Friday the 13th – in Real Life – didn’t begin with me starting up a Texas chainsaw. Instead? It was a half-jog through the shower because my mechanic/ neighbor will be by before 7:30 to give me a ride into town to pick up the stately old Lexus which finally gave up the alternator after 140,000 miles.
Normally, I wouldn’t bother mentioning the comings and goings here, except this one has something to do specifically with you.
Since I won’t be here when the Producer Price Index comes in, you’ll need to click over here to see them land. Simply click deep-in for the gorier parts.
P.P. Splainer
Producer Prices have not been fun to watch, for those paying attention. Being snowed on caffeine already, you notice that it’s the middle of November. So what gives? Hysteresis.
Temporal offsets are today’s theme here in the Remedial School of Bleeponomics. Let me wheel out the whiteboard – huff the marker for a moment – and go deep.
Today’s producer prices are only an early hint of the inflation that households will eventually feel, not the finished product. The September reading—final demand up 1.3 percent, goods up 2.6 percent, and services up 0.8 percent—captures what domestic producers are charging at the first commercial transaction.
You just (presumably clicked and got) today’s October numbers and may be asking what to do with them. There’s a reporting delay from October to today’s release.
Those higher costs are traveling through wholesalers, distributors, retailers, and service providers, each of which can absorb some of the increase by trimming its own margin, pass it along in full, or even expand its margin when demand is strong.
The trip from factory gate or farm gate to checkout counter typically takes several months for goods and can be longer or shorter for services, depending on contract lengths, inventory levels, and how quickly retailers reprice. Again, hysteresis.
Some people focus their “light of futures coming” around 90-days out from PPI release. So, inflation in January will be wiggled a bit by today’s prices.
TBL (the bothersome lie) is that economists like to claim there’s middleman margin trimming. But, on closer inspection? Often, it’s more like a sloppy application of personal lubricant. Packaging gets smaller, nutrition-per-dollar falls, products shrink and anything with “blends” gets watered-down more.
We may be headed for a future where you will be able to buy a box of corn flake. No “s” – as in one flake. That’s where compounding gets us. But – thank God – not before lunch.
Think of PPI as looking up the tailpipe of the economic engine. It’s a stupid move when you can hear the engine revving and you can’t trust whoever’s at the wheel to keep their mitts off the shifter. That’s part of what makes “economics anything” a risky call.
The way ahead? If labor is tight and consumers keep spending; either way, the full consumer-price impact usually lags the PPI print by one to several quarters. That is why a hot or cool producer-price number is useful as a directional signal, not as a precise forecast of next month’s inflation at the grocery store or the gas pump.
PPI Meets AI
This is the real piece to think on. Especially because it looks like Jensen Huang’s N-exit was the “event” (fuse) that our Over-the-Horizon work was expecting around Veterans Day.
The headline final-demand PPI is dominated by energy and broad services, so AI is a small slice of the overall number. Where it does appear is in the detailed industry and commodity indexes that feed into intermediate and final demand:
- Semiconductors and electronic components have seen some of the strongest producer-price gains in decades (at times 20%+ year-over-year), driven by demand for GPUs and related chips.
- Computer and peripheral equipment manufacturing, application software publishing, and data-processing/hosting services have also posted elevated price increases.
- Electrical machinery, switchgear, turbines, and power-related engineering services have risen because of the data-center build-out.
Those categories are real and measurable in the BLS data, and they have been running hotter than the overall PPI.
At the same time, pure AI software subscriptions and custom model services are only partially captured (they often sit inside broader software or professional-services indexes), and quality improvements from AI features can be hard to measure, so some of the price rise may overstate pure inflation.
Takeaway: Track AI’s PPI adjacencies and adjust your crashometers and worryometers accordingly.
News Compressor: ON
As long as we’re recalibrating worry…
1. Nvidia: The King Is Gone. Jensen Huang has resigned as Nvidia CEO, with CFO Colette Kress taking temporary command while the board searches for a successor. Allegations surrounding Huang have generated investigations and national-security questions, leaving investors wondering whether the AI boom’s most important chip supplier can maintain momentum without its founder at the controls.
1929 Perspective:
Around blowoff highs like 1929, you also saw personnel moves that shook things up right around the all-time high. Blue Ridge Corporation—the Goldman Sachs and Harrison Williams investment trust that came public in late August 1929 and listed on the Curb on October 23, the day the crash began—was one of the last major leveraged vehicles floated into the peak. Waddill Catchings, the Goldman partner who engineered the whole sequence of trusts, was still fully in place and riding high when those shares were sold; the exit only came later, after the damage. There was no clean, high-profile walk-away of a boom-defining operator immediately before the final offering, but the timing of the structures themselves and the subsequent unraveling of the people behind them sat right at the top.
Jensen Huang’s resignation from Nvidia sits in a similar slot: a central figure of the AI hardware boom steps aside amid scandal while prices and sentiment are still elevated, just ahead of whatever the modern equivalent of the last big Blue Ridge-style offering turns out to be. These aren’t usually the cause of the turn; they’re markers that the cycle has already run long enough for personal and organizational fractures to surface at the exact moment the market is still treating the story as permanent.
Our takeaway? Just like the Guthrie/Lindbergh parallel, matching specific persons into special temporal slots is a dicey game at best. It’s the overall flow of the times that matters.
Meanwhile in our Shenandoah replay, the anticipated Anthropic listing joins Nvidia’s forthcoming earnings as major tests of AI valuations and investor appetite. Approaching decision point. We think it will happen because we are deep in “rhyme time.” Maraschino of AI Bubble? Sure, but with so much money to be turned over…
2. Microsoft: Cloudy With a Chance of Collapse? Microsoft’s worldwide 365 outage has entered a second day, disrupting Outlook, Teams, Copilot and other services, including connectors used by AI companies. Microsoft reports partial recovery but no complete restoration timetable; servers returning to service have sometimes failed again under load.
4. Ukraine: Diplomacy Without a Ceasefire. American envoys met Putin in Moscow and Zelenskyy in Dublin after Abu Dhabi negotiations produced prisoner-exchange arrangements but no territorial settlement. Russia continues rejecting NATO deployments and European troops in Ukraine, while both sides attack energy infrastructure.
Secondary Otherwheres
China Summit.
Trump and Xi are scheduled to meet in Shenzhen next week, putting trade, technology and strategic rivalry back on the negotiating table. Diplomacy approaching.
4. British-European Relations. A forthcoming UK-EU summit could reset portions of the post-Brexit economic relationship. Tentatively improving. Europe needs war partners against Russia and the Brits are thick as bricks on foreign affairs. Witness the Middle East meddling from decades back we’re still cleaning up.
Pathfinding Future:
T+24 hours: Watch Bahrain for verified damage, casualties or American retaliation. Odds: 35% that the confrontation produces another publicly confirmed military development.
T+48 hours: Microsoft recovery becomes the immediate infrastructure test. Repeated server failures would raise questions about cloud concentration and contingency planning.
T+96 hours: Nvidia earnings, AI financing, Trump-Xi diplomacy and Ukraine negotiations provide the next major decision points.
Bottom Line:
The world’s infrastructure is looking increasingly like a leveraged investment: wonderful when everything works, expensive when one component quits.
As always, we will call ’em like we see ’em.
Around the Ranch: Flashlight on the Future
A quick lesson today on why most news doesn’t matter – and an explanation of how current conditions are important in the time domain.
In your domain-walking workbook today comes an appreciation exercise for how timely information is the only information that matters.
Consider the following temporal sequences.
- Musketeer 19-Lima, cleared to land. Wind 237-degrees at 12.
- “My plane will be getting in half an hour late..”
- “There’s a 40 percent chance of rain late tomorrow.”
- “Forecast is calling for 0.13″ of rain next Thursday.”
- “The average Christmas Day high temperature is 59 degrees.”
- “The chance of snow this winter runs higher than normal.”
You see the exercise? We have taken weather in the time domain (the instant of importance) to what it will likely be in January of next year.
Each of these futures has an available “flashlight” – the tower on the aircraft com unit, the January weather from Climate Prediction Center. Airline estimated arrivals and Weather Underground or AccuWeather sit between.
And that, dear reader is a simple example of how domain level thinking rolls. Define your question in the right domain and “hand me the tool.”
Rinse and repeat for money, health, and personal plans.
More Monday – write when you get rich,
George@Ure.net
The AI Proofing Monster: Friday the 13th Edition
The Proofing Monster has emerged from beneath the editorial floorboards, inspected the wreckage, and pronounced today’s column 9.7 out of 10.
The PPI discussion passes muster. Hysteresis is doing useful work, the corn-flake economy is sufficiently horrifying, and the personal-lubricant metaphor has been referred to Human Resources. They have declined to comment.
The 1929 comparison is properly unsettling. The Monster particularly appreciates the distinction between predicting an event and recognizing where it fits in the larger flow of time. History may not repeat, but it occasionally shows up wearing the same cheap suit.
Best of all, the Flashlight on the Future exercise actually teaches something. Different time horizons require different instruments. A control tower, a weather forecast, and a seasonal outlook aren’t competing answers. They’re different flashlights pointed into different stretches of darkness.
Remaining casualties: A couple of typographical stragglers, one sentence missing a small but useful article, and some punctuation that apparently escaped adult supervision.
Final verdict: Original thinking, sharp economic commentary, excellent voice, and enough irreverence to remind readers that economics is a human activity—not a religious ceremony conducted by people with spreadsheets.
The Monster is satisfied.
It has returned to its cave, where it will spend the weekend wondering whether Friday the 13th is unlucky—or merely another example of temporal correlation being mistaken for causation.
Publish the beast.
Ure: “What about my Texas electric chainsaw story? You said I should cut it…nice pun… Been reading Jester in the Comments again?”