Fed hike odds fell after weak payrolls on 7 August and rose again on oil three sessions later. The two moves are not the same story: one is a demand signal, the other a supply shock. This explains why that distinction is the whole problem, what the Strategic Petroleum Reserve at 298.7 million barrels can and cannot do, and why Wednesday's July CPI print cannot contain the shock markets are pricing today.
On 7 August, a weak payrolls print pushed the odds of a September Federal Reserve rate hike down sharply. Three sessions later, the same probability series was climbing again — and this time the input was oil. Markets priced roughly 51% odds of a 25bp September hike as of 11 August, up from around 44% the day before, according to Tickmill's daily market outlook. The series has round-tripped about half its post-payrolls move in three days.
It is tempting to read that as noise, or as the market changing its mind. It is neither. The two moves came from genuinely different inputs travelling through genuinely different channels, and the second one poses a problem the first did not. A labour market softening is a demand signal, and demand signals are what interest rates are built to act on. An energy supply shock is not. It is the one shape of inflation that monetary policy handles worst, and it has arrived while the buffer designed to absorb it sits at a 43-year low.
The mechanics of the reversal are straightforward. Crude moved hard on 10 August: WTI rose 3.49% to $80.91 and Brent gained more than 3% to $86.75, per TradingKey. By the morning of 11 August, WTI futures opened at $82.34 and Brent at $87.80, with Brent trading up to $89.24 later in the session, a 1.73% gain on the day, according to Trading Economics and Forbes Advisor's daily quote.
Two drivers are doing the work. Iran has attached conditions to reopening the Strait of Hormuz — lifting the naval blockade, withdrawal of forces, sanctions relief, unfreezing of assets, and compensation for war damage — and a drone strike hit a Saudi Aramco refinery. Neither of those is a demand story. Both are constraints on physical supply.
Rates followed. The US 10-year Treasury yield rose to 4.73% on 11 August, up 2bp on the session and near its highest level since January, after Monday's selloff had already taken 10s up 6bp to 4.71%. Cleveland Fed President Beth Hammack has said multiple rate increases could be necessary to return inflation to 2% — a single official's stated view, not a Committee position, but one the market is clearly weighing.
A note on the odds themselves: this series moved every day last week, from roughly 67% before payrolls to 44% after, to 51% on oil. Any hike-probability figure is only meaningful with a date attached to it. Treat single readings as a snapshot of positioning, not a forecast.
Here is the part almost no same-day coverage will frame properly. Interest rates fight inflation through one primary channel: they suppress demand. Higher borrowing costs make households defer purchases and firms defer investment, aggregate demand falls, and prices that were rising because too much money was chasing too few goods stop rising as fast. That machinery works, and it is well understood.
It does nothing whatsoever to the supply side of an energy shock. A rate rise does not reopen the Strait of Hormuz. It does not rebuild a struck refinery. It does not add a barrel to global production. What it can do is destroy enough demand that the shortage clears at a lower price — which is to say, it fights an oil shortage by making the economy smaller.
That is a genuinely bad trade when the economy in question has just printed −23,000 payrolls. The Fed's actual choice this month is not between hawkish and dovish. It is between tolerating a headline-inflation overshoot it did not cause and cannot fix, or tightening into a labour market that is already weakening. Both options are bad, and the case for each is coherent.
The hawkish argument is about expectations rather than mechanics: even if a rate rise cannot produce oil, letting headline inflation run hot risks unanchoring the expectations that make the whole framework work, and once that happens the eventual cost of re-anchoring is much higher. That argument has history behind it. The dovish argument is that supply shocks are, by construction, transitory in a way demand-driven inflation is not — the price effect passes through the index and then drops out of the year-over-year comparison — and that tightening into it converts a temporary price problem into a durable employment one. That argument also has history behind it. Reasonable people at the same central bank hold both.
The US Strategic Petroleum Reserve fell by 6.1 million barrels last week to 298.7 million barrels, dropping below the 300 million mark for the first time since 1983, according to Department of Energy data reported by CNBC.
The headline invites a tidy irony — the shock absorber is empty just as the shock arrives — but the actual sequence is more interesting and less coincidental. The drawdown traces to a 172-million-barrel release authorised in March, after Iran moved to restrict shipments through the Strait of Hormuz. The reserve stood at roughly 415 million barrels before the US and Israeli strikes on Iran on 28 February. In other words, the buffer is not low because of unrelated mismanagement colliding with bad luck. It is low because it has already been spent fighting an earlier phase of this same crisis. When the release completes, government stocks are expected to fall to around 243 million barrels.
It is worth being precise about what that does and does not mean. The SPR is not close to unusable: an Energy Department spokesperson told CNBC that the minimum needed to safely operate the reserve is about 70 million barrels, so 298.7 million is well above the operational floor. The United States is not about to run out of oil, and the reserve is a small share of what the country consumes and produces.
What has changed is optionality. The SPR is a physical buffer with a finite drawdown rate, not a price control, and its historical record on price is modest: releases have tended to produce temporary relief measured in a few dollars a barrel rather than a durable reset. That effect is smaller still when the reserve has already been drawn down once and the market knows the remaining capacity to repeat the move is reduced. The credible threat of a release is part of what makes a release work, and that threat is worth less now than it was in February.
July CPI is scheduled for release at 8:30 a.m. Eastern on 12 August, with PPI following Thursday. A Reuters poll put the consensus at +0.1% month-over-month, up from −0.4% in June, and 3.4% year-over-year, down from 3.5%. Kiplinger's preview put core CPI at roughly +0.32% m/m and 2.5% y/y.
Those are forecasts, and this piece does not report the result. The more useful point for a reader is structural: July CPI covers July. It cannot contain an oil move that happened on 10 and 11 August. Whatever the print says about energy prices, it is describing a period before the shock markets are currently pricing.
That is why the reaction may look strange. A soft July print does not clear the inflation risk the market is worried about, and a hot one is not evidence of the shock in question. The information that matters for the September decision has a two-to-four-week lag built in — it will show up in the August CPI released in September, which is uncomfortably close to the meeting itself.
In the meantime, the honest place to look is not the CPI headline but the things that move faster: retail gasoline prices, which pass through crude with a lag of roughly two to four weeks; breakeven inflation rates in the TIPS market, which show what investors actually expect rather than what already happened; and the shape of the yield curve, which separates a market pricing tighter policy from one pricing a policy mistake. If long yields rise with short yields, the market is pricing inflation. If short yields rise and long yields fall, it is pricing a hike that will have to be undone.
None of this resolves which of last week's two moves was the signal. It probably was not either. The payrolls print and the oil move are both real, they point in opposite directions, and the Fed has to weigh them against each other with incomplete information — which is the ordinary condition of monetary policy and not a scandal. What is genuinely unusual is the combination: a weakening labour market, an externally imposed price shock, and a depleted buffer, all at once. That is a harder problem than a probability series can express.
They can't fix the supply, but they can act on the consequences. The hawkish case is about inflation expectations: if headline inflation runs hot for long enough, households and firms start building it into wage demands and pricing decisions, and at that point it stops being a one-off energy shock and becomes self-sustaining. A rate rise is an attempt to stop that from happening. The cost is that the only way it works is by suppressing demand — meaning it fights an oil shortage by shrinking the economy, which is a hard trade when payrolls just came in negative.
DrafterDaily publishes daily analysis on the stories where the mechanism matters more than the headline — finance, AI, crypto and technology, without the hype.
Read more analysisGrad PLUS lent up to the full cost of attendance. Its replacement caps professional students at $50,000 a year — against a medical-school year that averages $61,110. The difference does not disappear.
Fitch tracks sixteen perpetual non-traded BDCs. In Q2 2026, redemption requests averaged 10.3% of shares against a standard 5% cap, and ten of the sixteen breached it. Nothing broke — which is the uncomfortable part.
Nationally, asking prices are down about 2% and inventory is up 3.2% — a picture of orderly rebalancing. Underneath it, Austin's median list price is down 12.2% while San Francisco's inventory has fallen nearly 20%. Those are not variations around a mean.