beats value             the reverse
   duration is rewarded           duration is punished
   leverage is nearly free        leverage costs something
   "there is no alternative"      cash yields something real

Reading inflation data

Two distinctions that come up constantly:

Headline vs core. Core strips food and energy — not because they don't matter, but because they are volatile and supply-driven, so core is the better read on the persistent trend policy can influence.

Levels vs rates of change. "Inflation is falling" almost always means the rate is falling. Prices are still rising, only more slowly. Actual price falls are deflation, which central banks fear more than moderate inflation, because it encourages deferred spending and makes debt harder to service in real terms — and because the policy rate cannot go far below zero to fight it (M4.2).

Source: for the market-implied measures, the Cleveland Fed and NY Fed both publish inflation-expectation methodologies that are explicit about the risk and liquidity premia in breakevens. Figures above from FRED series DGS10, DFII10 and T10YIE.