🦢 Black Swan Early Warning Monitor — September 26, 2026

Rod Fontecilla

Sep 26

24 min read

Updated: 6 days ago

Weekly scan of 13 early-warning indicators for systemic, stagflationary, and tail risk. Data as of Friday, September 25, 2026. The most recent prior report in the archive is dated August 29, 2026 (no reports were filed for September 5, 12 or 19), so all "week-over-week" comparisons below are measured against the August 29 readings — a four-week gap.

Updated September 27, 2026 with an addendum introducing Indicator 14 (AI Capital Window / Funding-Gap Signal) and a revised scenario outlook — see the ⭐ Addendum section below.

🚨 Active Conditions (🔴 Elevated)

The shock came back, and this time the Fed is hiking into it. Four weeks ago this monitor recorded the oil premium draining out of the market as Hormuz mines were cleared. That reversed. Houthi strikes on Saudi Arabia shut the East-West pipeline, the US blockade of Iran remains in force, and Brent spiked from $97.60 on September 22 to $108.40 on September 24 before easing to roughly $106 on reports of a phased US–Iran deal. Brent is up ~20% against the last report and roughly +68% against the pre-war baseline, re-tripping the shock threshold. US gasoline is $4.48, the highest late-September average AAA has ever recorded.

Domestically, the configuration this monitor has described since July arrived on schedule. The FOMC hiked 25 bps to 3.75–4.00% on September 16 in a 12–0 vote, the first increase since 2023; 16 of 19 participants see at least one more hike this year and October odds sit between 54% and 67%. The bond market responded with the deepest selloff of the cycle: the 10-year at 5.17–5.22%, its highest since 2007, the 30-year at 5.48–5.50%, its highest since 2004, German Bunds at a 17-year high and JGBs at their highest since 1996. Michigan sentiment collapsed to 48.1 with one-year inflation expectations jumping to 4.6%.

Against that, the growth data refuse to roll over: August payrolls +162,000 against a 53,000 consensus, initial claims 197,000, and the S&P Global US manufacturing flash PMI at 57.0, the strongest since 2022. That is precisely the stagflation shape — hot activity, hot prices, a tightening central bank, and a household sector that has stopped believing. Six of thirteen indicators now read 🔴, the highest count since this monitor was expanded in April.

🔴 Energy / Commodity Shock (Oil) (upgraded from 🟡) — Brent ~$106, WTI $92.41; Brent +20% vs. Aug 28 and ~+68% vs. the ~$63 pre-war baseline. Houthi attacks closed the Saudi East-West pipeline, forcing more crude through a contested strait; VLCC Oman–China rates hit $870,947/day.
🔴 Inflation / Stagflation Risk — August CPI 3.4% YoY, +0.4% MoM, energy +16.3% YoY, gasoline +27.4% YoY; Michigan 1-yr expectations 4.6% (from 4.0%), 5–10-yr 3.4%. The Fed hiked and projects PCE will not reach 2% until 2029.
🔴 Geopolitical Risk — US blockade of Iran in force; Houthi missile attacks on Saudi Arabia condemned by the UN Security Council, with Saudi Arabia, Turkey and Pakistan convening a chiefs-of-staff meeting; Iran offered on Sep 25 to reopen Hormuz within seven days in exchange for lifting the blockade and unfreezing assets. The Trump–Xi summit (Sep 24) produced a two-month tariff-truce extension and little else; Russia resumed strikes on Kyiv the day after US envoys left.
🔴 Valuation Extremes (CAPE) — Shiller CAPE 40.6–41.2; S&P 500 7,743.41, within 0.5% of its August 28 level despite a 45 bp rise in the 10-year.
🔴 Leverage & Speculation (Margin Debt) — August $1.45T, +2.6% MoM, +37.2% YoY; investor credit balance a record −$1,028.7B. July's de-levering lasted one month.
🔴 Consumer Sentiment & Demand — Michigan final September 48.1 (from 51.7), Expectations 46.3, a bottom-1st-percentile reading 42.5% below its long-run average. Conference Board (August) 89.4, Expectations 68.2; September print due Sep 29.

Early-Warning Precursors

#	Precursor	Now (Level & Status)	3–6-Mo View
1	Liquidity & Yield Curve	10y 5.17% (from 4.73%), 3m 4.18% (from 3.90%), spread +0.99% (from +0.83%); 2y 4.81% (from 4.34%); 30y 5.48–5.50%, highest since 2004; Fed funds 3.75–4.00% after the Sep 16 hike; QT ended — 🟡 (discretionary, held)	The mechanical rule reads 🟢 and the spread actually widened 16 bps — but it widened for the wrong reason. This is a bear steepener: the 3-month rose 28 bps on the Fed hike while the 10-year rose 44 bps and the 30-year 26 bps on inflation and supply fears. The 10-year is now up roughly 125 bps since early March and 70 bps since the June FOMC. Thirty-year mortgages are back above 7%. A curve that steepens because the long end is being repriced for higher structural inflation, while the front end is being pushed up by a hiking central bank, is not the benign steepening the threshold was designed to capture. Held at caution; the failure mode to watch is a disorderly long-end auction, which has not yet occurred.
2	Credit Stress (HY OAS)	ICE BofA US HY OAS 2.80% (Sep 24), from 2.63% — 🟢	Wider by 17 bps over four weeks (+4.5% on the week, +6.5% on the month) — the first sustained widening this monitor has recorded since spring, and it came alongside the bond rout and the oil spike. It is still 70 bps inside the 3.5% caution line and by any historical standard extraordinarily tight. Public high yield has now begun to notice; it has not begun to worry. The pattern to watch is whether widening continues into a fifth week — spreads have historically led equity drawdowns by two to four weeks in stress episodes.
3	Credit-to-GDP Gap (BIS) ⭐	Below US CCyB (0%); gap negative on the last confirmed release (Q4 2025) — 🟢 (BIS Sep 14 update could not be retrieved this run; re-check next week)	No confirmed new data. US total credit to the private non-financial sector was 140.3% of GDP at last reading with the gap substantially negative, far below the Basel III 2 pp activation threshold. Bank-channel credit remains structurally benign. As before, this cycle's leverage lives in margin accounts (see #6), corporate issuance and private credit, none of which this gauge captures.
4	Valuation Extremes (CAPE)	Shiller CAPE 40.6 (thetrading.tools) to 41.18 (GuruFocus, Sep 1); S&P 500 7,743.41, +0.4% vs. Aug 28 — 🔴	Unchanged at an extreme, which is itself the story. Over four weeks the discount rate on every equity cash flow rose 45 bps at the 10-year and the Fed hiked, and the multiple did not compress. The index fell to six-week lows near 7,500 during Warsh's September 16 press conference and recovered all of it within seven sessions. CAPE remains in the 98th–99th percentile of 145 years; the long-term average is 32.6 and the all-time high 44.2. The Russell 2000 at 2,837 is down 4.6% from 2,973 on Aug 28 while the S&P is up — the breadth divergence flagged last report widened.
5	Energy / Commodity Shock (Oil)	Brent ~$105.8 (Sep 25, from $88.29), WTI $92.41 (from $83.44); Brent +20% vs. Aug 28, ~+68% vs. ~$63 pre-war baseline, on track for a 17%+ September gain; US gasoline $4.48/gal (from $4.09) — 🔴 (upgraded from 🟡)	The improvement recorded on August 29 has fully reversed. The +50%-over-baseline threshold trips again by a wide margin. The catalyst was not Hormuz itself — non-Iranian transits actually rose to a three-week high of 97 for Sep 7–13 — but Houthi missile attacks on Saudi Arabia that closed the East-West pipeline, removing the one bypass route for Gulf crude and driving VLCC Oman–China rates to $870,947/day, with West Africa–China at an all-time high. Brent swung more than $10 in two days (Sep 22–24). The offset: Iran's Sep 25 offer to reopen the strait within a week, which knocked ~$2.50 off Brent, and an unexpected 2.97 mb US crude build. This is the single largest deterioration on the board and the one most directly wired into #9 and #12.
6	Leverage & Speculation (Margin Debt)	August $1.45T, +2.6% MoM, +37.2% YoY (+33.1% real); investor credit balance −$1,028.7B, a record — 🔴	July's 5.7% decline, which this monitor called the best structural development on the board, lasted exactly one month. Margin debt re-accelerated in August and the net credit balance crossed −$1 trillion for the first time. Real margin debt is up 570% since 1997 against 361% for the S&P 500. Historically margin debt has peaked two to six months ahead of major market tops; a re-acceleration after a one-month dip does not resolve the question of whether June was the peak, but it does mean the leverage that would amplify any correction has not left the system.
7	Vol Term-Structure (VIX)	VIX 14.87 (Sep 25), from 14.4; VIX3M 17.93, IVTS 0.829, day 119 of contango — 🟢	Contango intact for four months. Over the four weeks since the last report the Fed hiked for the first time in three years, the 10-year hit a 19-year high, Brent spiked $10 in two sessions, and Michigan sentiment fell to a bottom-percentile reading — and the VIX is 0.5 points higher. Equity volatility remains the cheapest insurance on the board against the widest set of outstanding hazards. The MOVE index and rates volatility are where the stress is showing; equity vol has so far declined to import it.
8	Geopolitical Risk ⭐	US blockade of Iran in force since Jul 14; Houthi attacks on Saudi Arabia escalating, East-West pipeline closed, UNSC condemnation; Iran proposed (Sep 25) reopening Hormuz within 7 days for blockade relief and frozen-asset access; Trump–Xi summit (Sep 24) extended tariff truce two months, no Iran breakthrough, no Chinese-bank sanctions announced; Russia struck Kyiv Sep 7–8 after US envoy visit — 🔴	Still two-sided, but the escalation side gained. Escalating: the war widened geographically to Saudi Arabia's infrastructure, drawing in Turkey and Pakistan militarily; commodity-ship transits through Hormuz fell to single digits by late September per regional reporting even as crude transits rose; Iran's military warned it "will respond powerfully to any attack." De-escalating: Tehran's seven-day reopening offer is the most concrete proposal since the June 17 MOU collapsed; Qatar is mediating; a senior US official confirmed frozen-asset access is on the table; the Trump–Xi summit passed without the Chinese-bank sanctions that were the main escalation rung flagged last report. Held at 🔴 — twelfth straight week.
9	Inflation / Stagflation Risk ⭐	August CPI 3.4% YoY (+0.4% MoM), core 2.4% (+0.3% MoM); energy +16.3% YoY, gasoline +27.4%, shelter 3.0%; Michigan 1-yr expectations 4.6% (from 4.0%), 5–10-yr 3.4% (from 3.3%); Fed hiked to 3.75–4.00% Sep 16, October hike odds 54–67%; August PCE due Sep 30 — 🔴	Headline CPI sits at the top of the caution band, but the sub-condition on rising expectations — which was not met on August 29 — is now met decisively: one-year expectations jumped 60 bps in a month and long-run expectations ticked up for the first time in four months. Energy is doing the work: a 2.1% monthly energy rise and 3.9% gasoline rise produced a 0.4% headline print with core still at 0.3%. The Fed's own projections put PCE back at 2% only in 2029. Warsh: "Inflation is too high and has been for too long." With Brent back above $105 in the last week of September, the October CPI print is very likely to be hotter than August's, and the Fed has told the market it will respond. The trapped configuration is no longer a forecast.
10	Dollar Strength / Currency Stress (DXY) ⭐	DXY 100.97 (Sep 25), from 99.67; +1.3% vs. Aug 28, +1.0% on the week, +1.8% on the month; ~0.7% below the late-July ~101.70 peak — 🟢	Back inside the 100–110 stability band and rising on rate differentials. The reflexive risk flagged last report — a hiking Fed driving the dollar sharply higher and relocating stress to EM borrowers and dollar-funded carry — is now the live direction, though at +1.8% on the month it is orderly. The dollar weakened against the yen after Japanese authorities reaffirmed intervention readiness. Watch Japan: JGB yields at their highest since 1996 alongside a US Treasury that has already intervened once in the yen this year is the one currency pairing on the board with the capacity to become disorderly.
11	Financial Conditions Index (NFCI) ⭐	NFCI −0.555 (week ending Sep 18), from −0.566 on Aug 21; prior week −0.557 — 🟢	Tightened by roughly a hundredth over four weeks that included a Fed hike, a 45 bp rise in the 10-year and a 17 bp widening in HY. That is not nothing, but it is remarkably little. Conditions remain well inside loose territory and 25 bps from the caution line. The plumbing continues to hold: SOFR inside the target range, no repo dislocation, no fund gates. The NFCI is confirming what the VIX and HY OAS are saying — the financial system is absorbing the rates shock without transmitting it. Whether that persists through an October hike and a 5.5% long bond is the question.
12	Consumer Sentiment & Demand ⭐	Michigan final September 48.1 (from 51.7), Current Conditions 50.9, Expectations 46.3 (−10.4% MoM); Conference Board August 89.4, Expectations 68.2 (September due Sep 29) — 🔴	Michigan is below the 55 threshold for a second month and now in the bottom 1st percentile of readings since 1978, 42.5% below its long-run average and down 15% year-to-date. Expectations fell more than 10% in a single month. Survey director Joanne Hsu cited "elevated fuel prices and re-escalating trade disputes." Conference Board Expectations have been below the 80 recession threshold for 19 consecutive months as of August. The hard data still disagree — payrolls +162K, claims 197K, and the S&P Global composite at 58.4 — but a household sector this pessimistic with gasoline at $4.48 and mortgages above 7% is not a base for the spending resilience the soft-landing case requires.
13	Global PMI / Growth Momentum ⭐	S&P Global US Mfg PMI 57.0 (September flash, from 53.9 August final; highest since 2022), Composite 58.4; J.P.Morgan Global Composite 53.5 (August, a 27-month high, from 52.7), global manufacturing output 53.0; August payrolls +162K, unemployment 4.1% — 🟢 (upgraded from 🟡)	The growth-momentum downgrade recorded on August 29 did not survive the data. US manufacturing output grew at its fastest pace since April 2022, new orders at the strongest rate in four and a half years, and factory employment at the highest since February 2021 — all five PMI components contributed. Globally, the composite hit a 27-month high with new export business rising for the first time in six months and employment growth the strongest in three years; only France, Canada, Brazil and Kazakhstan contracted. This is unambiguously 🟢 on the rule. The caveat is that it is also the reason the Fed feels free to hike: "remarkable resilience" (Williams) plus 3.4% inflation is the argument for tightening, and a growth reading this hot alongside a consumer reading this cold is the definition of a divergence that resolves one way or the other.

📊 Risk Dashboard

Status	Count	Indicators
🟢 Benign	6	HY OAS, Credit-to-GDP Gap, VIX, DXY, NFCI, Global PMI
🟡 Caution	1	Yield Curve
🔴 Elevated	6	CAPE, Margin Debt, Oil / Commodity Shock, Geopolitical, Inflation / Stagflation, Consumer Sentiment

Read: The scoreboard moved from 5 red / 3 yellow / 5 green to 6 red / 1 yellow / 6 green. Two indicators changed status and both moved outward: oil returned to red, and global growth moved to green. Six 🔴 readings is the trigger level at which this framework requires the combined stagflation and tail weight to exceed 40%.

The board has polarised. Every gauge that measures market pricing of stress — credit spreads, equity volatility, financial conditions, the dollar — is green. Every gauge that measures the underlying hazard — valuation, leverage, the energy shock, geopolitics, inflation, the consumer — is red. The one indicator that measures real activity is green and getting greener. There is no yellow left except the yield curve, and that one is yellow only because this monitor overrides its own rule.

Four weeks ago the framing was that risk had migrated from external-and-acute to internal-and-structural. Both are now present at once. The external shock is back at full strength — Brent +68% over baseline, gasoline at a seasonal record — and the internal configuration has hardened from a forecast into policy: the Fed has hiked, has told the market to expect more, and has published projections showing inflation not returning to target for three years. The market's answer, so far, is a VIX of 14.9 and an S&P within half a percent of where it started.

That is either the correct read of a resilient economy that can absorb 5% Treasuries and $105 oil, or it is the most extended gap between hazard and hedging this monitor has recorded. The next four weeks — an October FOMC, a September CPI with $105 Brent in it, and a Conference Board print on Sep 29 — will say which.

🔄 Notable Week-over-Week Changes

All comparisons are to the August 29, 2026 report; the intervening three weekly runs were not filed.

The Fed hiked (Sep 16). 25 bps to 3.75–4.00%, 12–0, the first increase since 2023. Sixteen of 19 participants project at least one more hike in 2026; four project two; median year-end dot 4.1% (from 3.8% in June). PCE projected to reach 2% only in 2029. Equities fell 1% to six-week lows during the press conference; the 10-year crossed 5% the same day. October hike odds now 54–67%; December odds imply an 89% probability of at least one further hike.
The long end broke to multi-decade highs. 10-year 4.73% → 5.17–5.22% (highest since 2007), 30-year 5.22% → 5.48–5.50% (highest since 2004), 2-year 4.34% → 4.81%, 3-month 3.90% → 4.18%. German 10-year Bund above 3.6% (17-year high), Japanese 10-year at highest since 1996. Freddie Mac 30-year mortgage 7.03%. Analyst consensus attributes the bulk of the move to repriced Fed expectations, the remainder to growth and oil.
Oil re-shocked. Brent $88.29 → ~$105.8 (+20%), WTI $83.44 → $92.41 (+11%); Brent touched $108.40 on Sep 24, a $10+ swing in two days, on Houthi missile claims against Saudi Arabia and an Iranian warning of escalation. Brent on track for +17% in September. Gasoline $4.09 → $4.48, the highest late-September average on record; September's month-to-date average of $4.30 is on pace to exceed the prior September record of $3.83 (2023).
Hormuz: more oil, fewer everything else. Non-Iranian transits rose to 97 for Sep 7–13 (three-week high; three-week average 88 vs 67 in late July–mid August), driven by crude as the Saudi East-West pipeline closure forced barrels back through the strait. VLCC Oman–China $870,947/day; West Africa–China $509,000/day (record, 2.7× early September); Suezmax West Africa–Europe $238,000/day (record). Regional reporting puts commodity-ship transits in single digits.
Iran offered a seven-day reopening (Sep 25). Phased deal: Hormuz reopens within a week if the US lifts its blockade and grants access to frozen assets; President Pezeshkian said he is "ready to strike a deal." No nuclear concessions. Qatar mediating. Brent fell ~2% on the report.
Trump–Xi summit (Sep 24) delivered a truce, not a deal. Tariff truce extended two months through January; an AI dialogue channel established; "no change" on Taiwan; no announced sanctions on Chinese banks; US ambassador said Trump told Xi that Chinese support for Iran's war effort is "unacceptable" and China "claimed commitment to cease." Described by multiple outlets as spectacle over substance.
August CPI (Sep 10): 3.4% YoY, +0.4% MoM; core 2.4%, +0.3% MoM. Energy +2.1% MoM, +16.3% YoY; gasoline +3.9% MoM, +27.4% YoY; food 2.7% YoY; shelter 3.0% YoY. The report was read as keeping the September hike in play — and it did.
Michigan sentiment fell to 48.1 (final September, from 51.7), Expectations 46.3 (−10.4%), Current Conditions 50.9. One-year inflation expectations 4.0% → 4.6%; 5–10-year 3.3% → 3.4%. Bottom 1st percentile since 1978.
Payrolls surprised massively to the upside. August +162,000 vs 53,000 consensus, the best month since March; unemployment 4.1% unchanged. Initial claims fell to 197,000 in the latest week. The −79K benchmark revision flagged last report was followed by the strongest print of the year.
US manufacturing PMI jumped to 57.0 (September flash, from 53.9), the largest monthly improvement since May 2022 and the highest level since April 2022; composite 58.4. Global composite PMI 53.5 in August, a 27-month high. Global growth upgraded to 🟢.
Margin debt re-accelerated. August $1.45T, +2.6% MoM, +37.2% YoY; net investor credit balance −$1,028.7B, crossing −$1T for the first time. July's decline lasted one month.
HY OAS widened 17 bps to 2.80% (Sep 24) — the first four-week widening since spring, +4.5% on the week alone. Still 70 bps inside the caution line.
NFCI tightened marginally to −0.555 (from −0.566); financial conditions remain loose through the hike and the bond rout.
DXY 99.67 → 100.97 (+1.3%), second consecutive weekly gain, back inside the stability band.
Equities flat, breadth worse. S&P 7,711.76 → 7,743.41 (+0.4%); Nasdaq 26,402 → 27,069 (+2.5%); Dow 53,560 → 51,829 (−3.2%); Russell 2000 2,973 → 2,838 (−4.6%). The Dow snapped a three-week losing streak on Friday. Gold $4,456 → $4,321 (−3.0%); bitcoin ~$77,500 → $83,954 (+8.3%).
VIX 14.4 → 14.87, VIX3M 17.48 → 17.93, IVTS 0.826 → 0.829; contango day 100 → day 119.
Little changed: CAPE (~40.6–41.2), BIS credit gap (Sep 14 update not retrieved; last confirmed reading unchanged), Conference Board (September print due Sep 29), August PCE (due Sep 30).

Net: the oil shock returned, the Fed hiked into it, the long bond broke to a 22-year high, the consumer broke down — and the growth data, the credit market and the VIX all shrugged.

🧭 Consensus Black Swan Outlook (3–6 Mo)

Scenario	Probability	WoW Change
🟢 Base-Case Soft Landing — growth resilient, inflation contained, no major shock	21%	▼ 3pp
🟡 Mild Risk-Off / Correction — equities −10–15%, no recession	32%	▲ 1pp
🔴 Stagflationary / Recessionary Downside — growth contracts, inflation stays elevated, Fed trapped	35%	▲ 2pp
⚫ Tail / Black Swan Cascade — systemic stress, equities −30%+, credit freeze	12%	— 0pp
	100%

Rationale: The combined downside bucket rises to 47% from 45%, clearing the >40% floor this framework imposes when six or more indicators read 🔴. The three points come out of the soft landing, which is now the least likely single outcome on the board.

Soft landing falls 3pp to 21%. The soft landing specifically requires inflation to come down without the Fed forcing it and without an external price shock. Both conditions failed in the same four weeks. Brent is back above $105 and will be in the September and October CPI prints; the Fed has hiked and told the market to expect more; and its own projections concede inflation does not reach target until 2029. What keeps this scenario above 20% is that the growth data are not merely resilient but accelerating — a 57.0 manufacturing PMI, +162K payrolls, 197K claims and a 27-month-high global composite. An economy growing this fast can absorb a great deal. But "growth resilient, inflation contained" now requires an oil price the market does not have and a Fed pause the Fed has ruled out.

Stagflation rises 2pp to 35%, a new cycle high and now the modal outcome by three points. Every leg of the configuration this monitor has tracked since July is now in place simultaneously and none of them is hypothetical: energy-driven headline inflation with expectations rising (4.6%), a central bank actively tightening (3.75–4.00% and guiding higher), a household sector in the bottom percentile of sentiment (48.1) with real purchasing power being eroded at the pump ($4.48), and a long bond at 5.5% pushing mortgages above 7%. The reason this is not higher is the same reason the soft landing is not lower: real activity has not yet contracted, and the stagflation scenario requires it to. The divergence between a 57 PMI and a 48 consumer is the widest this monitor has seen, and stagflation is the scenario in which the consumer turns out to be right.

The correction rises 1pp to 32%. A 41 CAPE against a 5.2% 10-year and a 5.5% 30-year is a multiple that has not repriced to its discount rate, and the breadth data say the repricing has begun beneath the index: the Russell 2000 is down 4.6% and the Dow 3.2% over the same four weeks the S&P rose 0.4%. HY spreads widened 17 bps — modest, but the first sustained move in months, and credit tends to lead equity by two to four weeks. The mechanical case for a 10–15% index-level correction is stronger than it was; the reason this bucket only gains a point is that some of what would have been "correction" probability has moved to "stagflation," because a correction that coincides with an oil shock and a hiking Fed has a harder time stopping at −15%.

The tail holds at 12%. The cascade mechanics that improved on August 29 have partly reversed — HY widened, the long end broke out — but the plumbing that would actually transmit a cascade remains sound. NFCI moved one hundredth in four weeks that included a hike and a 45 bp rates move. SOFR is inside the range. No fund has gated, no repo dislocation, no large-bank stress. The Treasury market has repriced violently but in an orderly fashion — there has been no failed auction. A margin-debt balance past −$1T and a VIX at 14.9 mean the amplification capacity for a tail event is as high as it has ever been; what is missing is a trigger in the financial system itself. The two candidates worth naming are a disorderly long-end Treasury auction and a JGB/yen dislocation that forces Japanese repatriation from Treasuries. Neither is visible yet. The tail stays at 12 rather than rising because the probability mass that would have gone there has, for now, a more likely home in the stagflation bucket.

Five things to watch this week: September 29 brings the Conference Board's September survey — if Expectations follow Michigan's 10% drop from 68.2, the consumer indicator moves from red to alarming. September 30 brings August PCE, the Fed's preferred gauge, with July's headline 3.7% / core 3.3% as the bar. October 1 brings the global manufacturing PMI headline and ISM for September. Watch whether Iran's seven-day reopening offer becomes a signed instrument or another collapsed MOU — it is the single variable that most directly moves indicators #5, #8, #9 and #12 at once. And watch the 30-year auction calendar: the long bond at 5.5% has cleared every auction so far, and a tail on a 30-year sale is the most plausible route from the rates shock into the plumbing.

⭐ Addendum (September 27, 2026) — New Indicator 14: AI Capital Window / Funding-Gap Signal

Added one day after publication to establish a baseline. Rationale: booms need continuous fresh capital and turn into busts when the window to raise new money closes — "burn rate" and "funding gap" entered the vernacular in March 2000 as the Nasdaq peaked, and Freddie Mac's failure to raise common equity in November 2007 signalled the window had shut for the whole financial system. None of indicators 1–13 measures whether the capital window funding the AI build-out is open. Five sub-signals, each with its own status, roll up to one headline reading.

#	Precursor	Now (Level & Status)	3–6-Mo View
14	AI Capital Window / Funding-Gap Signal ⭐	(a) Vernacular 🟡 — "cash burn" and "burn rate" attached to OpenAI in the FT (Sep 19: $278B cumulative negative free cash flow projected 2026–30; the March $122B raise "exhausted by 2028 at the current burn rate"), the WSJ (Sep 26, explicitly naming "funding gap" as the signal to watch) and Axios (Sep 25, Oracle data-center debt); Bloomberg and Semafor on derailed AI IPOs. "Funding gap" itself has appeared in one major outlet, as a warning rather than a description. (b) IPO window 🟡 — OpenAI IPO pushed to 2027; Anthropic slipped twice (September → mid-October → November, seeking ~$2T / up to $100B); Holtec suspended its ~$900M IPO on Sep 16 the day it was to price, citing market conditions; SB Energy delayed; X-Energy −37% and Standard Nuclear −21% below issue; only 3 IPOs priced in September and 5 of the 10 largest 2026 listings trade below issue. (c) Mega-round status 🟢 — OpenAI closed $122B at $852B (March) and is in investor-initiated talks above $1.2T (FT, Sep 15); Anthropic Series H $965B (May 29); xAI $20B Series E (January, Nvidia participating); SoftBank's junk-bond sale to fund its OpenAI stake drew >$20B of early orders. Rounds are still closing up. The caveat is structure: a growing share of the money is vendor or circular (Nvidia, SoftBank borrowing in high yield, Oracle counting on OpenAI revenue to service its own debt). (d) AI-infra credit 🟡 — Oracle issued a force majeure notice on its $18B Project Jupiter lease (Sep 24); its 2055 bonds traded at 77 cents, CDS rose, and data-center bonds widened intraday before partly recovering; Oracle stock −30% YTD. CoreWeave's $2.6B term loan was flexed from S+425–450 to S+550 at 97 OID (10.44% yield) in July; CoreWeave 5-yr CDS has been quoted near 675 bps (~42% implied 5-yr default probability) with $18.6B of debt against $1.9B of sales. Yet a CoreWeave-tied data center still placed $1.1B of junk bonds on Sep 23 — the market is repricing, not refusing. (e) Secondary market 🟢 — Anthropic voided unauthorised secondary transfers in May; OpenAI's secondary remains open near its $852B August tender level; retail pre-IPO wrappers trade at premiums (the Robinhood SpaceX vehicle was worth roughly two dollars per dollar of private stock before listing). Discounts are absent; froth is present. Headline: 🟡 Nervousness (first reading — WoW n/a)	This is the textbook "nervousness" configuration the framework was built to catch: the frontier labs can still raise whatever they want at rising valuations, while the periphery — nuclear suppliers, energy spin-outs, the most levered GPU landlord, the lowest-rated hyperscaler — is being repriced or turned away. In 2000 the window closed from the outside in, and that is the direction of travel here. Two things would move this to 🔴: OpenAI's $1.2T talks failing to close or closing at a flat/structured valuation, or Oracle's force majeure escalating into a missed lease payment, covenant breach or pulled financing that forces the OpenAI–Oracle–SoftBank circle to be marked. Two things would move it back to 🟢: Anthropic pricing in November at or above range, and CoreWeave/Oracle spreads retracing. Watch the SoftBank junk deal's final pricing and the October Anthropic Q3 numbers.

Revised Risk Dashboard (14 indicators): 🟢 6 (HY OAS, Credit-to-GDP Gap, VIX, DXY, NFCI, Global PMI) · 🟡 2 (Yield Curve, AI Capital Window) · 🔴 6 (CAPE, Margin Debt, Oil, Geopolitical, Inflation / Stagflation, Consumer Sentiment).

Revised Consensus Outlook (3–6 Mo), incorporating Indicator 14:

Scenario	Probability	Change vs. Sep 26
🟢 Base-Case Soft Landing — growth resilient, inflation contained, no major shock	21%	— 0pp
🟡 Mild Risk-Off / Correction — equities −10–15%, no recession	30%	▼ 2pp
🔴 Stagflationary / Recessionary Downside — growth contracts, inflation stays elevated, Fed trapped	35%	— 0pp
⚫ Tail / Black Swan Cascade — systemic stress, equities −30%+, credit freeze	14%	▲ 2pp
	100%

Rationale for the shift: A 🟡 capital-window reading adds two points to the Tail by rule, taken from the Correction bucket rather than from Stagflation or the Soft Landing, because the AI funding channel changes the shape of an equity drawdown rather than its likelihood. If a 10–15% correction begins while Oracle is already invoking force majeure, CoreWeave is paying 10.4% for term debt and the three largest private companies in history are all waiting on a window, the correction has a mechanism to become something larger: falling public multiples reprice the private rounds, which delays the IPOs, which strands the debt-funded builders that were counting on those proceeds. That is the Pets.com-to-Nasdaq sequence in reverse order of size, and it is the most plausible path from a 41 CAPE and a −$1T margin balance to a −30% outcome. The Tail at 14% is still below its May levels — the plumbing is sound and the frontier labs are still raising — but it now has a named transmission channel it did not have on Friday.

Indicator 14 will be tracked in every weekly report from October 3, 2026. Sources for this addendum: FT via Silicon Review – OpenAI cash burn $278B by 2030, PYMNTS / FT – OpenAI eyes $1.2T pre-IPO round, Bloomberg – OpenAI weighing funding round above $1.2T, The Decoder / WSJ / The Information – Anthropic postpones IPO, Forbes – Anthropic IPO slips to November as retail money floods pre-IPO funds, Bloomberg – Holtec suspends US IPO, Investing.com – Holtec postpones IPO citing market conditions, Bloomberg / Yahoo – Derailed IPO plans rattle US market ahead of Anthropic debut, Semafor – AI-related IPOs postponed as investors grow cautious, Axios – Oracle rekindles the data center debt risks, Yahoo Finance – CoreWeave completes $2.6B term loan after raising spread and yield, Bloomberg – CoreWeave-tied data center raises $1.1B in junk bonds, RIA – Oracle and CoreWeave CDS spreads widening, Value Add VC – Anthropic blocked secondary sales, TechCrunch – Anthropic warns investors against secondary platforms.

📚 Sources

⚖️ Disclaimer

This post is published for informational and educational purposes only. Nothing here is individualized investment advice, a recommendation tailored to any person's circumstances, an offer, or a solicitation to buy or sell any security or financial instrument. This post is not an investment advisory service, and no advisory or fiduciary relationship is created between me and any reader.

I am not a registered investment adviser, broker-dealer, or financial planner. The indicators, thresholds, status ratings, and scenario probabilities are my own subjective framework and judgment, not forecasts or guarantees; they can be and often will be wrong. Data are drawn from third-party public sources believed to be reliable but not independently verified, may be preliminary, revised, delayed, or contain errors, and are current only as of the date shown. Portions of this post are produced with the assistance of AI tools and may contain inaccuracies. Past performance and historical relationships do not guarantee future results.

Investing involves risk, including the possible loss of principal. You are solely responsible for your own investment decisions and should consult a licensed financial, tax, or legal professional before acting on anything you read here. I may hold positions in securities or asset classes discussed and may trade them without notice. To the fullest extent permitted by law, I disclaim all liability for any loss or damage arising from reliance on this content. Links to third-party sites are provided for convenience only and do not constitute an endorsement.

Auto-generated by Black Swan Monitor · September 26, 2026 · Addendum September 27, 2026 · Next run: October 3, 2026

ADANA TECHNOLOGIES

🦢 Black Swan Early Warning Monitor — September 26, 2026

🚨 Active Conditions (🔴 Elevated)

Early-Warning Precursors

📊 Risk Dashboard

🔄 Notable Week-over-Week Changes

🧭 Consensus Black Swan Outlook (3–6 Mo)

⭐ Addendum (September 27, 2026) — New Indicator 14: AI Capital Window / Funding-Gap Signal

📚 Sources

⚖️ Disclaimer

Recent Posts

AI Strategies

AI Pilots

AI Solutions