42bp at Year-End: The Floor or Ceiling of a Fed Hike Cycle, and How to Position Duration

Next week the Federal Reserve will probably raise rates for the first time in 2026. A hot August CPI report, released on September 11th, pushed the probability of a quarter-point hike at the September 15–16 meeting to roughly 90% on CME FedWatch, while the futures market overall prices about 42 basis points of additional tightening by year-end. That figure—about one and three-quarters quarter-point moves—is the whole debate in miniature. Is it the floor beneath a longer hiking cycle, or a ceiling that a one-and-done tightening will soon unwind?

A number that prices a hunch

It is tempting to read the answer from the inflation data. The trouble is that 42bp is not really a forecast about prices. It is a measurement of how much the market trusts a central bank that has gone out of its way to make its intentions unknowable.

Kevin Warsh, who became Fed chair in May, communicates the way a poker player bluffs—by saying less. He declined to place a dot on the June dot plot, cut the committee's statement to roughly 130 words, and replaced the pledge to support maximum employment with the declaration that the committee "will deliver price stability." He has questioned whether PCE inflation should remain the Fed's target and put the future of the dot plot itself before a task force. J.P. Morgan's economists note that Warsh has stridently asserted resolve while declining to specify how he will achieve it.

This opacity matters because a credibility-committed path is what turns market pricing into a floor. When a Fed chair stands behind a projected path, traders treat that path as a promise to be tested against data. When the chair refuses to say where he is going, they must guess at resolve from rhetoric—and a guess that swings with every release is exactly what 42bp is. As late as August the market put only about a one-in-three chance on a September hike, below even the single hike the Fed's own June projections had penciled in. Within weeks, after the hot print, that odds had roughly tripled. That swing is not the market anticipating a cycle; it is the market discounting the absence of guidance.

Two readings, one shock

The case that 42bp is a floor rests on inflation that is no longer merely an energy story. Core CPI rose 0.3% month-on-month in August, a tenth above consensus, and 2.4% year-on-year. The gains are broad in character: shelter rose 0.3%, ending a stretch of moderation; transportation services rose 0.5%; and an AI-driven squeeze on computer and electronics prices has, by TD Economics' estimate, added 0.7–0.8 percentage points to core inflation this year. Services breadth has narrowed but remains "elevated and sticky." Against that, a robust labour market—162,000 jobs added in August, unemployment at 4.1%—gives a hawkish Fed every excuse to continue. Three members voted for a hike in July, the most dissents in a decade.

Yet the honest reading is that the current print is still largely an energy shock wearing a core disguise. Gasoline rose 3.9% in August and accounted for more than a third of the headline's monthly gain; energy is up 16.3% year-on-year as oil trades above $100 a barrel on the US-Iran conflict. The pass-through of that shock into services has so far been minimal outside airfares, which have risen 7–11% since February; businesses appear to be treating the spike as temporary. If oil retreats—or merely stops rising—the headline will cool, and the core's 0.3% will look like noise against a 2.4% annual rate already down sharply from a 4.2% peak in May. On this reading the Fed delivers one credibility-preserving hike and stops, and the second, December leg embedded in the 42bp is precisely the part that unwinds.

The dots settle it

The September dot plot, published with the decision, is the instrument that will settle the argument. The June median penciled in a single hike by year-end, a target range of 3.75–4.00%, with nine of eighteen participants at or below the current range. The market now prices more than that—about 1.7 hikes, or 42bp. So the fresh dots arrive with a clear test. If the median ratchets up toward two hikes, a Fed majority will be telling the market the cycle runs longer and 42bp is indeed a floor. If the median holds at one hike while Warsh again declines to place his own dot, the central bank's own central tendency is telling the market it has priced one move too many. That is the cleanest evidence available that 42bp is a ceiling.

The core CPI prints that follow are the second lever. A sustained run of 0.3%-and-up core readings, or a re-accelerating supercore, validates the floor: inflation has become domestic and self-sustaining. A return to the 0.2% monthly pace that most forecasters still expect, with the headline held up only by petrol, validates the ceiling: hike once, then watch.

The falsification conditions follow directly. The claim that 42bp anchors the path breaks—in either direction—when one of a handful of things happens: the dots move decisively up, or hold while Warsh finally commits; core CPI accelerates with breadth, or snaps back to 0.2%; a Middle East de-escalation pulls oil back toward the $70s and removes the pass-through pretext; or the labour market cracks, denying the Fed cover for a second hike. Each is observable within two months.

Duration, dollar, and the credibility test

For Treasury investors the two readings have opposite consequences, which is why the positioning question is really a bet on the dots. Duration is a bond's sensitivity to rate changes—roughly, how much its price falls when yields rise. In the floor world of more hikes and a higher terminal rate, the whole curve reprices higher, the risk is higher-for-longer, and the trade is short duration: money-market funds and the front end, banks the 42bp at a discount. In the ceiling world of one-and-done, the front end falls as the December leg is priced out and the curve bull-steepens, rewarding extended duration in long Treasuries. Note that much of either move has already happened. The 10-year yield sits near 4.97%, its highest since October 2023 and up roughly 0.9 percentage points year-on-year, with the dollar near 99. The easy repricing is done; what remains is whether the dots confirm it or reverse it.

The dollar follows the same logic through interest-rate differentials: a lengthening cycle firms the currency, an unwound one-and-done lets it ease as the differential narrows.

And when is the market's 42bp a reliable forecast? Only when the two things that make futures pricing trustworthy are in place: a committed institution and a mean-reverting shock. The Fed currently supplies neither—guidance is withdrawn, and the shock is geopolitical and binary. Treat 42bp for what it is: a crowd's price for information the chair refuses to supply. It will prove accurate only if the Fed uses the September dots to bind itself to the path its rhetoric promises, and if core inflation keeps validating that promise one print at a time. If the dots ratchet to two hikes, duration stays short and the dollar firms. If the chair again declines to put himself on the page, the market is left holding a hike the Fed has not sanctioned—and the December leg of the 42bp is the first thing to go.