Billionaire Son Bets 66% of $18B Fortune on 1 Stock
Billionaire Masayoshi Son Bets $12 Billion on Intel: What You Need to Know
TL;DR: SoftBank’s founder Masayoshi Son put 67% of his entire US stock portfolio into Intel — a massive $12 billion bet that the struggling chip giant is about to stage a historic comeback.
Quick Read: The Big Picture
- SoftBank (led by famous investor Masayoshi "Masa" Son) disclosed on August 14, 2026 that it owns 86.9 million shares of Intel worth $12.14 billion
- This represents 66.81% of SoftBank’s entire US stock portfolio — an extremely concentrated bet
- The other holdings (Symbotic at ~10% and T-Mobile at ~9%) are tiny in comparison
- Intel just reported strong earnings (revenue up 25%, profits nearly double expectations)
- But Intel’s stock dropped 27% between June 30 and the filing date
- Regular investors should understand the thesis — but NOT copy the position size
What Exactly Did Son Buy?
The Portfolio Breakdown (as of June 30, 2026)
| Company | Ticker | % of Portfolio | Value |
|---|---|---|---|
| Intel | INTC | 66.81% | $12.14 billion |
| Symbotic | SYM | 9.85% | $1.79 billion |
| T-Mobile US | TMUS | 9.23% | $1.68 billion |
| Everything else (fintech bets like Klarna, Chime, Nu Holdings) | — | ~14% | ~$2.5 billion |
What’s a 13F Filing?
Think of it like a quarterly report card for big investment firms. Any firm managing over $100 million must tell the SEC (US securities regulator) what stocks they own. These become public 45 days after each quarter ends. So this filing shows what SoftBank owned on June 30, 2026, but we only learned about it on August 14, 2026.
Why Is Son Betting So Big on Intel?
1. Intel’s Business Is Actually Improving
Q2 FY2026 Results (reported July 23, 2026):
- Revenue: $16.13 billion — up 25.4% year-over-year
- Profit (non-GAAP EPS): $0.42 per share — nearly DOUBLE the $0.22 estimate
- Data Center & AI Revenue: $6.26 billion — up 59% year-over-year
- 7th straight quarter beating expectations
- CEO Lip-Bu Tan called it "the strongest revenue growth in more than 15 years"
2. They’re Building the Physical Infrastructure for AI
- Capital expenditures (capex) guided above $20 billion for 2026 — and even higher for 2027
- This means Intel is building factories, buying equipment, and creating the physical foundation that AI runs on
- Government support (CHIPS Act) helps fund this buildout
- Son has a history of betting on "critical asset" companies — things the world needs and can’t easily replace
3. He’s Been Building This Position for a While
- Q3 2025: SoftBank committed $2 billion as a strategic investor in Intel
- Same time: NVIDIA invested $5 billion in Intel
- This 13F shows: Son doubled down at scale — buying more while the stock was weak, not selling into strength
Important Context: The Timing Was Painful
The Ouch Factor
- June 30, 2026 (filing date): Intel stock at $139.63
- August 14, 2026 (filing public): Intel stock at $102.50
- Drop: 26.59% in 45 days
- SoftBank’s $12.14B position was underwater by ~$3.2 billion on paper when the filing became public
- But: Intel is still up 329% over the past year
What Analysts Think
- Consensus price target: $114.88 (about 12% above the August 14 price)
- Forward P/E ratio: ~81x — very expensive, meaning the stock price already assumes near-perfect execution
- The bet only works if: Intel’s foundry business (making chips for other companies) succeeds, not just its own CPU sales
What This Means for Regular Investors
DO: Follow the Logic, Not the Leverage
| Son’s Approach | What You Should Do |
|---|---|
| 67% of portfolio in ONE stock | Never put 67% in one stock |
| Billions to deploy, can handle volatility | Size positions to your risk tolerance |
| Long-term strategic bet on AI infrastructure | Invest in Intel if you believe the thesis — at a rational weight (e.g., 2-5% of portfolio) |
| Has inside access, strategic partnerships | Rely on public info + your own research |
The Sensible Takeaway
"Following Son’s conviction is defensible. Following Son’s position size is not."
- Intel is investable on its own merits at a normal portfolio weight
- The turnaround story has real fundamentals behind it
- But concentration risk (putting all eggs in one basket) is dangerous for retirement-focused investors
The Other Holdings: Not Just Filler
Symbotic (SYM) — 9.85% / $1.79B
- Warehouse automation + AI — robots that run fulfillment centers
- 35% owned by SoftBank via a joint venture called Exol
- $11 billion contract targeting a $500B+ market
- Ties into the physical AI infrastructure theme (robots need chips, data centers, connectivity)
T-Mobile (TMUS) — 9.23% / $1.68B
- Legacy from the Sprint merger (SoftBank owned Sprint)
- $18.2 billion authorized for stockholder returns in 2026 (buybacks + dividends)
- Steady cash flow generator — the "boring but reliable" piece
Summary: Key Takeaways
- Masayoshi Son made a $12.14B concentrated bet on Intel — 67% of SoftBank’s US portfolio
- Intel’s fundamentals are improving fast — 25% revenue growth, 59% data center/AI growth, 7 straight beats
- Massive capex ($20B+) signals Intel is building the physical backbone of AI — exactly Son’s sweet spot
- Stock dropped 27% after the quarter ended — position was underwater when filing went public
- Retail investors: Learn from the thesis (AI infrastructure, foundry turnaround, government support), not the sizing
- Symbotic and T-Mobile provide smaller but thematic exposure to AI automation and connectivity
FAQ: Your Questions Answered
1. Who is Masayoshi Son and why should I care?
Masayoshi "Masa" Son is the founder of SoftBank Group, a Japanese multinational conglomerate. He’s famous for:
- Early investment in Alibaba (turned $20M into ~$100B+)
- Creating the Vision Fund ($100B+ tech investment fund)
- Big bets on ARM, NVIDIA, Uber, WeWork (mixed results)
- When he concentrates this heavily, Wall Street pays attention
2. What is a "foundry" and why does it matter for Intel?
A foundry is a factory that manufactures chips designed by other companies. Think of it like: Apple designs the iPhone chip → TSMC (Taiwan) manufactures it.
- Intel is trying to become a major US foundry (competing with TSMC)
- This is strategically vital for US national security (CHIPS Act funding)
- If Intel succeeds here, it becomes a toll booth for the entire semiconductor industry
3. Why did Intel stock drop 27% if earnings were great?
Several reasons:
- "Sell the news" — stock had already run up 329% in a year
- Guidance for Q3 was modest ($15.8B–$16.8B revenue)
- Forward P/E of 81x means perfection is priced in — any hiccup hurts
- Macro fears (recession, AI spending slowdown) hit all semiconductors
4. Should I buy Intel stock now?
Only you can decide (this isn’t financial advice). Consider:
- Pros: Turnaround underway, AI tailwinds, government support, cheap vs. history
- Cons: Execution risk huge, foundry unproven, valuation demanding, competition fierce
- Middle ground: Small starter position, dollar-cost average, define your exit criteria before buying
5. What’s the difference between GAAP and non-GAAP earnings?
- GAAP = Official accounting rules (includes one-time charges, stock comp, restructuring)
- Non-GAAP = Company’s "adjusted" view (excludes those items to show "core" performance)
- Intel reported $0.42 non-GAAP EPS vs. $0.22 estimate — but GAAP EPS was likely lower
- Always check both — non-GAAP can be useful but sometimes hides real costs
Final Thought
Big investors make big, concentrated bets because they have to — they manage too much money to diversify like you can.
Your advantage as a small investor? You can diversify. You can size positions sensibly. You can sleep at night.
Take the insight (Intel’s AI infrastructure buildout is real and massive). Leave the leverage (67% in one stock) to the billionaires.
Article based on SoftBank’s 13F filing disclosed August 14, 2026, covering holdings as of June 30, 2026. All figures sourced from SEC filings and company reports. This article is for informational purposes only and does not constitute investment advice.