Cover of Brief 12, Capital Now Charges For Proof

September 30, 2026

Volume 1 · Capstone

The Economy Held.
Business Got Harder.

Volume 1 in review: how twelve weeks changed the cost of acting on steady demand, which calls held, and the signals that will test the fourth quarter.

Curated byPJ Bickett signature
01

The Quarter

Volume 1 ran from a closed strait in July to a unanimous rate increase in September.

The volume opened with executives still planning on cheaper money and cheaper energy for the second half. Within three weeks both assumptions were under strain: Brent reached $100.69, the 10-year Treasury moved above 4.7%, and Alphabet disclosed a $49.6 billion equity raise to keep building.

August showed the economy could carry those costs, and that the price of money would be what moved. Second-quarter revenue across the S&P 500 grew 15.0%, the fastest since late 2021, even with Hormuz moving a fraction of its prewar flow. Meanwhile the 30-year Treasury reached 5.31% on August 17, its highest since 2007, with the Federal Reserve still on hold.

September turned the quarter's open question into a fact. The Federal Reserve raised its range to 3.75% to 4.00% on September 16 by a vote of 12 to 0,1 and the 30-year closed September 25 at 5.49%, its highest close since January 2002.2 By then the companies that moved had already acted on their own evidence: Campbell's, Uber and The Trade Desk cut staff, and GE and Amazon moved to buy or bind suppliers they could not quickly replace.

Demand held for twelve weeks. The price of acting on it rose, in energy first, then in money, and by September in the suppliers and lenders behind each commitment.

02

The Decision SignalBusiness Friction Index

The quarter in one view

Four conditions, each scored 1 to 5 for every brief, measuring how hard it was to turn steady demand into results. Each week is scored on what that brief reported.

1 · Low2 · Manageable3 · Elevated4 · High5 · Severe

Capital carried the quarter.

ConditionFirst sixLast sixChange
Demand2.22.0-0.2
Capital3.24.3+1.1
Cost3.33.3+0.0
Execution3.23.6+0.4
Overall3.03.3+0.3

Capital friction rose from 3.2 across the first six briefs to 4.3 across the last six, the largest move of the four. Execution rose more slowly, from 3.2 to 3.6, as the constraint moved from shipping routes to suppliers. Cost held at 3.3 in both halves: energy stayed expensive but did not trend.

Demand is why the quarter never read as a downturn. It averaged 2.1, the lowest of the four, and ended the volume at 1.5. Customers kept buying through a closed strait and a rising long bond, and the friction sat in funding, pricing and delivering what they bought.

Index methodology and full record

The Business Friction Index measures four conditions that determine how difficult it is for companies to turn opportunity into results.

Demand
Can we sell?
Capital
Can we fund?
Cost
Can we protect margin?
Execution
Can we deliver?
Overall
The simple average of the four, weighted equally.
1 · Low
The condition helps the plan.
2 · Manageable
Present, and absorbed inside the normal plan.
3 · Elevated
Requires the plan to change its assumptions or its timing.
4 · High
Changes which commitments clear.
5 · Severe
Forces decisions the plan did not anticipate.

Each weekly score is an editorial judgment made against what that brief reported. Volume 1 scores are fixed as of this Capstone, and only a factual or arithmetic error changes them.

Volume 1 weekly friction scores
BriefDemandCapitalCostExecutionOverall
01 · Jul 61.53.03.52.52.6
02 · Jul 132.03.53.53.03.0
03 · Jul 201.52.02.53.52.4
04 · Jul 272.03.53.53.03.0
05 · Aug 32.53.54.03.53.4
06 · Aug 103.53.53.03.53.4
07 · Aug 171.54.03.02.52.8
08 · Aug 243.04.54.03.03.6
09 · Aug 312.04.53.04.03.4
10 · Sep 72.04.03.03.53.1
11 · Sep 142.04.04.04.53.6
12 · Sep 211.54.53.04.03.3
Average2.13.73.33.43.1
03

The Defining Shift

AI moved from the technology budget to the capital budget.

The Index shows capital carrying the quarter, and much of the reason traces to that change in how the largest companies pay for AI. The buildout outgrew technology budgets and operating cash and moved onto the capital budget, increasingly financed with debt: AI-related borrowing reached $220 billion this year against $12.5 billion a year earlier.

That move reached companies that never bought AI. Brief 08 found issuance funding AI and data centers among the forces lifting the 30-year Treasury, which sets the price of long-duration money for every borrower. Brief 05 found Apple raising prices because AI data centers were bidding away memory chips, and the Federal Reserve's own minutes named the buildout as a force keeping chip and electricity prices high.

AI now competes with sales, people, acquisitions, infrastructure and margin for the same dollar, whether the program costs $100 billion or $1 million. Every executive faces the same allocation question: what else could this dollar fund? Capability earns AI a place in that conversation, and return decides how much of the dollar it gets.

04

The Record

Twelve briefs made thirty-six calls.

Three show the range of the volume's judgment: one the quarter confirmed, one still open and one that came too early. A week-by-week view of the posture follows, including the one call that overreached.

On the money

Cheaper Money Never Arrived

↗ Brief 01, Signal 02: Betting Against the Fed↗ Brief 05, Signal 02: Three Votes For Higher↗ Brief 08, Signal 01: 164 Points Over Policy

Brief 01 warned that second-half plans rested on rate cuts the Federal Reserve had stopped signaling. Brief 05 recorded three dissents for a higher rate, and Brief 08 showed the long bond already pricing long-duration money 164 basis points above policy.

The Fed raised, unanimously.

The committee that split 9 to 3 in July voted 12 to 0 for an increase on September 16,1 and sixteen of eighteen policymakers expect at least one more by year-end. Plans that budgeted for a cut ended the quarter paying more at both ends of the curve than they did on July 1.2

Too soon to call

AI Spend Is Ahead Of The Proof

↗ Brief 09, Signal 01: 70% Capex, 23% Cash Flow↗ Brief 09, Signal 03: 3 Months To 2 Years↗ Brief 12, Signal 01: $11B For A $10B Bet

Brief 09 showed hyperscaler spending on course to pass operating cash flow this quarter, with roughly 33% of it expected to be financed with debt. Brief 12 showed lenders charging for the gap between commitment and evidence: SoftBank raised $11 billion for its OpenAI commitment while its 2031 bonds yielded 8.2%, against 6.7% in January.

The return half is still open.

Goldman Sachs puts the lag between investment and monetization at three months to two years. The crossover itself becomes visible, or does not, when the largest spenders report third-quarter results in October.

Swing and a miss

More Compute Did Not Give Buyers Control

↗ Brief 01, Signal 01: Compute Becomes Inventory↗ Brief 03, Signal 03: The Megawatt Is the Asset↗ Brief 05, Signal 01: A Hundred Year Flood

Brief 01 read Meta's plan to rent out unused computing capacity as the moment leverage moved to the buyers of AI capacity. The market added $179 billion to Meta's value in a single session.

That was early.

The scarcity moved to what compute depends on. Brief 03 placed the binding constraint in power two weeks later, and by August Apple was telling investors that AI data centers were bidding memory chips away from its products. Buyers gained options on compute and lost them on nearly everything underneath it. When one constraint loosens, the useful question is where it moved.

Posture By Week

01Plan02Plan03Mobilize04Commit05Plan06Deploy07Commit08Mobilize09Plan10Commit11Plan12Commit

The weekly calls moved ten times in eleven weeks, since each one read a different week's signals. Averaged across the volume they land at Commit, which is also where the volume closed.

One call overreached. Brief 06 called Deploy on August 10 for an instruction to rerun the operating review on units, customers and headcount. That was a change in measurement, and Commit would have matched it.

05

The Verdict

Volume 1 opened with a brief titled Operators Are Planning On Cuts, about second-half plans that counted on lower interest rates. It closed with the Federal Reserve raising rates and lenders charging more to any borrower that could not prove its case.

Steady demand was the quarter's trap. The common response was to read it as permission to keep the plan written in June, which assumed cheaper money, calmer energy and replaceable suppliers, and to wait for clarity before changing it. In August, Brief 07 found COOs waiting on a recovery from a downturn that never came, and by September Brief 10 found CEOs who had stopped waiting. Waiting had a price: the 30-year Treasury closed at 4.99% the day Volume 1 opened and at 5.49% on September 25,2 so the same long-term commitment cost more to finance in September than in July.

Four signals from this volume will test the fourth quarter: the 30-year Treasury, transaction counts at the large retailers, oil through Saudi Arabia's pipeline around Hormuz, and supplier timelines in third-quarter results. The first readings arrive with the EIA's energy outlook on October 64 and the Federal Reserve's decision on October 28.3 Most fourth-quarter commitments will be made before all four have reported, so the work now is deciding in advance which reading would change each one.

Respectfully,

PJ Bickett

PJ Bickett signature
06

The Decision

Where is your business getting harder faster than your plan is changing?

Two perspectives. One clear picture.Strategic intelligence on Monday. Business conditions on Friday.