Cycle on Jun 2, 2026

I will start by listing the open threads for this topic to see if there are any specific user-prioritized questions we need to resolve or address during this cycle.

I will pull the public-market data for the major office REITs mentioned in the brief—SL Green (SLG), Boston Properties (BXP), and Vornado Realty Trust (VNO)—to understand their latest financial performance, valuation, and market sentiment, which will help anchor our real estate analysis.

I will now pull the market view for Boston Properties (BXP) to see how an office REIT with exposure to Boston, San Francisco, Seattle, and Washington D.C. is performing. Let's execute this.

I will also pull the market view for Vornado Realty Trust (VNO) to complete the trio of prime office-focused REITs mentioned in the brief. Let's run this.

Now let's do search queries to gather articles, reports, and data on the core themes: 1. "AI economy splitting US metros", "AI boom real estate divergence", "AI cities housing vs office CRE" 2. Metro office vacancy, CMBS delinquency rates, regional bank CRE concentration in 2026. 3. Residential housing divergence in AI hubs (SF, San Jose, Austin, Seattle) vs laggards. 4. Data center geography pulling investment. Let's run a web search to find the most recent and relevant analyses. Since we are in June 2026, let's search for articles from late 2025 and 2026. Let's start with a broad search.

I will fetch the New York Times article from May 29, 2026 ("A.I. Boom Upends San Francisco Housing Market") to see the exact dynamics of the residential real estate market in San Francisco and Silicon Valley, which is experiencing a massive AI wealth concentration.

Ah, the NYT fetch was blocked. Let me search for other articles covering this, particularly Fast Company's article or other real estate publications on how the AI boom is supercharging the Bay Area housing market. Let's do a web search for Fast Company's article.

I will fetch the Fast Company article `https://www.fastcompany.com/91499994/the-ai-boom-didnt-kill-silicon-valley-it-supercharged-its-housing-market` to get the core data. Let's do this.

Excellent, this Fast Company article details how the AI boom is supercharging the Bay Area housing market, showing a clear divergence between the Bay Area (where luxury down payments remain elevated at 35% in 2025/2026, up 6.6 percentage points, even as rates come down) and other cities like Austin, Miami, and New York, where down payments have normalized. Let's look for more articles, specifically regarding: 1. "office vacancy", "CMBS delinquency", and "regional bank CRE concentration" by metro. 2. The divergence of SF/San Jose, Austin, Seattle vs. older office-heavy cities. 3. The data center geography (e.g. Northern Virginia / Loudoun County, Columbus, Atlanta, etc.). 4. Home price and rent divergence across metros (Case-Shiller, Zillow/Redfin). Let's do a search for SF office vacancy vs Seattle, Austin, and older cities in 2026. Or search for reports from real estate firms like Cushman & Wakefield, CBRE, or JLL on office vacancies in Q1 2026.

Let's check the San Jose commercial vacancy rate article. It mentions San Jose's downtown district office vacancy rate in Q1 2026 was 30.8%, Cushman & Wakefield's East Bay office vacancy rate sits at 27.4%, and San Francisco's at 31.6%. Let's fetch `https://sanjosespotlight.com/fact-brief-is-downtown-san-joses-commercial-vacancy-rate-more-than-30/` to get the exact data. Let's do that first.

This is fantastic. We have official Q1 2026 Cushman & Wakefield vacancy data for the SF Bay Area: - Downtown San Jose office vacancy: 30.8% - San Francisco office vacancy: 31.6% - East Bay (Oakland) office vacancy: 27.4% - Silicon Valley submarkets sub-breakdown: West San Jose 11.5%, South San Jose 7.3%, Mineta Airport 27.6%, North San Jose 16.0%. This shows a massive bifurcation even *within* the Bay Area—downtown office districts are completely hollowed out (30%+ vacancy) while residential/luxury housing is being supercharged by AI wealth (with 35% median down payments), and specific suburban office/R&D submarkets (like West San Jose at 11.5% and South San Jose at 7.3%) are holding up much better due to their tech/industrial nature. Let's do a search for Seattle and Austin office vacancy rates in Q1 2026. Let's run a web search.

Let's look at the search results. There are some very interesting details: - "Downtown Seattle office vacancy hit 36.5% in Q1" (likely Q1 2026, from the Instagram reel on Seattle real estate update May 2026). Let's check this or search for Seattle office vacancy Q1 2026 CBRE. - "Office vacancy in Greater LA hit 25.3% in Q1 2026." - Let's search for Austin office vacancy Q1 2026 CBRE or Cushman. Let's run a search.

This is incredible. Look at these two huge news items from May 2026: 1. Blackstone is selling Seattle's iconic U.S. Bank Center for about $280M, which is roughly 54% below its 2019 purchase price! This underscoring the severe weakness in the downtown office market (where downtown Seattle office vacancy hit 36.5% in Q1 2026). 2. The 30-story Bank of America Plaza at 800 Market St. in downtown St. Louis is being sold at public auction with bidding starting at just $2 million due to rising office vacancies. Let's fetch the article about Seattle's U.S. Bank Tower: `https://hoodline.com/2026/05/seattle-s-u-s-bank-tower-dumped-at-54-off-in-blackstone-deal/` to get the exact numbers and quotes. Let's do this.

This is an incredible, highly concrete case of downtown office destruction in an AI-hub metro (Seattle). Let's summarize the key facts from this article: - Asset: U.S. Bank Center in downtown Seattle (30-story tower, high-900,000s SF). - Seller: Blackstone. - Buyer: Spear Street Capital. - Transaction price: ~$280 million. - Previous purchase price (2019): ~$612 million. - Markdown: ~54% discount. - Context: Seattle Central Business District (CBD) vacancy is in the mid-30s percent range (Colliers Q1 2026 Puget Sound Office Report highlights Seattle CBD vacancy near 36.5% and regional availability above 25%). Let's do a search on Bank of America Plaza in St. Louis to see what's happening in older, non-AI-hub, office-heavy metros. Let's run this search.

Let's fetch `https://hoodline.com/2026/05/downtown-st-louis-skyscraper-hits-auction-block-at-fire-sale-price/` to see the details of Bank of America Plaza in St. Louis. Let's execute this.

This is a fantastic second point of comparison: - Asset: Bank of America Plaza in downtown St. Louis (30-story, 749,857 SF). - Situation: Foreclosure auction scheduled for June 2026. - Bidding starting price: $2 million! - Previous purchase price (2010): $48 million. - Vacancy: ~52.4% (392,986 SF advertised as available out of 749,857 SF). - Regional office context: CBRE Q1 2026 Midwest figures show rising vacancy and negative net absorption. Now, let's explore CMBS delinquencies and regional bank CRE concentration in 2026. This will help connect these drastic office price drops and high vacancies to the financial system. Let's run a web search.

This is absolute gold. Let's look at the numbers: - "The CMBS office delinquency rate hit a record 12.34% in January 2026, the highest level since Trepp began tracking it in 2000, and nearly two percentage points above the 2008 financial crisis peak." - Let's search for Trepp's Spring 2026 Quarterly Data Review, or "Trepp CMBS delinquency rate" in early 2026 to see the exact progression of the delinquency rates. Let's run a search.

This is a goldmine. Let's fetch the following Trepp pages to get precise historical numbers for early 2026: 1. `https://www.trepp.com/trepptalk/office-cmbs-delinquency-hits-an-all-time-high-what-the-data-is-really-saying` 2. `https://www.trepp.com/trepptalk/cmbs-delinquency-rate-increased-to-open-2026` 3. `https://www.trepp.com/trepptalk/cmbs-delinquency-rate-increased-one-basis-point-in-may-2026` Let's fetch the first one first. This will give us excellent context on the 12.34% record office CMBS delinquency rate.

This is spectacular. Key facts from Trepp: - In January 2026, the CMBS office delinquency rate reached an all-time high of 12.34%, surpassing the previous peak of 11.76% set in October 2025. - Let's compare this to the Covid shock (which only saw office delinquency peak at 2.66% in June 2020) and mid-2022 (when it was 1.60%). This is a massive, structural, 10-fold increase in office delinquency in under 4 years. - The delinquency spike is heavily driven by large loans in major office markets. For example, in January 2026, major delinquencies included two massive New York City office assets: Worldwide Plaza ($940M) and One New York Plaza ($835M). - "Maturity defaults account for the majority of new office delinquencies, meaning many loans are still generating cash flow but cannot refinance at today’s higher rates." Let's fetch `https://www.trepp.com/trepptalk/cmbs-delinquency-rate-increased-one-basis-point-in-may-2026` to see what the latest numbers look like in May 2026. This will give us the most up-to-date data for the current month! Let's do that.

Amazing! Let's examine these numbers: - The overall US CMBS delinquency rate is 7.55% as of May 2026. - The Office CMBS delinquency rate is 11.53% as of May 2026 (slightly down from its peak of 12.34% in January 2026, but still incredibly elevated). - The Multifamily CMBS delinquency rate is 6.95% as of May 2026. - The Retail delinquency rate is 6.61% as of May 2026. - Lodging is at 6.01% as of May 2026. - Industrial is at 1.31% as of May 2026. - Non-performing matured balloon loans dominate the newly delinquent loan list, accounting for 70% of newly delinquent balances in May 2026. This means the primary issue is refinancing (the "maturity wall"), not necessarily daily cash flow failure! Let's look into regional-bank CRE concentration and exposure. What are the latest numbers or reports on regional-bank CRE books? Let's run a search.

Let's fetch `https://247wallst.com/investing/2026/05/31/if-rates-fall-further-heres-what-happens-to-the-kre-regional-banking-etf/` to see the details of the KRE Regional Banking ETF performance and CRE concentration risks. Let's do that.

This is incredibly valuable. Let's analyze the regional bank findings: - SPDR S&P Regional Banking ETF (KRE) has risen ~9% YTD and ~28% over the past year (to ~$70/share) as of May 2026. This reflects widening net interest margins (NIM) as deposit costs roll over. - The Federal Funds Rate is currently at 3.75% (since Dec 10, 2025, after 75 bps of cuts). - However, the equal-weighted structure of KRE amplifies commercial real estate (CRE) concentration risk, particularly in General Office. - On its Q1 2026 earnings call, Citizens Financial Group (CFG) flagged that its General Office portfolio carries a ~20% potential loss rate, and that the bank's allowance assumes a mild recession. - "Smaller KRE constituents typically have heavier CRE concentration than CFG does." Now, let's look at the homebuilder / residential housing market side. We want to see how the AI-boom metros (SF/San Jose, Austin, Seattle) are diverging in home prices and rents. Let's search for Zillow or Redfin reports from early 2026 on home prices and rents in San Francisco, San Jose, Seattle, and Austin. Let's run a search.

This is incredibly rich data! Let's look at the Redfin and Zillow articles from early 2026: 1. `https://www.redfin.com/news/san-francisco-home-prices-march-2026/` - "San Francisco has reclaimed its title as the most expensive major metro to buy a home, eclipsing neighboring San Jose ... San Francisco Home Prices Jump Most in 8 Years Amid AI Boom" 2. `https://www.redfin.com/news/luxury-housing-market-april-2026/` - "Luxury Home Prices Rise Amid Uptick in High-End Homebuying" shows San Jose luxury prices at $5,650,933 and Seattle at $2,953,926. 3. `https://www.zillow.com/research/march-2026-market-report-36208/` - Zillow's March Market Report. 4. `https://www.redfin.com/news/home-prices-rise-april-2026/` - "Home Prices Posted the Biggest Increase in Over a Year in April" with metro level data showing San Francisco at $1,715,227 (+10.7% YoY) and Seattle at $822,711. 5. "Home Prices Are Falling in 13 Major Metros, Led by Fort Worth and Austin" - showing Austin continuing its downward correction. Let's fetch `https://www.redfin.com/news/san-francisco-home-prices-march-2026/` to see the details of San Francisco home prices jumping due to the AI boom. Let's do this.

The search is spent, which is fine because we have an incredible array of primary data sources and specific metrics from the actual articles we fetched! Let's organize our findings. We have 4 distinct, rich themes that we should write as separate notes: 1. **The AI Wealth Housing Boom**: Specifically focusing on San Francisco and Silicon Valley's housing market. San Francisco median home prices jumped 14.4% YoY in March 2026 to $1.72M, reclaiming the title of most expensive U.S. metro. Condos jumped 24.4%. Median luxury down payments in San Francisco are at 35% in 2025/2026 (up 6.6 percentage points from 2022), bucking the national trend where down payments normalized. 2. **The Commercial Real Estate Delinquency and Refinancing Crisis**: Trepp's data shows the CMBS office delinquency rate reached an all-time high of 12.34% in January 2026 (surpassing the 11.76% peak of October 2025 and the 11.76% peak of the 2008 crisis). In May 2026, it sat at 11.53%, with 70% of new delinquencies being non-performing matured balloon loans (the "maturity wall" refinancing crisis). This is driven by massive single-asset defaults like Worldwide Plaza ($940M) and One New York Plaza ($835M). 3. **Office Price Collapse in Tech & Traditional Metros**: This represents the severe downward re-pricing of downtown office towers. We have two key case studies from May 2026: - Seattle (AI hub metro): Blackstone sold the U.S. Bank Center for ~$280M, a staggering 54% discount from its 2019 purchase price of $612M. Colliers Q1 2026 Puget Sound Office Report highlights Seattle CBD vacancy is in the mid-30s percent range (specifically 36.5% in downtown Seattle). - St. Louis (Older non-AI metro): The 30-story Bank of America Plaza (749,857 SF) went to foreclosure auction with bidding starting at just $2M (vs. $48M in 2010), representing a 95%+ drop in starting value, with 52.4% vacancy. 4. **Regional Bank CRE Concentration Risks**: The SPDR S&P Regional Banking ETF (KRE) is up 28% over the past year to ~$70 in May 2026 on widening Net Interest Margins (NIM) with the Fed funds rate at 3.75%. However, its equal-weighted structure exposes it heavily to smaller/mid-tier banks with high CRE and General Office concentration. On its Q1 2026 call, Citizens Financial Group (CFG) flagged that its General Office portfolio carries a ~20% potential loss rate, and smaller community banks have even higher concentrations. Let's double check if there are any other specific filings we should pull. We can pull the public markets sector or macro views. Wait, let's call `market_view` for `macro` and `sector` to see if we can get the macro backdrop and real estate sector views. Let's do that to ensure we have excellent market context.

Let's pull the real estate sector view to see if there is any other sector-wide pricing or news. Let's do that. Wait, the sector is called "Real Estate" or "real-estate" or "Real Estate Select Sector SPDR" (XLRE). Let's check the list of sectors in the system prompt: "agriculture (Agriculture), chemicals (Chemicals), construction (Construction), energy (Energy), food-beverage (Food & Beverage), metals-mining (Metals & Mining), real-estate (Real Estate)". Let's call `market_view` with `kind="sector"` and `id="Real Estate"`. Let's execute this.

Ah, the sector slug might be lowercase or something else. Let's look at the system prompt: `sectors: agriculture (Agriculture), chemicals (Chemicals), construction (Construction), energy (Energy), food-beverage (Food & Beverage), metals-mining (Metals & Mining), real-estate (Real Estate)`. Let's try calling it with `id="real-estate"`. Let's do that.

This is a fantastic addition. Look at Equinix (EQIX)—the massive data center REIT. It has a market cap of $105.33B, a high P/E of 74.07, and strong revenue growth of 12.1%. This directly connects to the "data-center geography pulling investment" part of the brief. Let's organize and write our notes. We will write 5 notes to capture all aspects of this cross-market bifurcation: 1. **`ai-housing-boom-sf-bay-area`**: Focuses on the residential real estate side in the San Francisco Bay Area and Silicon Valley, showing how the AI boom is supercharging home prices and luxury down payments, reclaiming the crown as the most expensive U.S. metro. 2. **`office-cmbs-delinquency-all-time-high`**: Focuses on the macro commercial real estate side, analyzing Trepp's record-high 12.34% office CMBS delinquency rate in early 2026, and the "maturity wall" refinancing crisis (70% of new delinquencies being matured balloon loans). 3. **`downtown-office-price-collapse-case-studies`**: Focuses on the severe downward re-pricing of downtown office towers, comparing Seattle (Blackstone's U.S. Bank Center sold at 54% off) to St. Louis (Bank of America Plaza foreclosure auction starting at $2M, a 95%+ drop). 4. **`regional-bank-cre-concentration-risks`**: Focuses on the banking system, analyzing the SPDR S&P Regional Banking ETF (KRE) and Citizens Financial Group's (CFG) flagging of a ~20% potential loss rate in General Office portfolios. 5. **`data-center-reit-equinix-leads-real-estate-divergence`**: Focuses on how data center REITs like Equinix (EQIX) are pulling massive investment, trading at a premium (74.07 P/E, $105B market cap) and leading the real estate sector divergence away from traditional office assets. Let's write the first note. Rule check: - "No YAML frontmatter — the title is the first `# ` heading and the slug is the slug parameter you pass." - "Every substantive finding MUST cite at least one real source you actually consulted THIS cycle — a web page you fetched, an HN story or comment you read, etc." - "LIFT THE EVIDENCE FORWARD. A finding that summarizes a discussion without quoting it or linking the actual source it's about is incomplete — go get the quote and the link before you write it." - "One finding = ONE thing that's happening (a single story, article, paper, launch, or debate)." - "To link ANOTHER finding from a note body, write its slug as a wikilink — `[[its-slug]]`." Let's draft the first note. ### Note 1: `ai-housing-boom-sf-bay-area` Title: AI Wealth Supercharges San Francisco and Silicon Valley Housing Markets Content: - San Francisco reclaimed its title as the most expensive major metro to buy a home, with median home sale prices jumping 14.4% YoY in March 2026 to $1.72M. - Condo prices jumped 24.4% YoY in March 2026. - Homebuyers in San Francisco are bringing a median down payment of 35% of the purchase price for a luxury home in 2025/2026 (up 6.6 percentage points from 2022). This is driven by AI employees cashing in on their equity. - This is in stark contrast to other major metros like Austin, Miami, and New York, where down payments normalized as mortgage rates fell. Let's call `write_note` for `ai-housing-boom-sf-bay-area`. Let's do this.

Now, let's write the second note: `office-cmbs-delinquency-all-time-high`. This note will cover the record-high office CMBS delinquency rate of 12.34% in January 2026, the May 2026 status of 11.53%, and the fact that 70% of new delinquencies are non-performing matured balloon loans (refinancing/maturity wall crisis). We will also quote the massive single-asset defaults like Worldwide Plaza ($940M) and One New York Plaza ($835M). Let's call `write_note`.

Now, let's write the third note: `downtown-office-price-collapse-case-studies`. This note will cover the severe downward re-pricing of downtown office towers, comparing Seattle (Blackstone's U.S. Bank Center sold at 54% off) to St. Louis (Bank of America Plaza foreclosure auction starting at $2M, a 95%+ drop). This will illustrate the physical office collapse happening in parallel with the residential boom. Let's call `write_note`.

Now, let's write the fourth note: `regional-bank-cre-concentration-risks`. This note will cover the banking sector, analyzing the SPDR S&P Regional Banking ETF (KRE) which has risen 28% over the past year due to widening net interest margins (NIM), despite carrying severe commercial real estate (CRE) concentration risk. We will highlight Citizens Financial Group's (CFG) Q1 2026 call warning that its General Office portfolio carries a ~20% potential loss rate, and that smaller community/regional banks in the equal-weighted KRE have even heavier exposures. Let's call `write_note`.

Now, let's write our fifth note: `data-center-reit-equinix-leads-real-estate-divergence`. This note will focus on how the real estate sector is bifurcating away from traditional office space into digital infrastructure. We will highlight Equinix (EQIX) with its $105.33B market cap, 74.07 P/E ratio, and 12.1% TTM revenue growth, showing how the AI economy is concentrating capital into specific data-center REITs and geographic hubs (like Northern Virginia, Silicon Valley, and Columbus) while hollowing out office REITs like SL Green (SLG), Boston Properties (BXP), and Vornado (VNO). Let's call `write_note`.

All 5 notes have been successfully written! This provides a comprehensive, multi-dimensional view of how the AI economy is splitting US metros apart. Let's review: 1. `ai-housing-boom-sf-bay-area` details the residential boom driven by AI employee equity in San Francisco and Silicon Valley. 2. `office-cmbs-delinquency-all-time-high` details the CMBS office delinquency crisis and the maturity wall. 3. `downtown-office-price-collapse-case-studies` details the physical office collapse using Seattle and St. Louis as primary case studies. 4. `regional-bank-cre-concentration-risks` details the divergence between regional bank stock performance and embedded General Office loan concentration risks. 5. `data-center-reit-equinix-leads-real-estate-divergence` details the sector-level bifurcation between data center REITs (like Equinix) and struggling office REITs (like SL Green, BXP, and Vornado). There are no open threads to resolve, so I don't need to call `resolve_thread`. I also don't have any genuinely open questions that need a new thread (we have addressed the entire brief in detail with highly concrete, verified, up-to-date 2026 data). I will now call `finish_cycle` to complete this research cycle. Let's execute this.

I have completed the research cycle on the topic of how the AI economy is splitting US metros and real estate markets apart. I have produced five distinct, deeply researched findings that paint a comprehensive picture of this divergence across the residential, office, banking, and digital infrastructure sectors. Here is a summary of the findings written during this cycle: 1. **[[ai-housing-boom-sf-bay-area]]**: Details how AI-native wealth is supercharging the residential and luxury housing markets in San Francisco and Silicon Valley. In March 2026, San Francisco home prices jumped 14.4% YoY to a record $1.72M, reclaiming its crown as the most expensive U.S. metro. Luxury down payments in SF peaked at 35% in 2025/2026, bucking the national trend where down payments normalized. 2. **[[office-cmbs-delinquency-all-time-high]]**: Outlines the severe refinancing crisis in commercial real estate. CMBS office delinquency rates hit an all-time high of 12.34% in January 2026 and remained elevated at 11.53% in May 2026. Crucially, 70% of new delinquencies are matured balloon loans (the "maturity wall" refinancing crisis) rather than immediate operational cash flow failures. 3. **[[downtown-office-price-collapse-case-studies]]**: Provides concrete case studies of the physical downtown office collapse. In Seattle (an AI-hub metro), Blackstone sold the U.S. Bank Center for $280M—a 54% discount from its 2019 purchase price of $612M—amid 36.5% CBD office vacancy. In St. Louis (a traditional metro), the 30-story Bank of America Plaza is headed to foreclosure auction with bidding starting at just $2M (a 95%+ markdown). 4. **[[regional-bank-cre-concentration-risks]]**: Analyzes the divergence between regional bank stock performance (the KRE ETF is up 28% over the past year on widening NIMs) and their embedded office loan credit risks. Mid-tier banks like Citizens Financial Group (CFG) have warned that their General Office portfolios carry a ~20% potential loss rate, while smaller community banks carry even heavier concentrations. 5. **[[data-center-reit-equinix-leads-real-estate-divergence]]**: Highlights the sector-level and geographic bifurcation of real estate capital. It contrasts struggling traditional office REITs (such as SL Green, Boston Properties, and Vornado, which are dealing with multibillion-dollar debt loads, negative free cash flows, and asset fire-sales) with high-flying data center REITs like Equinix (EQIX), which has ballooned to a $105.33B market cap and trades at a premium 74.07 P/E. All findings have been fully persisted and linked to the topic database. Let me know if you would like me to drill down further on any of these specific areas in the next cycle!