Forecasting: What the Market Already Thinks

You do not need to forecast inflation yourself. A liquid market prices it every day, for free, at every horizon, and it will beat your opinion more often than not.

Breakeven inflation. Treasury issues both nominal bonds and TIPS of matched maturity. The nominal bond compensates you for expected inflation; the TIPS does not need to, because its principal adjusts. The difference is the market’s implied inflation rate:

Both legs are published daily, and the spreads are on FRED as T5YIE and T10YIE. If the 10-year nominal yields 4.3% and the 10-year TIPS yields 1.9%, the market expects about 2.4% average inflation over the next decade. That number is the hurdle every long-horizon plan has to clear, and it took you ten seconds to obtain.

This is the Fisher relation in practice. Exactly:

where $i$ is the nominal rate, $r$ the real rate, and ${\pi }^e$ expected inflation.

The honest caveat. A breakeven is not a clean expectation. It bundles three things:

Investors demand compensation $\rho$ for bearing inflation uncertainty, which pushes breakevens above true expectations. TIPS trade in a smaller, less liquid market, so they carry a yield concession $\lambda$ that pushes breakevens down. The two partly offset, and both move around — violently in a liquidity crisis, when breakevens can collapse for reasons that have nothing to do with expected prices. Use breakevens as the market’s best available guess, not as a measurement.

Forward breakevens. The 5-year, 5-year forward strips out the near term to isolate long-run expectations — what the market thinks inflation will average over the five years beginning five years from now:

Central bankers watch it as a gauge of whether expectations are “anchored.” Treat it with care: section “Five Fallacies of Fixed-Income Markets” explains why its predictive power over realized inflation is weaker than its prominence suggests.

Surveys, as a cross-check. Markets can be distorted; surveys can be naive. Consult both, and worry when they diverge:

• Survey of Professional Forecasters (Philadelphia Fed, quarterly) — the consensus of paid economists; historically the hardest single benchmark to beat.
• Cleveland Fed inflation expectations — a model blending breakevens, surveys, and swaps to strip out the risk premia above. The closest thing to a clean ${\pi }^e$ available for free.
• NY Fed Survey of Consumer Expectations and the University of Michigan survey — what households believe. Consistently biased high and heavily influenced by gasoline prices, but they matter because expectations feed wage demands, and wage demands feed actual inflation.

What to do with any of it. Not trade. The forecast’s job is to set the assumptions in your plan and to price specific decisions:

Set your planning deflator from the breakeven, not from history.

If the 30-year breakeven says 2.3%, using a 1970s-flavored 4% makes you save more than necessary; using a 2010s-flavored 1.5% makes you save too little. Use the market’s number and revisit annually.

Price nominal versus inflation-linked bonds directly.

The breakeven is the exact indifference point. Buy nominal Treasuries if you believe realized inflation will come in below the breakeven; buy TIPS if you believe it will come in above. There is no third answer, and this is one of the rare cases where a view is cleanly expressible.

Decide fixed versus floating debt.

A fixed-rate 30-year mortgage is a short position in long-dated inflation — you repay in depreciated dollars while the asset reprices upward. When breakevens are low and fixed rates are locked, that position is cheap. It is, for most households, the single largest inflation hedge they will ever own, and it is acquired by doing nothing more exotic than not refinancing into a floating rate.

Stop treating cash as riskless.

Cash has zero nominal volatility and a guaranteed negative real return whenever the yield sits below inflation. Size your cash to a liquidity need measured in months, not to a feeling of safety (section “Sizing the Fund”).