The Macro Investing Tool

The Fed Can Hike Into the Oil Shock, but the Dollar and the Front End Still Decide the Damage

RAOUL'S TAKE · RAOUL PAL · 14 SEPTEMBER 2026 · THE COMPRESSION, MACRO INVESTING TOOLS

At some point, if this continues, gas prices are going to start to become an issue in the United States and will become an election issue. The national average is already $4.30 a gallon, up from $3.85 a month ago, and the midterms are on the 3rd of November. A resolution has to be found as fast as possible, and maybe China is offering a path for that.

Which brings me on to the China talks... Xi is in Washington on the 24th for the second summit of the year, and it's an important meeting, because there's a lot on the table. Xi offered back in May to help broker something on Iran and has since put a four-point peace proposal down, so the real question is whether the US is now prepared to let China act as the moderator, and if it is, then that wraps everything together into one China conversation, i.e. trade, where an extension of last October's truce is reported to be all but certain, the AI trade, with a dialogue on AI being set up ahead of the meeting, a weaker dollar, China buying US Treasuries again, access to the swap lines via the Middle East, Japan and the others, or at least access to liquidity, some sort of solution to Iran, and potentially even a solution to Russia and Ukraine. Not that we'll get all the details from the meeting itself, but that's how much is riding on it, and it remains unclear how far it gets.

Bond yields are caught in the trap here, where ongoing issuance is still too large for the Treasury's buyback programme to absorb. The 10-year is at 4.97%, its highest since 2023, and the 30-year touched 5.36% on Friday, a 19-year high, and that's after the bigger buybacks started on the 9th. The first of those operations bought $5.19bn against a $6bn maximum, and yields went straight back up. So we should expect some more changes on that front to try and manage the bond market, because the current size clearly isn't enough.

This week we also get the Fed, who are highly likely to hike rates, maybe 25 basis points, with the market pricing it at around 85% going into Wednesday. But hiking into an oil price shock doesn't solve it. Oil is at $103 and has been rising since July, and a hike just punishes people further who are already being punished by high mortgage rates, high interest rates and high oil prices. The 30-year mortgage is at 6.76% and rising, so the consumer is being squeezed from three directions at once, and none of that is something a rate hike fixes.

A lot of this is going to come down to our two measures of financial conditions, as to which is going to be the most likely path for markets. If it's the old measure, using the 3-month rate plus the dollar, the damage is relatively limited, because it has gone more or less sideways all year as the dollar has offset the front end. But using the newer measure, with 10-year rates plus the dollar, the damage is substantial, i.e. it has fallen from 56.9 in February to 49.3 in August, harder over six months than in all but around 3% of its history, and it will only get worse if rates stay this high.

Again, to reiterate, not all of the tightening legs are firing. Only two of the four are squeezing, the long end and oil, while the front end and the dollar are both still easing, so the breadth of the financial conditions tightening is still relatively benign.

When you look at the ISM against the new measure, financial conditions ex-oil, the chart still points at the ISM drifting back down towards 50 over the next nine months or so, which is the same picture as last week.

Against the old measure with 3-month rates, it suggests the business cycle goes sideways for a while instead, which is the version I'm erring towards. The ISM itself is at 54.6 with new orders at 53.7, so the survey has come off the boil a bit, but the services side is running at 61.7 and there's nothing in there yet that looks like real damage.

My hunch is that the dollar is the most important variable here, along with 3-month rates. The dollar is at 99.5 and has stayed under 100, which is what we wanted to see, and as long as that holds the old measure keeps the upper hand. As you can see on the chart of the Nasdaq versus the old FCI, this period of weakness should be over in the next couple of weeks or so, with the usual caveat that the lead shifts around and it's a tendency rather than a timetable. That intuitively makes more sense knowing where we are in the cycle overall. The Nasdaq is about 5% off its high and down 1.7% on the week, so it's still inside the same choppy range, and the dollar chart says much the same thing.

Other good news is that core CPI continues to drift lower. Friday's print had it at 2.4% year on year, from 2.8% in May and 3.0% a year ago, and the three-month annualised rate is running at 2.0%, so the trend is still down even though the monthly number came in a touch hot at 0.3%. Our forward indicators suggest it continues lower in due course, i.e. the breadth of inflation, the share of the CPI basket that's rising quickly, is down to 17% from 34% in May, and our acceleration measure has been falling for four months.

Supercore, on our measure of core CPI ex-shelter, is also incredibly benign and stuck at around 2%, i.e. 2.0% year on year and 1.9% annualised over three months. So the market clearly is not chasing inflation trends, and the 10-year breakeven at 2.3% says the same. What's driving the long end is the issuance and the lack of buying by China and others, and that's a function of the dollar, as I've talked about before. The weaker dollar starts to bring back the foreign central bank buyers of Treasuries, which is required for this whole system to work.

The rise in US liquidity on the broad measure suggests the Nasdaq rises over time, but not so sharply, i.e. more choppy into the midterm elections. Broad liquidity picked up to 4.3% year on year this week from 3.7% the week before, which was the biggest move on the dashboard this week.

The narrow measure suggests further strength for Bitcoin, which is at $78,400 and still in the uptrend on the Risk Monitor that started in August. The narrow level has been rising since the spring, but the excess version of it, the bit left over once the real economy has been served, is still scarce and got scarcer in August, so really we would expect the price action after October to be more interesting, once that excess starts to build.

Overall, I would expect to see continued choppiness and a slowdown in the rate of growth, as shown by the ISM, because of the tightening of financial conditions that's occurred. It's more likely that the ISM goes sideways for a period of time, as opposed to any real economic damage occurring. There's just too much tailwind from the supercycle of AI to change anything right now, but we always must stay alert for any changes in that picture.

My view remains that the supercycle ends up overpowering the liquidity and debt refinancing cycle, and it combines with a cycle that has periods of slowdowns but continues to grow all the way through, maybe even till 2030. Obviously, that's still a hypothesis on my behalf, and I will assess the data as it comes.

Into next week I'm watching the Fed on Wednesday and what Warsh says about the path from here, whether Treasury changes the buyback again and how the long end takes it, the dollar staying under 100, Xi in Washington on the 24th and anything on Iran, and whether the Nasdaq finally joins crypto as positive on the Risk Monitor.

Raoul