The Fed keeps fighting inflation with higher interest rates, but some of today’s biggest price pressures may have little to do with monetary policy.
KEY TAKEAWAYS
Headline and core PCE both came in better than expected. Headline PCE slowed to 3.4% year over year while core fell to 3.0%, giving markets an immediate reason to celebrate.
The measuring methodology changed at the same time. The BEA’s annual revision altered the treatment of portfolio management, legal services, and computer software and revised historical data going back to 2021.
Energy remains a major inflation driver. Gasoline and other energy goods are responsible for roughly two-thirds of the year-over-year increase in nondurable goods prices, reversing their previous disinflationary role.
Healthcare and housing remain stubborn service components. Housing is cooling gradually while healthcare is accelerating, leaving two large inflation categories that respond only imperfectly to Fed policy.
The Fed may be tightening against inflation it cannot directly reach. Higher rates cannot produce oil, alter healthcare reimbursement contracts, or quickly increase housing supply–and may actually restrain new home construction.
MY HOT TAKES
The inflation report was genuinely better, but the magnitude of the improvement deserves skepticism. Part of the cooling came from changing how certain prices are measured rather than from households suddenly paying materially less.
The composition of inflation matters more than the headline number. When energy, healthcare, and housing are doing much of the work, simply raising the Fed Funds Rate becomes an increasingly blunt instrument.
Energy may be the fastest path toward meaningful headline disinflation. Lower energy prices would hit gasoline directly while also relieving some utility and transportation costs throughout the economy.
More Fed tightening could create unintended consequences in housing. Expensive financing discourages construction precisely when additional supply could help reduce long-term housing inflation.
Investors should focus on what continued tightening does to rate-sensitive assets. If the Fed keeps attacking supply-driven inflation with demand-side medicine, homebuilders, small caps, utilities, and long-duration bonds become increasingly important areas to watch.
You can quote me: “A big chunk of the ‘cooling’ you read about yesterday came from a spreadsheet, not from the checkout line.”
Sweater or no sweater. It’s just a habit, but I find sticking to routines helpful with getting through my action-packed life. Yes, my ritualistic behavior can sometimes annoy my family, but thankfully, they get it. Every morning the espresso is pulled in the same way, same weights, same temperature, same beans, same roaster, same grind size. Before the crema even touches my lips I look out my window at the skyline–for practical reasons–for weather of course. Again, and I have told you this before, I know that I can simply look on my smartphone–rituals. As if that look at the street below wasn’t enough, there is the local news weather lady. She gives TWO numbers. The temperature–and the "feels like." I learned a long time ago to ignore the first one and dress for the second. Here is something I didn't know until I went down a rabbit hole at 4:45 AM this morning–after the espresso of course. For more than sixty years, the wind chill number came from a 1939 experiment by two Antarctic explorers, Paul Siple and Charles Passel (you don’t have to write down the names–there won’t be a quiz), who timed how long it took water to freeze outside their hut. Then, on November 1, 2001, the National Weather Service swapped in a brand-new formula–one built around how human skin actually loses heat–and practically overnight, winter got a little less brutal. Same air. Same wind. Same frozen ears waiting for the crosstown bus. Just a different formula. I thought about that yesterday at 8:30 in the morning, when the government told us inflation had cooled.
The Bureau of Economic Analysis reported that the PCE Price Index (Personal Consumption Expenditures–the Fed's preferred inflation gauge) rose 3.4% from a year ago in August, down from 3.7% in July. Core PCE (which strips out food and energy) came in at 3.0%, down from 3.3%. Economists were looking for 3.7% and 3.3%. Both numbers beat–both numbers BEAT–and the market exhaled. But read the fine print, because this was not an ordinary release. It was the BEA's annual update, and the revisions reach all the way back to January 2021. The entire post-pandemic inflation history was redrawn. Even July got recalibrated–its monthly readings were trimmed from 0.2% to 0.1%. So when you compare August to the July number you saw a month ago, you are comparing a reading on the new thermometer to a reading on the old one.
So what changed? The BEA reworked how it measures prices in three categories. The first is portfolio management and investment advice–yes, Wall Street. The second is legal services, where the old consumer price series had become so erratic it was essentially unusable. The third is computer software and accessories, where the old index was blending products with wildly different price trends into one number. The fix for legal services and software was to lean on producer price indexes (what businesses receive for what they sell) instead of relying on consumer prices alone. The fix for portfolio management was a different approach altogether–building the price from the industry's employment, hours, and earnings data. Is any of this a conspiracy? No. Statisticians fix their thermometers all the time, and most of these changes are probably improvements. Wall Street estimates had the changes shaving roughly 0.1 to 0.2 percentage points off core inflation. According to analysts, software alone accounts for the lion's share of that–the price of the programs on your laptop, not the price of anything you eat, drive, or live in. Think about that for a second. A big chunk of the "cooling" you read about yesterday came from a spreadsheet, not from the checkout line. But notice what was NOT on that list. Gasoline. Groceries. Rent. Your doctor. The new formula didn't warm the room–it changed the reading in the corner where the lawyers, the software, and the money managers sit.
Take a look at this chart, because it tells you what is actually holding the temperature up. Each bar is headline inflation broken into its pieces (my favorite Bloomberg ECAN chart). Of August's 3.4%, services (the rust-colored bars) account for about 2.3 points. Durable goods (the purple–cars, appliances, furniture) add less than 0.3. And nondurable goods (the yellow–things you use up, like gas and groceries) add about 0.8. Now look at how that yellow bar behaves. For most of 2024 and 2025 it was barely there–at times it was actually subtracting from inflation. Then, early in 2026, it came roaring back. The headline line (in white) peeled away from core (in blue) right along with it. That gap between headline and core is mostly energy–and most of that energy lives in the yellow bar. The methodology change never touched it. So what is inside the yellow bar? The following chart opens it up.
Nondurable goods prices are up about 4.1% from a year ago, and the green bars–gasoline and other energy goods–account for about 2.7 points of that. Two-thirds. Look at the history. Through 2024 and most of 2025, green sat below zero, pulling inflation DOWN. Starting in March of this year, it exploded to the top of the chart and has lived there ever since. That is the Strait of Hormuz showing up at your local gas station. By my own back-of-the-envelope math, gasoline and other energy goods alone account for roughly half a percentage point of headline inflation. And that is just the fuel you pump–the electricity and natural gas that heat your home are counted elsewhere, which I'll get to in a moment. Here is the uncomfortable part. No rate hike ever refined a barrel of oil. The Fed can make your mortgage more expensive. It cannot make a tanker sail faster. Now back to the rust-colored bar (services) from the first chart–the one that has refused to go away since COVID. Check out this next chart then keep reading.
Services prices are up 3.5% from a year ago, and two pieces do most of the heavy lifting. Housing and utilities contribute about 0.87 points. Healthcare contributes about 0.82. A year ago, that same comparison was 1.20 for housing and just 0.75 for healthcare. Housing is slowly–slowly–giving ground. Healthcare is quietly climbing. They are now running neck and neck, and at this pace healthcare is about to take the lead. Why doesn't healthcare respond to the Fed? Because a big share of what PCE counts as healthcare isn't what you pay at the counter. It's what Medicare, Medicaid, and your employer's insurance company pay on your behalf–prices set by reimbursement schedules and contracts negotiated long before anyone at the Fed raised a hand to vote. Interest rates do not reach those prices. And don't miss the word utilities in that housing bar–electricity and natural gas are hiding in there, too. Energy shows up twice!
Put the three charts together and here is what you get. Energy goods, healthcare, and housing–each worth roughly half a percentage point of headline inflation. Three blocks, about the same size, and together roughly half of all the inflation in the report. Not one of them answers directly to the Fed Funds Rates.
Which brings me to the Fed. In September, the FOMC raised rates by a quarter point to a range of 4.00%–its first hike since 2023–and its own projections pencil in one more before year-end. Just yesterday, Minneapolis Fed President Neel Kashkari said it plainly: "inflation is still too high." He sees one more hike this year and another in 2027–though he was careful to call the projections a snapshot in time. I understand the instinct. Inflation at 3.4% is not 2%. It has been above target for more than five years now, and a central bank that blinks too often loses the one thing it can't print–credibility. But look at the housing piece again, because this is where the medicine can make the patient sicker. Higher rates push mortgage rates higher–though this is not directly tied to Fed Funds, just indirectly. People sitting on cheap old mortgages stay put. More directly, however, builders face more expensive financing and build less. Fewer homes for sale and fewer homes being built means rents and housing costs stay sticky–STICKY–longer. The one big component that was actually cooling on its own is the one that rate hikes work against.
So what does fix it? The clearest lever is energy–get it back down, and the green bars shrink, the yellow bar shrinks, and the utilities piece of housing gets relief, too. That lever doesn't sit at the Fed. It sits with the executive branch–diplomacy around the Strait of Hormuz, supply, how the Strategic Petroleum Reserve gets used, how we treat our own refining capacity. Those are policy questions, not political ones, and the answers will matter more to your inflation rate over the next six months than anything said at the next Fed press conference. For investors, the question is just as practical. If the Fed keeps tightening into inflation it can't reach, what happens to rate-sensitive corners of your portfolio–homebuilders, utilities, small caps, long-dated bonds? And are you positioned for a Fed that is fighting the wrong bars on the chart?
Tomorrow morning, between espressos, the weather guy will give me two numbers again. I'll still dress for the "feels like." Because here is what the 2001 wind chill change taught us–a new formula can make the reading friendlier, but it never changed how cold your ears get at the bus stop. Yesterday's PCE report is the same story. The formula changed. The wind didn't. And the Fed is reaching for the thermostat in a room where the furnace, the windows, and the front door all belong to somebody else.
YESTERDAY’S MARKETS
Stocks ended the third quarter mostly lower yesterday. The S&P 500 slipped 0.25% and the Dow Jones Industrial Average fell 443 points, or 0.86%, to 50,906, while the Nasdaq Composite rose 0.24%. The S&P 500 and the Dow both finished September with monthly losses as the 10-year Treasury yield climbed to about 5.29%, a 52-week high. West Texas Intermediate crude rose roughly 0.9% to about $90.17 a barrel.
NEXT UP
Initial Jobless Claims (September 26th) came in at 197k slightly lower than last week’s 198k.
ISM Manufacturing (Sept). May have inched up to 55.0 from 54.6.
Construction Spending (August) was probably flat after slipping by -0.5% in the prior period.
Fed speaking bonanza today: Kashkari, Barkin, Collins, Schmid, Waller, Jefferson, Boman, Cook, and Logan.