Video summary

I'm Changing How I Invest My Money Because of AI

Main summary

Key takeaways

Finance

Finance-focused summary (AI-driven portfolio changes)

Core issue raised: AI concentration + valuation risk in the S&P 500

  • The speaker argues their prior strategy—buying a low-cost, passive S&P 500 index fund—has historically worked well.
  • They claim the S&P 500 has returned “over 10% per year” on average (approximate; no specific start/end period given).
  • After discussing a prior “AI bubble,” they argue the S&P 500 has become over-concentrated in large AI-exposed firms:
    • For every $1 invested in the S&P 500, 40 cents is allocated to just 10 companies.
    • Those top 10 companies are:
      • Nvidia (NVDA)
      • Microsoft (MSFT)
      • Apple (AAPL)
      • Alphabet (GOOGL)
      • Amazon (AMZN)
      • Broadcom (AVGO)
      • Meta (META)
      • Tesla (TSLA)
      • Berkshire Hathaway (BRK.B)
      • JP Morgan (JPM)
    • They state these top companies comprise ~40% of the entire index (as-of filming).
    • They describe Nvidia as receiving ~7–8 cents per dollar due to market-cap weighting.

Valuation concern

  • They argue that for these AI companies’ current valuations to be justified, they would need about $2 trillion in revenue.
  • They add this is more than the combined 2024 revenue of:
    • Nvidia, Microsoft, Apple, Alphabet, Amazon, and Meta
  • Their concern is that earnings expectations are:
    • future-driven, and
    • tightly linked to one technology theme (AI).

Funding/risk angle

  • They claim these companies are financing AI spending with “a ton of debt.”
  • They cite a Sam Altman / OpenAI example:
    • Altman reportedly willing to commit $1.4 trillion to AI infrastructure spending
    • while OpenAI has ~$13–20B in yearly revenue.

Mechanistic explanation of the “feedback loop” (passive + concentration)

  • Larger companies attract more passive inflows → higher prices → bigger index weights → even more passive dollars, reinforcing concentration.

Macro/regime caution

  • They reference a claim from Deutsche Bank that without AI spending, the US economy “would already be in a recession.”

What they changed (explicit portfolio construction actions)

1) Reduced exposure to the market-cap-weighted S&P 500 (but kept most stock exposure there)

  • They state they are putting less money into the S&P 500, but keeping the majority of their stock allocation in the regular market-cap-weighted version.

2) Considered/evaluated equal-weighting to reduce AI concentration risk

  • They discuss an equal-weighted S&P 500 index fund:
    • Top-10 concentration drops to ~2% because each company is treated equally.
  • Downside: a “negative momentum approach.”
    • They explain it sells companies after they “take off” and buys laggards to maintain equal weights.
    • This implies higher turnover → higher costs that can “eat into your gains.”
  • Net takeaway: less concentration benefit, more trading-cost risk; therefore they don’t fully replace market-cap weighting.

3) Added global diversification via a single global ETF/fund approach

  • They argue leadership shifts across history and the S&P 500 alone misses major non-US businesses.
  • They recommend/describe:
    • On Trading 212, ticker VWRP as a global stock market fund.
    • Coverage claim: ~3,700–3,800 companies across 45+ developed and emerging countries (including US/UK/Europe/Japan and also China/India).
    • They note holdings are still US-dominated, but global funds rebalance automatically as country performance changes.
    • Expense ratio stated: 0.19% (about $1.90 per $1,000 annually).
  • A US-based alternative suggested:
    • Fidelity international fund (ticker: FSPSX).

4) Added a “zone-based” stock selection framework

They describe a portfolio mindset divided into four categories:

  • Crowded zone (avoid / de-emphasize)

    • Large market cap, expensive valuation
    • Examples: Nvidia, Tesla, Meta

  • Defensive zone (relative preference / value-like quality)

    • Examples: McDonald’s, Walmart, Coca-Cola
    • Rationale: predictable demand/earnings and cash generation

  • Speculative zone (avoid; “gambling”)

    • Examples: Beyond Meat, Peloton, AMC (noted as “during hype phases”)
    • Rationale: valuations far above current revenue; momentum-driven

  • Overlooked zone (their “AI opportunity” thesis)

    • Thesis: many companies are betting on building the best AI model, but value may shift to users of AI tech—especially when models become interchangeable.
    • They cite an observed pattern: startups/smaller firms allowing users to choose among Claude, ChatGPT, Gemini.
    • They claim this could lead to AI price wars that harm heavily capitalized incumbents.
    • Therefore they prefer small/mid-cap firms that apply AI cheaply without “billions in debt.”
    • Action: they say they’re investing in small and mid-cap funds and also actively looking for AI startups.

5) Increased gold allocation as an AI-era “instability hedge”

They frame gold as benefiting from:

  • central bank demand, and
  • regulatory/asset-class treatment.

Key claims mentioned:

  • China gold purchases reportedly rising sharply up to 2025 (chart mentioned; no exact tonnage).
  • A “gold corridor” and BRICS usage with Brazil, Russia, India, South Africa.
  • Claim: gold overtook US Treasuries as the largest foreign reserve asset held by central banks since 1996.
  • Basel III change: gold reclassified as a “Tier 1 asset” (as of 2025), letting banks treat it like cash/US Treasuries on balance sheets.
  • They cite Bank of America guidelines suggesting gold reserves should rise from ~20% to ~30% (roughly +10% allocation).
  • Speculative forecast: gold could be double the current price in 5 years (“not unreasonable”).

Implementation they describe:

  • Buy physical gold (long-term).
  • Dollar-cost average into iShares Physical Gold ETF on Trading 212 (ticker not included in subtitles).

6) Increased cash reserves for optionality and drawdown protection

  • They emphasize an emergency fund/cash buffer to avoid forced selling during downturns.
  • They cite Warren Buffett cash levels:
    • As of Q1 2025, $347.7B in cash and cash equivalents.
  • Their implied action:
    • Hold slightly more cash so market crashes don’t harm life and to buy opportunities during downturns.
  • They explicitly state: “cash is king during a stock market crash.”

Explicit recommendations / cautions summarized

  • Do not rely solely on passive market-cap-weighted S&P 500 due to AI concentration and valuation feedback-loop risk.
  • Equal weighting reduces concentration but may increase costs from rebalancing/trading (negative momentum risk).
  • Diversify globally to reduce “one-market” risk and capture non-US growth (examples provided: TSMC, Samsung, Toyota, Tencent, AstraZeneca, HSBC).
  • Favor small/mid-cap “AI overlooked” opportunities rather than crowded mega-cap AI winners, based on the idea that model-building dominance may matter less than AI application/usage when models become commoditized.
  • Add gold as a strategic hedge (not an all-in bet).
  • Maintain adequate cash reserves to avoid selling at the worst time.

Methodology / frameworks mentioned

S&P 500 weighting comparison

  • Market-cap weighted: winners get larger weights; no need for frequent trimming.
  • Equal weighted: trims rising outperformers and buys laggards to maintain equal exposure.

“Four zone” stock framework

  • Crowded: expensive mega-caps
  • Defensive: quality/value-like
  • Speculative: hype/momentum
  • Overlooked: small/mid-cap AI adopters (bet on cheaper application vs expensive model arms race)

Global diversification thesis

  • “Global cake” concept: overall world market stays constant while country “layers” change; a global fund rebalances automatically.

Key tickers / instruments mentioned

  • S&P 500 index / S&P 500 index fund (general)
  • Nvidia (NVDA)
  • Microsoft (MSFT)
  • Apple (AAPL)
  • Alphabet (GOOGL)
  • Amazon (AMZN)
  • Broadcom (AVGO)
  • Meta (META)
  • Tesla (TSLA)
  • Berkshire Hathaway (BRK.B)
  • JP Morgan (JPM)
  • McDonald’s
  • Walmart
  • Coca-Cola
  • Beyond Meat
  • Peloton
  • AMC
  • Global ETF/fund: VWRP
  • US-based alternative: Fidelity international fund (FSPSX)
  • Gold ETF: iShares Physical Gold ETF (ticker not specified)
  • Treasuries (referenced conceptually; no ticker)
  • Cash / cash equivalents (referenced)

Disclosures / disclaimers

  • No explicit “not financial advice” disclaimer appears in the provided subtitles.
  • A sponsorship disclosure is mentioned:
    • Trading 212 agreed to sponsor a segment and offers free fractional shares.

Presenters / sources mentioned

  • Speaker/creator: not explicitly named in the subtitles (described as the speaker’s personal investing changes)
  • Sam Altman (OpenAI) as an example of AI infrastructure spending
  • OpenAI (revenue referenced)
  • Deutsche Bank (argument about AI spending avoiding recession)
  • Warren Buffett (cash level and value-investing behavior cited)
  • Bank of America (gold reserve allocation guidance)
  • Trading 212 (broker/platform; sponsorship)
  • Fidelity (international fund recommendation)
  • iShares (gold ETF issuer)

Original video