San Francisco has $2 trillion in AI wealth and can’t fix its own city. That’s every city’s problem

The article argues that AI wealth in San Francisco and the Bay Area has not translated into broad urban prosperity, despite OpenAI and Anthropic being valued at a combined $2 trillion and 91 additional AI unicorns adding $600 billion in private-market capitalization. It highlights improving but still strained urban metrics, including San Francisco office leasing up 43% year over year to 3.4 million square feet in Q1 2026, NYC office absorption of 6.9 million square feet, and a cut in San Francisco housing-permit processing time from 605 days to 280 days. The core message is that cities have the data and technology, but governance and procurement systems remain too slow to respond in real time.

Analysis

The market implication is not “cities are recovering,” but that urban asset owners are facing a regime change from occupancy beta to utilization beta. If weekday demand remains bifurcated while permitting, curb access, and transit capacity stay slow-moving, the value pool shifts toward operators that can reprice space, storage, and services intraday rather than those dependent on static long-duration leases. That is modestly positive for high-quality office managers with optionality and mixed-use exposure, but negative for commodity office landlords tied to legacy floorplates and fixed operating costs.

CBRE and CIGI should benefit from the growing mismatch between real-time demand and municipal response because advisory, leasing, and occupancy optimization become more valuable when every basis point of utilization matters. The second-order effect is that as clients chase adaptive reuse, conversion, and portfolio repositioning, consulting and transaction activity can outperform transaction volumes alone. The risk is timing: a faster permit pipeline would be a 12-24 month tailwind, but a true governance breakthrough would compress fees only gradually, while any renewed macro slowdown would quickly re-freeze deal activity.

The contrarian point is that the crowded bearish view on urban real estate is now too simple. The durable underperformance is not in “cities” broadly but in assets and business models built for stable utilization curves; that makes the opportunity more about selection than sector direction. Near term, the best trade is to own intermediaries and urban infrastructure enablers while fading the weakest legacy office cash flows, because the market is still pricing a linear recovery when the real upside comes from a nonlinear reconfiguration of how cities monetize space.

Key Decisions for Investors

• Long CBRE vs. short a basket of legacy office REITs (e.g., SLG, BXP) over 3-6 months — CBRE benefits from advisory/repositioning demand, while office landlords remain exposed to structurally lower utilization; target 1.5-2.0x relative return if leasing momentum persists.
• Add CIGI on pullbacks for a 6-12 month horizon — the stock should capture increased demand for workplace strategy, transaction support, and adaptive reuse services; risk/reward is attractive if urban re-tenanting continues but would fade if macro volatility freezes capital markets.
• Sell downside in high-quality office names via put spreads 6-9 months out — monetize the market’s overconfidence in a smooth recovery while limiting premium outlay; best where occupancy can improve but rents and NOI lag.
• Pair long CBRE / short JLL if looking for cleaner urban-transition exposure — CBRE has more operating leverage to portfolio repositioning and global advisory; JLL is more sensitive to slower capital markets digestion.
• Avoid broad bullish exposure to office beta until permit and transit data show sustained improvement for at least 2-3 quarters; otherwise, the trade is a tactical rally in utilization, not a structural re-rating.