The Red Flags Keep Piling Up: OpenAI Cuts Its Funding Target, AI Stocks Fall Hard and the Stressors Build
Published: November 1, 2026
I wrote in May that the S&P 500 at 7,400 looked like irrational exuberance. The red flags have kept piling up... like waves building into a tsunami.
OpenAI and Anthropic
A big red flag this week is OpenAI cutting its funding target. Some investors have reduced what they were putting in. The Anthropic IPO is under close watch. Strong demand can be expected, but a tepid response would hit the whole AI trade.
The AI stocks
SpaceX's record IPO in June was priced at $135 a share. On Friday it closed at $134.72, 36% below its June high. Seven of the nine big AI names in the deck are 15% to 60% below their 2026 highs. Thursday and Friday were a bloodbath for some. Alphabet fell 13% on Thursday, its worst day ever. Intel fell 29% on Friday, its worst since the 1970s.
AI will pay off
The technology is revolutionary and will pay off. We have seen this before: valuations boom, then bust, and the real value comes after.
AI valuation is only one part of the story
The bigger worry is how many stressors are building at the same time. Debt, high bond yields, costlier insurance on AI debt, oil, possible write-downs, a stretched US consumer and AI-related job cuts. The build itself is running into limits on power, equipment and local opposition. New York has paused state permits for new hyperscale data centres for up to a year. As if that were not enough, a euro-zone debt crisis is brewing, led by France. To top it off, hardly a day goes by without news of a big hack, and AI is adding to the risk. This week Binance was hit by a $6.3 billion theft, the largest crypto theft on record.
More analysis at tigzig.com/analysis
A few nuances the deck could not fit
CNBC reported on 31 Oct that OpenAI now seeks up to $20 billion, down from about $30 billion. The FT puts the valuation at $1.25 trillion. The round is still open, so the figures can change.
SpaceX listed on 12 Jun at its $135 IPO price and opened at $150. Its highest close was $211.39 on 16 Jun. It also closed below $135 in July and August, as low as $108.27 on 5 Aug.
Alphabet's 13% fall on 29 Oct was its steepest drop ever, and Intel's 29% on 30 Oct its worst day since the 1970s, per CNBC.
Free cash flow is operating cash flow less capital spending and finance lease repayments, from the company filings. Oracle's quarter ends a month early.
Gartner's $8.5 trillion is the cumulative AI revenue needed by 2030 to cover announced spending, financing costs and a 12% return on capital. Its forecast is at most $3.6 trillion.
The year-ago default swap levels are worked back from the changes CNBC reported.
Brent's 30 Oct price is the December contract, which expired that day. The January contract settled at $113.43.
The Binance theft affected $6.3 billion of client assets on 29 Oct, per CNBC.
A few earlier analysis, for quick reference
What kind of drawdown are we looking at this time? 24 Oct: tigzig.com/post/drawdowns-this-time-different-oct2026
A crescendo of warnings on AI debt and returns, 23 Oct: tigzig.com/post/ai-warnings-crescendo-oct2026
The biggest tech build meets the bond rout, 28 Sep: tigzig.com/post/ai-hyperscalers-debt-bond-rout-sep2026
France's sovereign debt crisis, 25 Oct: tigzig.com/post/france-sovereign-debt-crisis-oct2026
US jobs, September 2026, 4 Oct: tigzig.com/post/us-jobs-september-2026-update-oct2026
S&P at 7,400: irrational exuberance, 17 May: tigzig.com/post/sp500-irrational-exuberance-7400-may2026
Informational and educational analysis from public sources, not investment advice.
The red flags keep piling up
Browse the slides or download the PDF
Full Deck Content (Text Format)
Text below was extracted from the source deck. Chart visuals stay in the PDF and as slide images above the post.
Slide 1
TIGZIGGlobal Macro 1 November 2026
A bust likely before AI pays off
To me this looks like the dot-com bust and 2008 happening together.
It will take time. The players left standing may be different.
The five biggest AI builders now spend $1.18 on capex and leases for every dollar of operating cash. Prices have run ahead.
Gartner sees AI revenue at less than half of what is needed, and over $3.5 trillion could face write-downs by 2030.
Oil is near its wartime high and US households are stretched. Euro-zone debt is under strain, led by France.
Past busts fall, years to recover S&P 2000 -49% 7.2 yrs S&P 2008 -57% 5.5 yrs Sensex 2000 -56% 3.9 yrs Sensex 2008 -61% 2.8 yrs
AI names now % below 2026 high CoreWeave -60% Intel -53% Oracle -47% Broadcom -32% Alphabet -30% Meta -21% Amazon -15% Nvidia -14% Microsoft -4% S&P 500 -3%
AI cash flowNegative ValuationsOvervalued AI creditStress Interest ratesHigh US jobsWeak US householdsStretched Iran warOngoing Euro-zone debtStrain
Amar Harolikar, ACA · Decision Sciences & Applied AI1 of 14tigzig.com
Slide 2
In the dot-com and 2008 busts, prices fell by half or more and took years to come back
The S&P 500 took 7.2 years after 2000 and 5.5 after 2007. The Sensex took 3.9 and 2.8.
S&P 500 Log scale 1995 2005 2015 2026 -49%, 7.2 yrs -57%, 5.5 yrs 7,580 BSE Sensex Log scale 1999 2008 2017 2026 -56%, 3.9 yrs -61%, 2.8 yrs 74,460
Gartner, 14 Oct Write-downs “Failure to achieve sustainable ROIC on rapidly depreciating assets would almost certainly necessitate write-downs.” More than $4 trillion could be exposed through 2032.
Over $3.5 trillion of AI infrastructure spending could be exposed to impairment by the end of 2030.
80% of neoclouds are expected to cease independent operations by 2030.
SourcesYahoo Finance via tigzig.com APIs; Gartner; Earlier: 24 Oct, 17 May
Amar Harolikar · Decision Sciences & Applied AI2 of 14tigzig.com
Slide 3
Seven of nine big indices are 9% to 37% below their 2026 highs. The S&P 500 and DAX are 3% off
5 of the 9 are still up for the year. India, China and France are down.
End-2025 to 30 Oct. Red dot: the 2026 high. Navy dot: latest close.
S&P 500 +11% YTD, 3% off high Jan May Sep 7,819 7,580 DAX, Germany +5% YTD, 3% off high Jan May Sep 26,570 25,818 CAC 40, France -2% YTD, 9% off high Jan May Sep 8,726 7,969 Taiwan +56% YTD, 9% off high Jan May Sep 49,823 45,212 Kospi, Korea +36% YTD, 37% off high Jan May Sep 9,115 5,731 Nikkei, Japan +26% YTD, 13% off high Jan May Sep 72,366 63,290 Hang Seng -9% YTD, 17% off high Jan May Sep 27,968 23,294 Shanghai -9% YTD, 15% off high Jan May Sep 4,243 3,599 Nifty 50, India -12% YTD, 12% off high Jan May Sep 26,329 23,053
SourcesYahoo Finance via tigzig.com APIs; Earlier: 24 Oct, 19 Oct
Amar Harolikar · Decision Sciences & Applied AI3 of 14tigzig.com
Slide 4
Alphabet fell 13% in a day, its worst ever. Intel fell 29%, its worst since the 1970s
Seven of these nine AI names are 15% to 60% below their 2026 highs.
End-2025 to 30 Oct. Red dot: the 2026 high. Navy dot: latest close.
CoreWeave -60% off high, -23% YTD Jan May Sep $138 $55 Intel -53% off high, +78% YTD Jan May Sep $141 $66 Oracle -47% off high, -32% YTD Jan May Sep $246 $131 Broadcom -32% off high, -5% YTD Jan May Sep $480 $326 Alphabet -30% off high, -10% YTD Jan May Sep $402 $280 Meta -21% off high, -7% YTD Jan May Sep $778 $615 Amazon -15% off high, +4% YTD Jan May Sep $284 $240 Nvidia -14% off high, +10% YTD Jan May Sep $239 $205 Microsoft -4% off high, +9% YTD Jan May Sep $543 $522
SourcesYahoo Finance via tigzig.com APIs; worst days: CNBC 30 Oct, 30 Oct; Earlier: 23 Oct
Amar Harolikar · Decision Sciences & Applied AI4 of 14tigzig.com
Slide 5
Oil is up 93% this year and US diesel is at $6.35 a gallon. Gold has broken below $4,000
The 30-year mortgage rate is 7.2%, up from 6.2% at the end of 2025.
End-2025 to 30 Oct. Red dot: the 2026 high. Navy dot: latest close.
US 10-year yield Peak 5.31% on 5 Oct Jan May Sep 5.31% 4.93% 30-year mortgage rate Peak 7.40%, highest since Nov 2023 Jan May Sep 7.40% 7.24% Brent crude +93% YTD, near its high Jan May Sep $118 $118 US diesel +81% YTD, $ a gallon Jan May Sep $6.53 $6.35 Gold -10% YTD, 27% off high Jan May Sep $5,318 $3,899 Silver -30% YTD, 57% off high Jan May Sep $115 $49
SourcesYahoo Finance via tigzig.com APIs; EIA, Freddie Mac via tigzig.com Tremor. Gold, silver, Brent: futures. Brent is the December contract, which expired 30 Oct; January settled at $113 (CNBC); Earlier: 25 Sep, 30 Oct
Amar Harolikar · Decision Sciences & Applied AI5 of 14tigzig.com
Slide 6
Two more pressure points: the Iran war and euro-zone debt
Oil swings on each attack and each hint of a deal. France now pays more to borrow than Italy.
The Iran war, month nine Brent crude, $ a barrel
- War 28 Feb Deal 17 Jun Attacks Jan Apr Jul Oct $118
A deal signed on 17 June broke down weeks later.
October has had the most ship attacks of the war: at least 30 vessels.
Qatar is mediating new talks. The US has warned of strikes after the 3 Nov midterms.
Euro-zone debt Gap over Germany’s 10-year, bp. Pale: January. Bold: 30 Oct France 165 Italy 118 Belgium 106 Greece 102 Spain 63 Portugal 47
The gap has widened for all six. France’s has more than doubled.
Fitch cut its outlook to negative on 30 Oct. Moody’s kept a negative outlook.
More than half of France’s €3.46 trillion debt is held abroad.
My deep dive: France’s sovereign debt crisis, 25 Oct
SourcesCNBC 30 Oct, 31 Oct, 23 Oct; Fitch via Investing.com on Yahoo Finance; 10-year yields: CNBC; Brent via tigzig.com APIs; Eurostat
Amar Harolikar · Decision Sciences & Applied AI6 of 14tigzig.com
Slide 7
The builders say the cash comes later. Gartner sees AI revenue at less than half of what is needed
Gartner says over $3.5 trillion of AI infrastructure spending could be exposed to impairment by the end of 2030.
Capital spending, $bn a year 2026 plan (Alphabet, Meta: midpoint). Hatched: 2027 estimates 220 280+ Amazon 200 300+ Alphabet 140 210+ Meta 170 n/a Microsoft
AI revenue to 2030 Needed for a 12% return, and the most Gartner forecasts $8.5T Needed $3.6T Forecast
Shiller CAPE Price to 10 years of real earnings, since 1950 1960 1980 2000 2020 40 Dot-com peak 44.2 40.7
My deep dive: A correction in AI valuations is looking more likely, 31 Aug · The biggest tech build meets the bond rout, 28 Sep
SourcesCNBC 30 Oct, 28 Oct, 29 Oct; Gartner; Shiller via tigzig.com Tremor
Amar Harolikar · Decision Sciences & Applied AI7 of 14tigzig.com
Slide 8
Last quarter the five spent $1.18 for every dollar of cash their operations brought in
Their free cash flow turned negative, the first time in nine years of data. Four of the five now burn cash.
Free cash flow, all five $bn a quarter 59 2.2 -35 Q4 24 Q3 26
Share of cash flow spent % of operating cash flow 0% 50% 100% 57% 118% Q4 24 Q3 26
Free cash flow, quarter to Sep $bn Microsoft +17.6 Meta -2.1 Oracle * -5.4 Alphabet -10.2 Amazon -35.3
Free cash flow is operating cash flow less capital spending and finance lease repayments. Oracle’s quarters end a month early, so its bar is the quarter to 31 Aug.
SourcesSEC filings of Microsoft, Amazon, Alphabet, Meta, Oracle; Earlier: 28 Sep, 14 Sep
Amar Harolikar · Decision Sciences & Applied AI8 of 14tigzig.com
Slide 9
The bill runs into trillions, and more of it is now borrowed
Free cash flow can no longer pay for the build, so the builders are turning to bond markets.
OpenAI has cut its funding target to $20 billion, from $30 billion
Some investors reduced their allocations. The round values OpenAI at $1.25 trillion and is still open. CNBC and the FT, 31 Oct.
Hyperscaler capex Goldman Sachs, 23 Sep: 2026 on track. Hatched: estimates $800bn 2026 $1.2tn 2027 $1.4tn 2028
Hyperscaler bonds issued $bn, running total since 2024, J.P. Morgan. Hatched: estimates 17 2024 126 2025 405 2026 625-705 2027
The scale Data centre build to 2030 $5tn J.P. Morgan Of it, borrowed in bonds $2tn J.P. Morgan Anthropic’s commitments $800bn+ my sum, IPO filing AI debt issued in 2026* ~$500bn Goldman Sachs
- To early August. J.P. Morgan’s $5tn and $2tn are estimates.
My deep dive: The biggest tech build meets the bond rout, 28 Sep
SourcesGoldman 23 Sep, 5 Aug; J.P. Morgan AM 5 Aug; Anthropic S-1; CNBC 31 Oct
Amar Harolikar · Decision Sciences & Applied AI9 of 14tigzig.com
Slide 10
The build itself is running into limits: power, equipment and local opposition
Bernstein expects only about 35% of planned US data-centre capacity to be built
It expects the big hyperscalers to finish theirs. The shortfall is mostly newcomers. Business Insider, Oct.
How long it takes from application or order Power plants: grid hookup ~5 yrs Goldman, median High-voltage transformers 3-5 yrs Goldman Gas turbines 3+ yrs Siemens Energy
Data-centre projects since January Goldman Sachs count Cancelled or badly delayed 86 Of them, over local opposition 76
Goldman says delays push back returns, but also curb overbuilding.
In July New York paused state environmental permits for new hyperscale data centres, for up to a year.
71% of Americans oppose a data centre in their area (Gallup, March).
“Power is probably the forever bottleneck,” says Jason Tofsky of Goldman Sachs.
SourcesBernstein via Business Insider; Goldman Sachs 19 Oct; Siemens Energy via Utility Dive, Aug; New York governor, 14 Jul; Gallup via CNBC 19 Oct
Amar Harolikar · Decision Sciences & Applied AI10 of 14tigzig.com
Slide 11
Lenders want more for AI debt. CoreWeave’s default swaps are at distress levels
S&P cut Meta’s outlook to negative on 30 Oct and sees a $71 billion cash shortfall for it in 2027.
Cost to insure debt against default 5-year swaps, basis points 1,000: distress level CoreWeave, a year ago 400 approx. CoreWeave, 21 Oct 1,067 Oracle, a year ago 76 approx. Oracle, 19 Oct 295 A year ago: from CNBC’s reported change over the year. Oracle’s eased to 277 by 28 Oct, after a peak earlier in the month. Meta’s and Nvidia’s hit records in October.
CoreWeave, a neocloud 60% below its high, $ Jan May Sep $138 $55
Nebius, a neocloud 46% below its high, $ Jan May Sep $287 $154
SourcesCNBC 21 Oct, 19 Oct, 23 Oct, 28 Oct; S&P via Yahoo Finance; prices via tigzig.com APIs; Earlier: 23 Oct
Amar Harolikar · Decision Sciences & Applied AI11 of 14tigzig.com
Slide 12
US jobs: low hiring, low firing, longer searches, and pay behind prices
Unemployment is 4.2%, down from 4.5% last November. Job growth is 51K a month, up from 10K in 2025 but below 122K in 2024.
Hiring rate Near 2009-13 levels 2008 2016 2024 3.3% Layoff rate Near a 25-year low 2008 2016 2024 1.0% Time out of work Median weeks, Sep 2024 2026 11.5 Pay minus prices Behind prices since April 2024 2025 2026 -0.7 Jobs added a month 3-month average 2024 2026 51K Unemployment rate Down from 4.5% in Nov 2024 2026 4.2% Saving rate Lowest since 2022 2024 2026 3.8% Mortgages newly late 90+ days, most since 2015 2012 2018 2024 1.52% My deep dive: US jobs, September 2026, 4 Oct US housing is showing cracks, 22 Oct
SourcesBLS (hiring and layoff rates: JOLTS, Aug; 2020 spike cut off), BEA, NY Fed via tigzig.com Tremor
Amar Harolikar · Decision Sciences & Applied AI12 of 14tigzig.com
Slide 13
The other side: the builders see strong demand
Amazon Andy Jassy, CEO, 29 Oct Amazon has a “clear line of sight to attractive financial returns.” He expects AWS to grow over 40% next year.
Google Cloud Thomas Kurian, CEO, 28 Oct “We are seeing extraordinary demand and strong growth from both AI labs, capital markets, financial service institutions, and enterprises.”
Microsoft Amy Hood, CFO, 28 Oct Azure revenue grew 48% and is now Microsoft’s biggest single business. She expects 49% to 50% growth next quarter, at constant currency.
Oracle 28 Oct Expects free cash flow to turn positive in the second half of fiscal 2028. It caps new debt and equity for fiscal 2027-30 at $20 billion.
Nvidia Jensen Huang, CEO, Oct “Demand for the neoclouds is off the charts.”
Gartner 14 Oct Hyperscalers “have massive and highly diversified revenue sources that could plausibly sustain AI investments for years.”
IMF 13 Oct Its scenario: a steep AI pullback cuts world output 1.1% by 2028. Petya Koeva Brooks: not likely to mean a global recession.
SourcesCNBC 29 Oct, 28 Oct, 28 Oct, 28 Oct, 23 Oct, Gartner, Reuters
Amar Harolikar · Decision Sciences & Applied AI13 of 14tigzig.com
Slide 14
My earlier analysis
More at tigzig.com ➜ Analysis
AI cash, debt and valuations 26 OctAnthropic’s IPO filing and its must-pay deals 23 OctA crescendo of warnings on AI debt and returns 17 OctIMF warnings, and AI credit under strain 28 SepThe biggest tech build meets the bond rout 14 SepThe Fed is caught 31 AugA correction in AI valuations is looking more likely 16 JulThree warnings on the AI boom: BIS, Buffett, S&P Markets, debt and war 30 OctGold is testing $4,000 for the third time 25 OctFrance’s sovereign debt crisis 24 OctWhat kind of drawdown are we looking at this time? 19 OctForeign money is leaving Indian stocks 25 SepBond markets have started acting 12 JunMarkets are celebrating an Iran deal 8 JunIndex at a record, peak valuation, record-low sentiment 17 MayS&P at 7,400: irrational exuberance The US economy 22 OctUS housing is showing cracks 12 OctUS consumers: the pressure is building 4 OctUS jobs, September 2026 24 SepThe US job market, below the headline 10 SepMacro charts showing the stress they showed before 2007
Not investment advice. This is my analysis of public sources, for information only. Care has been taken, but errors are possible. Use of AI. Parts of this work were prepared with AI under my direction. I reviewed, edited and approved every page. The analysis, conclusions and judgment are mine.
Amar Harolikar · Decision Sciences & Applied AI14 of 14tigzig.com
Working on something similar? How I work covers the rates, the availability and what I take on.