Evidence as of October 2, 2026, with a supplementary capture closing October 3, 2026 (12:23 EDT). Every figure below is attributed to the publisher and the date the evidence record carries. Where a source refused automated retrieval, that is stated. Comuvia's own forecast numbers are labelled as judgement, not evidence.
Two weeks apart, the same economy produced two defensible stories. Central banks described activity as solid and kept tightening. Then a weak jobs report landed, long-term borrowing costs stayed high, and the 2008 forecasters got a fresh hearing. Watch the video record for the speakers in their own voices; this article sets the dated claims against each other in text.
What the officials put on the record
The Federal Open Market Committee approved its September 16, 2026 statement by a 12–0 vote and decided to raise the target range for the federal funds rate — the band the Fed sets for overnight lending between banks — by a quarter of a percentage point, to "3-3/4 to 4 percent". That statement said "Economic activity is expanding at a solid pace", that job gains "have kept pace with the workforce", and that "Inflation remains elevated" (FOMC statement).
The European Central Bank raised its three key rates by 25 basis points — hundredths of a percentage point — on September 10, 2026, saying inflation was "set to remain well above target for an extended period" (ECB decisions). At the European Parliament's economic affairs hearing on September 28, Christine Lagarde said that "while the shock is too large to look through" — too large to treat as temporary and ignore — "we view a measured response as appropriate" (BIS mirror); two days later Isabel Schnabel argued the repricing of interest rates "will dampen global aggregate demand" (ECB speech).
The Bank of England, in minutes published September 17, 2026, voted "by a majority of 6–3" to hold Bank Rate at 3.75 percent, three members preferring a rise to 4 percent; UK consumer-price inflation "increased to 3.1% in August and is likely to rise further over coming quarters" (Monetary Policy Summary).
Then the tone shifted. On October 1, Fed Vice Chair Jefferson said of the next rate decision that "my colleagues and I will need to come to our own judgment, which may take more time", with markets "reassessing" the outlook (Reuters via Kitco).
Indicator by indicator
| Indicator (as of) | What the record shows | Reading |
|---|---|---|
| Fed policy language, Sept 16 | Activity "expanding at a solid pace"; rate raised a quarter point | Solid |
| US real GDP, Q2 2026 (BEA third estimate, Sept 25) | +2.2% at an annual rate, after +2.5% in Q1 | Solid |
| September payrolls (CNBC, Oct 2) | "only 29,000 jobs in September, below estimates for a gain of more than 80,000" | Calm before |
| Announced layoffs, September (CBS News / Challenger, Gray & Christmas, Oct 1) | 43,281, down 20% year on year, "the lowest level in four years" | Solid |
| Initial jobless claims (US Dept of Labor, Oct 1) | 197,000 initial, 1,701,000 continuing — held by locator only, no located excerpt | Solid |
| 10-year Treasury yield (US Treasury daily curve) | 5.29% on Sept 30; 5.28% on Oct 2 | Calm before |
| Rate expectations (CME FedWatch via CNBC, Oct 2) | "only a 17% chance" of an October quarter-point rise, against roughly 36% a week earlier | Calm before |
| Gasoline (AAA, Oct 1) | September's monthly average $4.33, above the previous September record; the national average then fell to $4.41 | Mixed |
| Household debt, Q2 2026 (New York Fed, Aug 11) | Fell $13bn to $18.8 trillion; 4.7% of balances "in some stage of delinquency", improved slightly | Solid |
| Outside cycle model, 2026 Q2 (GDP Dynamics, retained Oct 3) | Phase "Expansion", strength "Above Normal", direction "Improving"; recession warning "Low" at one quarter, six months and one year | Solid |
The pattern: the strongest "solid" evidence is mostly backward-looking and quarterly; the strongest "calm before" evidence is recent and monthly. That is a Comuvia judgement about the table's shape, not a finding in any source.
Who could absorb a shock
The Fed's Economic Well-Being of U.S. Households in 2025 report, released May 13, 2026, found that 73 percent of adults were "doing okay or living comfortably" financially — below the 78 percent high of 2021; that 63 percent would "cover a $400 emergency expense using cash or its equivalent"; and that 55 percent "said they had set aside money for three months of expenses" (press release, savings chapter).
Alongside it sit the Fed's Distributional Financial Accounts, last updated September 18, 2026, and household debt-service ratios — the share of disposable income absorbed by required debt payments — published September 22, 2026. None of this says how a shock would land; it says where households start.
The financing chain behind the AI boom
SoftBank Group announced on September 24, 2026 senior notes of "approximately USD 11.1 billion" due 2030 to 2034, and on October 1 that it had "executed the third and final tranche of the follow-on investment" in OpenAI — USD 10.0 billion, completing a USD 30.0 billion programme (notes, tranche).
Bank of England Governor Andrew Bailey, in a BBC interview published October 1 and syndicated by KBC, said "Everybody is currently priced to be a winner" and, asked whether an AI bubble could burst, "You could see some correction of asset prices at some point" (KBC). The record carries the syndicated version because the paywalled originals refused retrieval.
What the video's fragments do and do not prove
The video quotes Steve Keen — whose own upload on the record is titled "I Predicted the 2008 Crash, Here's What Always Happens Before a Crisis" — from four recordings, and uses fragments from Rebel Capitalist, Money & Macro, Patrick Boyle and two GDP Dynamics livestreams. For each, the record retains the identity of the recording — title, channel, watch or oEmbed page — and notes the excerpt was cut locally. It does not retain a verified transcript of what was said inside each clip. A reader can check which recording a fragment came from, but not yet a quoted sentence against a located excerpt: these are sourced attributions, not verified claims.
One attribution has already been corrected. The narration calls the business-cycle modeller "an independent modeller". He is Ty Keynes, author of the GDP Dynamics channel, whose book Beyond Equilibrium was released October 3, 2026 and launched in a livestream with Steve Keen; the record's review log notes he said there that private debt and credit dynamics are not yet in his model. His own model page states its cards are "State-model signal categories, not recession probabilities."
Comuvia's sealed numbers — labelled as judgement
Comuvia's ForeGlass verification ledger, built September 9, 2026 and retained October 3, carries these open claims: probability 0.55 that a recession dated by the National Bureau of Economic Research begins on or before June 30, 2027; probability 0.35 that the S&P 500 suffers a peak-to-trough fall of 20 percent or more with the trough on or before December 31, 2027, which the ledger itself calls "modestly above the ~0.25 base rate… elevated, not a crash call"; and, as annual means for 2027, US unemployment of 4.3–6.5 percent (central 5.2), real GDP growth of −1.5 to +1.8 percent (central +0.6), and a household debt-service ratio of 11.2–12.6 percent (central 11.9). A macro-fragility tripwire "stays BELOW its 0.40 elevated criterion" through 2026 Q4 — in the ledger's words, "this model's conditional hazard, not a forecast of the world". First public scoring: October 15, 2026.
These are Comuvia's forecasts, sealed with a hash before the outcome. A hash shows the content has not changed since it was recorded; it does not make the forecast right.
What this record cannot settle
The Bureau of Labor Statistics refused automated retrieval (HTTP 403), so the September Employment Situation is held by locator only — payrolls +29,000, unemployment 4.2 percent — and the figure the video quotes comes from CNBC. Two FRED series and an Associated Press report timed out; the OECD and IMF outlooks are retained by locator only. The record's own summary calls the 10-year yield a 2007 high, but the evidence packet retains the Treasury daily rows rather than the historical series, so read that comparison as the video's framing, not a checked figure. The record's caveats state that attribution, source identity and located excerpts are checked, while "truth, rights clearance and public release require editorial review". The video uses AI-generated narration and a synthetic presenter.
Not investment advice.
Not established: which reading is correct, the next Fed move, or any fragment claim not yet matched to a located excerpt.