30-Year Treasury Yield Hits 17-Year High, Cursor Seals $60B Exit

Key Takeaways

1. US fiscal stress intensified: debt grew ~$600B in two months, and the 30-year yield hit a 17-year high of 5.31%.
2. AI is entering a “Reed’s Law” phase, with the US-China intelligence race doubling every six months, according to Raoul Pal.
3. Cursor’s $60B acquisition and OpenRouter’s ~$8B exit mark a new era of AI startup consolidation.
4. Meb Faber’s Nike study highlights valuation multiple expansion/contraction as a key risk for high-P/E stocks.
5. Active funds keep losing to benchmarks—only 13% beat over 10 years—reinforcing the case for index investing.

1. US Fiscal and Bond Market Stress

1. US national debt has grown by roughly $600 billion since July 1, and Charlie Bilello argues that the Treasury’s “debt buyback” is merely swapping old debt for new issuance, not real debt reduction, calling the “deficit reduction” narrative a lie. — via 1 2
2. The 30-year Treasury yield climbed to 5.31%, its highest level since June 2007, suggesting bond markets are “debunking” the Fed and the administration’s “low-inflation” narrative. — via 1
3. Interest payments on US federal debt now consume 3.3% of GDP, a record high; Lyn Alden warns that even with ~8% nominal private-sector growth, the deficit keeps expanding, creating a “fiscal wage-price spiral” that makes debt/GDP difficult to reduce. — via 1 2

2. AI: Race, Models, and Major Exits

1. Raoul Pal says AI is following Reed’s Law—a “squaring” of Metcalfe’s Law—producing exponential-within-exponential growth never seen in prior tech adoption curves; this means old economic frameworks will fail to describe the system. Meanwhile, the US and China are locked in an unavoidable intelligence race, with intelligence roughly doubling every six months; a single 18-month stall would mean falling three years behind, and the evolution from AGI to ASI could lead to one system dominating everything. — via 1 2
2. Cursor has completed the largest private-company acquisition ever at $60 billion, while OpenRouter reportedly exited for about $8 billion (unverified), signaling a wave of consolidation in AI developer tools and infrastructure. — via 1
3. Replit Free Mode is now powered by OpenAI GPT-5.6 Luna, making frontier-level AI available to everyone—another step toward broad consumer access. — via 1
4. An unnamed organization reportedly paused frontier RL training to meet alignment, safety, and monitoring standards (unverified); @jason comments “here we go again,” suggesting AI-safety pauses could become a recurring pattern. — via 1

3. Equity Markets and Investment Signals

1. Meb Faber argues that Nike’s decline is largely driven by multiple compression: its P/E rose from ~10x in 2009 to ~70x by 2020 and has now fallen back to ~20x. He warns that valuation expansion/contraction is a major driver of stock returns and highlights Costco’s ~50x P/E as a question mark, asking which other “normal” stocks are pricing in extreme expectations. — via 1 2 3
2. Only 27% of large-cap US equity funds beat their benchmark over the past year, and just 13% did so over ten years; Charlie Bilello cites Jack Bogle’s advice to “buy the whole market” rather than search for exceptions. — via 1
3. Energy is now the No. 1 S&P sector in 2026 after ranking 8th in 2025, a notable rotation signal for commodity and macro investors. — via 1
4. Clifford Asness defends cap-weighted indices as “reality”: they are simply the aggregate of all investors holding stocks at agreed prices, pushing back on The Economist’s claim that indices no longer reflect equity reality. — via 1