Skip to content
← Where the fuck are we going?

The Watch Window

26 signals · quarterly cadence with monthly micro-updates · last update 2026-05-15

Current state
12of 26 at threshold1 at alarm · 11 at watch · 14 in range

Cards highlighted with an orange left-border carry a current reading past at least one defined threshold. Click any card for the threshold values and the cadence that updates them. Signals without an orange border are inside their historical range or have no threshold defined for the current direction of travel.

The Watch Window is the live counterpart to the manuscript. Each signal here corresponds to an arithmetic claim made in the long-form work; each is updated manually, on a cadence that forces me to look at the underlying number before publishing it. Thresholds are numeric. Readings are descriptive. The dashboard refuses to predict timing or path — that is not what these signals are for.

A signal crossing a watch threshold is not a forecast that something will happen by a date. It is a statement that one of the load-bearing assumptions in the diagnosis has moved into a range that no longer cancels out the others. The discipline is to record that movement honestly when it occurs, and to resist the temptation to attach a story to it before the next quarter's data is in.

Diagnosis

US Federal Debt / GDP

On track to pass the postwar peak this year and to keep climbing through the 2030s under any CBO baseline.

31med 55118
~100% as of 2026-Q1

Net interest vs national defence spending

Crossover happened in FY2024 and the gap has widened in every fiscal year since. One dollar in five of every tax dollar collected services debt; the fifty-year average is twelve cents.

0.1med 0.41.1
1.06× (FY2025: $970B vs $916B) as of FY2025 actual (OMB Table 3.2)

Shiller CAPE ratio (S&P 500)

Second-highest reading in 144 years of data. December 1999 was the only higher observation. Current level is being held at 4–5% on the long bond, not zero.

4.8med 1644
~42 as of 2026-04

US market cap / GDP (Buffett indicator)

Highest reading in the postwar series, by a wide margin. Buffett once called this 'probably the best single measure of where valuations stand at any given moment.'

32med 75252
230–252% (range across published series) as of 2026-Q1

Mag 7 share of S&P 500 market cap

Concentration alone is not a thesis; concentration combined with the AI capex-to-revenue gap is.

10med 1731
~31% as of 2026-04

Bottom 50% household wealth share

The floor the sovereign math rests on. The five-year equity bull run was almost entirely a wealth event for the top quintile.

1.5med 2.73.5
~2.5% as of 2025-Q4

US renter households housing-cost-burdened

A record share. The 2020–2024 inflation has not been reversed in rent — it is the new baseline household cost structure, and the wage curve has not closed the gap.

24med 3849
~49% as of 2024 (Harvard JCHS)

Social Security worker-to-beneficiary ratio

The denominator that makes the entitlement math work or not. Trustees project the descent to 2.3 by 2040 — at which point the OASI cut becomes statutory rather than political.

2.3med 3.45.1
~2.7 as of 2025

US total fertility rate

Below replacement, structurally. The migration assumption in the entitlement math now does all the demographic work — the fertility series alone implies a shrinking native-born labour force.

1.6med 2.13.8
~1.63 as of 2024 NCHS provisional

Gold price (USD / oz)

Not a hedge thesis — a fiscal-credibility signal. The manuscript frames gold as the market's reading of how much trust remains in the reserve-currency arithmetic. The trend, not the spot price, is what matters.

252med 11004700
~$4,700 as of May 2026

OASI trust fund depletion year

Statutory cut of 23–28% triggers on depletion absent legislative action. The default position is the cut; the political system decides who absorbs it.

Pending

World Beyond

USD share of global FX reserves

Slow-moving signal. The dollar's reserve status is the largest non-fiscal lever the US Treasury has. Reserve diversification is a path, not an event.

57med 6572
~57% as of 2025-Q3

EM hard-currency credit spreads (EMBI)

Where dollar-funding stress shows up first. The South African vantage matters: SA is the clinical-extreme test bed for what dollar-tight cycles do to EM credit.

175med 380900
~315 bps as of 2026-05-13

USD share of global trade invoicing

The deeper measure of dollar dominance than reserve share. Invoicing moves slowly because the network effect is real; when it does move, it does not snap back.

Pending

SA debt service / consolidated revenue

The clinical-extreme vantage. SA is what the US fiscal arithmetic looks like decades earlier on the curve — the question is whether the US trajectory bends or simply repeats the EM playbook.

8.9med 14.522.1
~21.9% (FY2025/26) as of May 2025 Budget update

Political Amplifier

Mechanism

US credit card revolving APR

The structural transfer from the bottom half of the wealth distribution to the financial sector, in one number. The marginal household has no buffer — this is the cost of that fact.

12.5med 14.522
~22% as of 2026-Q1

Top 10% share of US consumer spending

Aggregate retail-sales resilience can mask median-household stress because the median household is no longer the marginal driver of aggregate consumption. The mechanism by which the K-shape hides.

27med 3250
~50% as of 2025

Treasury auction bid-to-cover (10y / 30y)

Where fiscal stress shows up at market speed. Watch the trend, not the individual print — single auctions are noisy.

1.8med 2.63.5
~2.4× as of 2026-05-13

Subprime auto 60+ day delinquency

The household-balance-sheet signal that surfaces first because the loans are short-tenor and the marginal borrower has no buffer.

2.5med 4.26.5
32-year record (per Fitch / Cox Auto) as of 2026-Q1

AI infrastructure capex-to-revenue gap

The single most concentrated bet in the postwar period. Hyperscaler FCF turn is the threshold where the bet has to start paying or the capex curve breaks.

1med 516
~16× as of 2026-Q1

FINRA margin debt

Late-cycle leverage signal. The size of the margin-call pool that becomes a mechanical seller when the index draws down.

200med 5001280
~$1.28T (ATH) as of Jan 2026

Counterweight

Methodology

Signals are selected to mirror the manuscript's five movements (Diagnosis, World Beyond, Political Amplifier, Mechanism, Counterweight). Categories are descriptive, not prescriptive — the dashboard refuses to encode timing or path predictions. Update cadence per signal is shown inline. The Falsifiability Method documents the diagnostic discipline used to select and interpret each signal.