Why Bond Yields Matter More Than Fed Decisions

Here is the framing we use, and it is the sentence this piece exists to install: the Fed decides the overnight rate; the bond market decides the discount rate. A policy decision matters precisely to the extent that it moves term yields, and term yields spend most of the year moving for reasons that have nothing to do with the meeting calendar — growth data, inflation data, fiscal supply, the global demand for duration. An investor who watches decisions is watching an input. An investor who watches the curve is watching the output, and the output is the only thing that touches asset prices.

One administered price, thousands of financed decisions

Start with what actually gets financed and at what maturity. A household buying a house borrows for decades, at a rate built off long-term Treasury yields plus a mortgage spread. A company funding capacity issues bonds at intermediate and long maturities, at the matching Treasury yield plus a credit spread. An equity investor valuing a business discounts cash flows that stretch out ten, twenty, thirty years — against a long risk-free yield plus an equity risk premium. In each case, the reference price is a term yield. The overnight rate appears nowhere in the arithmetic except as one distant input into the market's estimate of where overnight rates will average over the life of the loan.

That last clause is the mechanism in miniature. A ten-year yield is, to a first approximation, the market's expected average of overnight rates over the next decade, plus a term premium for bearing duration risk. (The decomposition of a yield move into those two components is its own framework, and we treat it separately — the point here is prior to it.) A single meeting contributes one small step to a decade-long average. The information that moves a ten-year yield is information about the whole path — how fast the economy is growing, how persistent inflation looks, where the destination rate sits — and that information arrives continuously, through data, not quarterly, through decisions. This is why term yields so often move more violently on an employment or inflation release than on the decision the release will eventually inform: the release revises the entire expected path at once, while the decision usually confirms a step the market had already priced.

Three layers of why

This is also the discipline behind how Sterling's Embedded Intelligence processes a policy event: the decision itself is logged as one input, but the object that gets analyzed is the curve's response — which maturities moved, in which direction, and whether the move was carried by the expected path or by the premium for holding duration. Sterling treats the announcement as a hypothesis and the curve as the verdict, because the verdict is what a portfolio is marked against.

When the curve refuses to follow

The strongest case against this framework

The serious objection runs: the Fed controls the curve too. Forward guidance moves the expected path directly, and large-scale asset purchases showed that a central bank can compress term premia at will. If the Fed can move any point on the curve it chooses, then the decision — broadly defined to include guidance and balance-sheet policy — is still the thing to watch, and yields are just the instrument panel.

We accept the premise and note that it concedes the argument. Guidance and asset purchases exist precisely because the overnight rate alone does not transmit — they are the Fed's own acknowledgment that term yields are the variable that matters. But conceding that the Fed can influence the curve is very different from claiming the Fed determines it. The curve aggregates the Fed's intentions with global savings flows, fiscal issuance, inflation risk, and every investor's estimate of the destination rate. The Fed is the largest single voice in that auction, not the auctioneer's hammer. When its guidance is credible and the data cooperate, the curve follows; when either fails, the curve overrules — and it is the curve, not the guidance, that the economy borrows at.

There is a second, narrower objection that deserves its own answer: the overnight rate matters mechanically. Floating-rate borrowers reprice off short rates within months. Money-market yields on trillions of dollars of cash reset with the policy rate, and bank deposit and lending economics key off the front end. All true — and the right way to hold both facts is a distinction this research library uses across clusters: the policy rate is a cash-flow variable for short-duration balance sheets; the curve is the valuation variable for everything else. If you own leveraged loans, hold large cash balances, or underwrite bank earnings, the front end is your instrument. If you own equities, real estate, or anything whose value lives in distant cash flows, your discount rate is set at term, and the meeting calendar is at best an occasional supplier of news to it.

What this changes about how you watch policy

The practical output is a reordering of attention. The two-year yield is the market's running summary of the policy path over the horizon the Fed can actually reach — it is a better policy monitor than any statement, because it updates on every data point rather than eight times a year. The ten-year yield, and specifically its inflation-adjusted component, is the input that belongs in an equity valuation. And the response of the curve to a decision is the only honest grade of that decision: a cut that steepens the long end because the market smells inflation has delivered no easing to a single mortgage borrower, whatever the press conference said.