Video summary
A.I. Mania Creating Twin Bubbles In Stock Prices AND Earnings | New Harbor Financial
Main summary
Key takeaways
Finance-focused subtitle summary
Macro/market backdrop & risks
- The market is described as extremely overvalued for an extended period, with common valuation indicators (e.g., cyclically adjusted P/E, the Buffett indicator) “not mattering” so far.
- Geopolitics: the Iran conflict is noted as starting/worsening at least ~3 months prior to the recording date (June 3), raising uncertainty and concern that “something else” may appear.
- Risk framing:
- The team believes parts of AI/tech are in a valuation + earnings “bubble” phase.
- They argue distortions could eventually unwind, but it does not automatically mean an immediate market top.
- Late-stage bubble dynamics: momentum can reverse quickly when expectations shift—e.g., when analysts cut forecasts and multiples re-rate downward.
Key portfolio/hedging methodology mentioned (systematic approach)
- They use a “dashboard of indicators” rather than predictions or black-box timing.
- Indicator panel includes:
- Bullish % breadth measures for S&P 500, NASDAQ, NYSE
- Point-and-figure charting (breadth)
- Relative strength of sectors vs the rest of the market
- Moving averages and other “old school” technical tools
- Implementation rules / risk management:
- Indicators have been whipsawing (flipping bullish/negative), but equity exposure is not fully swung.
- Equity exposure capped around ~48% (framed as less aggressive than many peers).
- If bullish tilt returns, they remove index put hedges; if it flips negative again, they add similar hedges.
- Specific hedge action:
- Took off an index put: July S&P 500 7,000 put
- Sold it at a loss, treated as the “cost of insurance,” while preserving remaining value.
- Options as “forgiveness”:
- They emphasize using options hedges to manage uncertainty instead of trying to be perfectly right.
Market performance & “narrow leadership” warning signs
- Market breadth is described as poor/narrow:
- Main leaders: Energy and Technology/AI
- S&P 500 up ~11% from a roughly June 1 reference point, but broad participation is limited.
- Sector laggards:
- Financials: XLF “hovering around ~50 stubbornly”
- Healthcare/pharma: XLV described as lagging
- Semiconductors / AI momentum:
- SMH moved from about ~250–300 to >600 quickly.
- Concern: sharp vertical/rush moves historically lead to real pullbacks, not necessarily a resolution via sideways action.
“Blowoff top” / timing milestones discussed
- The broader concern: the bull cycle may end in a blowoff top (plausible endpoint, not necessarily immediate).
- Specific “milestone” expectation (Mike Preston):
- S&P 500 +500 points within a couple weeks as a signal.
- They reference an S&P breakout above 7,000, reaching around 7,500 in ~3 weeks.
- At discussion time, the market is cited around ~7,600.
- They argue the blowoff may already be underway given the speed/extent of the move.
- John Lodra adds historical context:
- A chart from NASDAQ Dorsey Wright: approaching 9 straight up weeks
- 8 consecutive up weeks is described as rare historically.
- They caution: extremes alone are not reliable timing tools—follow-through after extreme streaks has often been positive on average, so turning points may be abrupt but not guaranteed imminent.
AI/earnings valuation discussion (bubble vs fundamentals)
- Semis/AI rally is framed as largely:
- “A large part” narrative + momentum chasing, not purely justified by earnings.
- Example of sentiment sensitivity:
- NVIDIA CEO Jensen Huang comments about a “next trillion dollar company,” cited as associated with the stock moving roughly ~30% in a day.
- Valuation/earnings bubble nuance:
- Both multiples and forward earnings are described as optimistic/inflated.
- Example: Microsoft pulling back on AI/coding spending (agents/AI coding spend described as “too much,” per commentary).
Real-economy spending analog & recession uncertainty
- Economy-side claim:
- US data center construction spending jumped +28% YoY in April to a record annualized ~$50B.
- Public transportation spending was ~$49.9B, so data center spending outpaced it “for the first time in history.”
- Implication:
- It’s “hard to see” a near-term recession while this AI/spending “tsunami” persists.
- But they also argue:
- Even if spending continues, asset prices could overprice it and then correct.
- Negative scenario considered: if ~$1T (this year, as stated) only comes in at ~60% of expectations, markets would likely repric downward.
Sector rotation vs “true reset”
- The question posed: would a downturn be
- Capital rotation (money leaving AI/tech into value/other sectors), or
- A bigger correction/reset where a large portion of value “goes poof.”
- John’s response (risk-agnostic):
- The broader market isn’t cheap overall, so “relative-cheap-to-index” charts can mislead.
- They’ll rely on relative strength / breadth indicators to judge whether it’s rotation vs heavier damage.
IPO window as sentiment indicator + caution
- IPO activity is viewed as a “wind/weather” indicator, but explicitly not a timing tool.
- Mentioned IPOs/issues:
- SpaceX (target price around ~$135/share mentioned; another referenced view cites ~half that per Morningstar)
- Anthropic
- ChatGPT (implied ecosystem/AI)
- Google (massive raise; Berkshire Hathaway taking a chunk mentioned)
- SpaceX valuation discussion:
- SpaceX is cited as potentially achieving about a ~$2T market cap on first day.
- IPO outcome statistics:
- Median IPO down ~10% after one year
- Median IPO down ~40% three years
- Average cited around ~5% down after ~3 years (with wording noting averages can be influenced by winners).
- Presenter’s takeaway:
- They do not plan to buy SpaceX at IPO; they’d “buy and wait” for a better valuation.
Precious metals / crypto brief notes
- Precious metals: “no change”, floating along correction lows; requires patience.
- Bitcoin: referenced as breaking key support levels; specifics deferred to the next week.
Extracted instruments / tickers / sectors
- Indices
- S&P 500
- ETFs
- SMH (semiconductor ETF)
- XLF (financials ETF)
- XLV (healthcare/pharma ETF)
- Options
- July S&P 500 7,000 put
- Companies / equities
- NVIDIA
- Microsoft
- Google (Alphabet implied)
- Berkshire Hathaway
- SpaceX (IPO discussed; not a public ticker in the subtitles)
- Sectors
- Energy
- Technology / Semiconductors / AI
- Financials
- Healthcare / Pharmaceuticals
Key numbers and timelines
- Recording date: June 3; “5 months out of 12” gone.
- Geopolitics: Iran conflict at least ~3 months.
- S&P 500
- Up ~11% from about June 1
- Breakout above 7,000
- Target framework: ~7,500 within ~3 weeks
- Discussion-time level cited around ~7,600
- Options / hedging:
- Hedge removed: July 7,000 put (sold at a loss)
- Semiconductors:
- SMH: ~250–300 to >600 in a short time
- Historical streak stats:
- 8 consecutive up weeks: only 20 times since 1950
- Longest streak mentioned: 13 weeks
- Macro/economy:
- Data center construction spending: +28% YoY (April), annualized ~$50B
- Transportation public spending: ~$49.9B
- Scenario math (spending):
- Expected AI-related spend: ~$1T this year (US figure as stated)
- If only ~60% materializes: expect repricing down
- IPO statistics:
- Median IPO -10% after 1 year
- Median IPO -40% after 3 years
Disclosures / disclaimers
- “This is not personal financial advice…” (including the SpaceX discussion).
- IPO usage is cautioned as “bad timing indicators”.
- Modeling/tooling notes:
- Their planning/modeling tools are not perfect; assumptions matter.
- Monte Carlo approaches can be overly optimistic if built with unrealistic return assumptions.
Presenters / sources mentioned
- Adam Tagert (Thoughtful Money founder/host)
- John Lodra (New Harbor Financial)
- Mike Preston (New Harbor Financial)
- Darius Stale (referenced regarding outcomes; source for “more likely outcome”)
- NASDAQ Dorsey Wright (source of chart on market streaks)
- Morningstar (referenced regarding SpaceX pricing estimate)
- Mentions of Goldman Sachs and John Hussman (for forward return expectations referenced)
- Jensen Huang (NVIDIA CEO; used as an example)
- Kevin Worsh (referenced regarding rate-hike expectations)