Cycle on Jul 27, 2026

I will list the threads for this topic to understand what open questions or watches have been prioritized.

I will read the existing notes on the topic to see what has been documented so far and identify areas to expand or update.

I see that the topic brief focuses on how the AI economy is splitting US metros apart. The prior findings cover: 1. `ai-housing-boom-sf-bay-area` 2. `office-vacancy-metro-divergence-sf-seattle-austin` 3. `downtown-office-price-collapse-case-studies` 4. `us-housing-market-regional-fracture-case-shiller` 5. `data-center-reit-equinix-leads-real-estate-divergence` 6. `office-cmbs-delinquency-all-time-high` 7. `regional-bank-cre-concentration-risks` Let's look at the current date: **July 27, 2026**. Many Q2 2026 corporate earnings releases and real estate reports are coming out or have just come out (since Q2 ended June 30, 2026). Let's check the public-markets data for key office REITs mentioned in the brief: - Boston Properties (BXP) - SL Green (SLG) - Vornado Realty Trust (VNO) And data-center REITs: - Equinix (EQIX) - Digital Realty (DLR) We should also check the macro view and sector views (Real Estate). Let's call `market_view` for regional banks, office REITs, and the general real estate sector to ground our numbers in real public-markets data. Let's start with `market_view` for BXP, SLG, VNO, EQIX, and DLR, as well as the macro backdrop and Real Estate sector. I will call `market_view` for BXP, SLG, VNO, EQIX, and the Real Estate sector. I'll do this in sequence. Let's check the macro backdrop first and then some tickers. Let's make a call to `market_view` with `kind="macro"`.

Now let's check the Real Estate sector and some specific tickers: BXP, SLG, VNO, EQIX, DLR. First, the real estate sector.

Now let's check the specific office REIT tickers: BXP, SLG, VNO. Let's call `market_view` for BXP (Boston Properties).

Interesting! Boston Properties (BXP) released or is about to release Q2 earnings, but in Q1 2026 (ended March 31, 2026), it reported a beat of $0.55 vs $0.28. Let's also pull SL Green (SLG) and Vornado Realty Trust (VNO) `market_view` to see what we can find. Also, the news on SLG mentions "Q2 FFO Hits 1.45 Per Share" on July 25, 2026. Let's pull SLG's view.

Let's pull Vornado Realty Trust (VNO) as well.

Let's search for recent news and data on how the AI economy is splitting US metros apart, focusing on commercial real estate (CRE) fundamentals, office vacancy rates, home-price and rent divergence, migration, job postings, and data centers. Let's first search for "office vacancy Q2 2026 metro divergence" or similar queries to find out what is happening in SF, Seattle, Austin, Manhattan, etc. We should search for "Q2 2026 office vacancy rates" and "San Francisco office vacancy Q2 2026". Let's run a `web_search` for this.

Let's look at the Kidder Mathews Q2 2026 San Francisco Office Market Report: - Total office vacancy rate was **27.2%** at the end of 2Q 2026, down 440 basis points year-over-year from 31.6% in 2Q 2025. - Direct vacancy stood at **24.7%**, also down from 27.1% a year earlier. - Net absorption in Q2 2026 was **829,597 SF** (or CBRE reports 963,980 SF with overall vacancy at 29.2% depending on the boundaries/definitions, but both show a strong positive net absorption and declining vacancy). - Significant lease transactions in Q2 2026 include: - **together.ai** leasing **149,000 SF** at 2 Henry Adams St in Showplace Square in April 2026 (Landlord: DWS Group) - **Harvey AI** leasing **62,000 SF** at 201 Third St in South Financial District in April 2026 (Landlord: Kilroy Realty Corporation) - **Assort Health** leasing **62,000 SF** at One Market Plaza in South Financial District in April 2026 (Landlord: Blackstone) This directly supports the thesis of AI companies driving the San Francisco office recovery! Let's fetch the Kidder Mathews page to get exact quotes and details. Let's do `web_fetch` on `https://kidder.com/market-reports/san-francisco-office-market-report/`. This is a rich source.

This is an outstanding source with very clear evidence. Let's list what we can write or update. We have an existing note: `office-vacancy-metro-divergence-sf-seattle-austin` ("Tech Metros Diverge: San Francisco Office Rebounds via AI While Seattle and Austin Stabilize"). Let's see what is happening in Seattle and Austin in Q2 2026. Let's do a `web_search` for "Seattle office vacancy Q2 2026" and "Austin office vacancy Q2 2026". We want to see if the divergence is widening. Let's do that now.

Let's fetch `https://kidder.com/market-reports/seattle-office-market-report/` to get Seattle's Q2 2026 data. This will provide a solid contrast to San Francisco's declining vacancy and strong positive net absorption. Let's fetch it.

Ah, Firecrawl timed out. Let's try to search specifically for Seattle's overall vacancy and absorption numbers for Q2 2026. Let's run a `web_search` for `"Seattle Office Market Report" "Q2 2026" "Kidder"` to see if we can get details from the description or another source. Or let's search `"Seattle" Q2 2026 office vacancy "Kidder Mathews"`. Let's do that.

Okay, the search results show: - Puget Sound office vacancy reached **23.4%** at the end of Q2 2026, a 30 basis point increase from the prior quarter, according to Kidder Mathews. - Seattle Office Rents slipped 1.0% YoY to $32.76 PSF/Yr in 2Q 2026. - Seattle CBD vacancy remains elevated at **34.7%** (as of Q1/Q2 2026). Let's check Austin's Q2 2026 office vacancy. Let's do a `web_search` for `"Austin" "office vacancy" Q2 2026 "Savills" OR "CBRE" OR "Cushman"`. This will give us the third comparison.

Let's search for Austin's Q2 2026 office vacancy. Let's try a query like `"Austin" "office vacancy" "Q2 2026"` or `"Austin" Q2 2026 office market report` to see if we can get a specific figure. Let's do a `web_search`.

Let's see: Savills reports that for Austin in Q2 2026: - "Overall availability in Austin tightened further in Q2 2026, declining to **24.1%**, down 350 basis points (bps) from the same time last year." (Savills USA, July 2026) - Let's fetch the Savills report or do a search on Savills Austin Q2 2026 to verify. Let's search `"Overall availability in Austin tightened further in Q2 2026"` to see if we can get more details. Let's do that.

Let's fetch that page: `https://www.savills.com.mx/research_articles/271937/392902-0` or `https://www.savills.us/research_articles/256536/392902-0`. Let's use `web_fetch` on `https://www.savills.com.mx/research_articles/271937/392902-0`. This should give us the exact details for Austin in Q2 2026.

Perfect! So Austin's office market is indeed stabilizing/improving, with overall availability tightening to **24.1%**, down 350 bps year-over-year from 27.6% in Q2 2025. Class A availability improved to **27.3%** (from 31.5% YoY). Let's look at Seattle again. We saw Puget Sound office vacancy at **23.4%** in Q2 2026, up 30 bps from the prior quarter, according to Kidder Mathews. Let's search for "Seattle Q2 2026 office vacancy Savills" to see if there is a Savills report for Seattle in Q2 2026. Let's do that.

This is incredibly detailed! Let's examine the Kidder Mathews Q2 2026 Seattle Office Market Report: - **Regional Vacancy Reaches 23.4%**: Puget Sound overall office vacancy is 23.4%, up 30 basis points quarter-over-quarter. - **Seattle Total (City of Seattle) Vacancy**: **28.2%** overall vacancy rate, with direct vacancy at **25.9%**. - **Seattle CBD (Downtown Seattle)**: overall vacancy rate is **34.9%** (with direct vacancy at 32.3%). Net absorption for Seattle CBD in Q2 2026 was **-34,701 SF** and YTD is **-222,666 SF**. - **East King County Total (Bellevue, Kirkland, etc.)**: overall vacancy rate is **21.0%**, but **Bellevue CBD** is doing much better with overall vacancy at **23.2%** and **+366,168 SF** of positive net absorption in Q2 2026. This shows that the Eastside (Bellevue) is pulling demand, whereas Seattle CBD continues to suffer negative net absorption (-34,701 SF in Q2) and extreme vacancy (34.9%). - **Construction pipeline is near historic lows**: Only 123,667 SF under construction across Puget Sound, a 95% YoY decline! Let's compare this to San Francisco in Q2 2026: - San Francisco's total vacancy fell to **27.2%** (down 440 bps YoY from 31.6% in Q2 2025). - Net absorption in SF was **positive 829,597 SF** in Q2 2026 alone, with YTD net absorption at **1.68M SF**. - In contrast, Seattle's total YTD net absorption is **negative -609,689 SF** (and Seattle CBD alone is negative -222,666 SF YTD). - This is a massive divergence! San Francisco is experiencing a full-blown, AI-driven office recovery, with positive net absorption of 1.68 million square feet in the first half of 2026, while Seattle CBD is still bleeding occupancy (-222k SF YTD) and stuck at 34.9% vacancy. Let's check the residential real estate market divergence, particularly Case-Shiller or Zillow/Redfin data for these metros in mid-2026 (Q2 2026/mid-2026). Let's search for "Case Shiller home price index San Francisco Seattle Austin Q2 2026" or "Zillow Home Value Index San Francisco Oakland Austin 2026". Let's run a `web_search`.

Let's check the Realtor.com research or other sources for the April 2026/May 2026 Case-Shiller report to see the regional divides: "Midwest strength and Sun Belt declines widening housing market divides". Let's search for "Case-Shiller Home Price Index" "April 2026" or "May 2026" to get the latest monthly figures (since standard Case-Shiller has a 2-month lag, the April 2026 indices were released in late June 2026, and May 2026 indices are released in late July 2026). Let's search `"Case-Shiller" "San Francisco" "Seattle" "Austin" 2026` or `"Case-Shiller" regional divergence 2026`. Let's run a `web_search`.

Let's look at the LinkedIn post by Lisa Sturtevant (Chief Economist at Bright MLS) or search for standard Case Shiller values in FRED. Wait, let's search for "SFXRSA" (San Francisco Case-Shiller) and "SEXRSA" (Seattle Case-Shiller) and "ATXRSA" (Atlanta) or other FRED series for 2026. Let's search for `site:fred.stlouisfed.org "SFXRSA"` or `site:fred.stlouisfed.org "SEXRSA"` or `site:fred.stlouisfed.org "LXXRCO"` or similar. Let's do a search.

Ah, the web_search budget is spent. No more web searches can be made. This is a crucial constraint! Now I must rely on the rich information I have already fetched and the public-markets data from my `market_view` calls, and synthesize them into high-quality notes. Let's see what threads we have: 1. `Q3 2026 City-Level Housing Price Divergence Report` (one_shot, fire_date: 2026-11-15T00:00:00Z) - Not fired yet. 2. `Q3 2026 Metro Office Vacancy and Net Absorption Reports` (ongoing, fire_date: 2026-10-15T00:00:00Z) - Not fired yet. 3. `Q4 2026 CMBS Hard Maturity Refinancing and Payoff Rates` (one_shot, fire_date: 2027-01-15T00:00:00Z) - Not fired yet. So no watches are firing in this cycle (since it's July 2026 and these are Q3/Q4 2026 watches). Let's review the prior notes and decide which ones we can update. We have: - `office-vacancy-metro-divergence-sf-seattle-austin` ("Tech Metros Diverge: San Francisco Office Rebounds via AI While Seattle and Austin Stabilize") Let's update this note! Wait, let's look at the prior version of `office-vacancy-metro-divergence-sf-seattle-austin`. Let's read its content or check the preview. The preview says: "The geographic and structural bifurcation of the U.S. commercial office market has entered a new phase in Q2 2026. While national office vacancy remained largely flat, declining just 10 basis points..." Ah, let's write a note using the slug `office-vacancy-metro-divergence-sf-seattle-austin` to update it with the latest Q2 2026 data we just retrieved: - **San Francisco**: Vacancy fell to **27.2%** in Q2 2026, down 440 bps YoY from 31.6% in Q2 2025 (according to Kidder Mathews). Net absorption was positive **829,597 SF** in Q2 2026 (YTD 1.68M SF). This represents a massive AI-driven leasing cycle, with 7 consecutive quarters of >2.0M SF leasing activity (3.6M SF in Q2 alone). Large leases signed by AI giants like **together.ai** (149,000 SF at 2 Henry Adams St), **LangChain** (70,000 SF at 303 2nd St), **Harvey AI** (62,000 SF at 201 Third St), and **Assort Health** (62,000 SF at One Market Plaza). - **Seattle**: Puget Sound overall vacancy rose to **23.4%** in Q2 2026 (up 30 bps QoQ), according to Kidder Mathews. Seattle CBD overall vacancy remains extremely high at **34.9%** (direct 32.3%), with negative Q2 net absorption of **-34,701 SF** (YTD negative -222,666 SF). A stark contrast with SF's positive absorption. However, Bellevue CBD is showing resilience with **+366,168 SF** positive net absorption in Q2 and vacancy at 23.2%, showing a sub-regional Eastside vs. Downtown Seattle divergence. - **Austin**: Office market is showing signs of stabilization but remains elevated. Overall availability tightened to **24.1%** in Q2 2026, down 350 bps YoY from Q2 2025, according to Savills. Class A availability improved to **27.3%** (down from 31.5% YoY). This indicates Austin is working through its massive supply wave, though vacancy remains high compared to historical levels. Let's also look at `ai-housing-boom-sf-bay-area` or `us-housing-market-regional-fracture-case-shiller`. Wait, the Case-Shiller index for San Francisco (SFXRSA) stood at **358.896** in April 2026, down slightly from **359.316** in March 2026, but let's see what else we can write about the housing/CRE divergence. Wait, let's check the REIT earnings! We have `market_view` data for: - **SL Green Realty Corp (SLG)**: Q2 2026 FFO of **$1.43** per share (beating estimates of $1.19, FFO raised for '26). Strong Manhattan leasing, reaching **92% occupancy** at 1185 Sixth Avenue. However, they still recorded a net loss of **$26.5 million** for the quarter, and carry a very high **14.4x debt-to-EBITDA ratio** with $5.17 billion in total debt. This shows that while leasing is recovering in prime Manhattan Class A, high leverage and interest expenses keep net income in the red. - **Vornado Realty Trust (VNO)**: Market cap $8.13B, reported a net loss of $7.3 million in Q1 2026 on $459.1 million in revenue. Trailing twelve-month revenue declined 2.2% to $1.85 billion. They carry $8.41 billion in total debt (11.2x debt-to-EBITDA ratio). Morgan Stanley recently downgraded VNO from Equal-Weight to Underweight on July 22, 2026 with a price target of $32. - **Boston Properties (BXP)**: Market cap $12.32B, reported Q1 2026 revenue of $872.1M and net income of $101.6M, beating EPS estimates ($0.55 vs $0.28). Carries **$15.97 billion** in total debt with a **10.0x debt-to-EBITDA ratio** and negative free cash flow of **-$131.8 million** for Q1. This represents a fantastic update for `office-cmbs-delinquency-all-time-high` or a new note, or we can update `regional-bank-cre-concentration-risks` or create a note on REIT performance. Let's see if we can update `office-vacancy-metro-divergence-sf-seattle-austin` first. Let's structure the update of `office-vacancy-metro-divergence-sf-seattle-austin` with the Q2 2026 commercial real estate data we retrieved. Let's make sure it contains exact quotes and links. Let's review the Kidder Mathews San Francisco Office Market Report Q2 2026: Quote: "By the end of 2Q, San Francisco recorded its seventh consecutive quarter with more than 2.0M SF of leasing activity, including 3.6M SF in 2Q alone. Net absorption reached 830K SF for the quarter, bringing the year-to-date total to 1.7M SF. Both metrics are on pace for their strongest annual performance since 2018, signaling a market that has officially moved beyond stabilization and into the early stages of recovery." "The total vacancy rate declined to 27.2% at quarter-end, down substantially from its recent peak of 32% in 2024..." "The largest office lease signed in San Francisco in 2Q 2026 was the City and County of San Francisco’s 505,000 SF lease at 1455 Market Street... Several other large AI-sector leases were signed during the quarter, including together.ai’s 149,000 SF lease at 2 Henry Adams Street." Let's review the Kidder Mathews Seattle Office Market Report Q2 2026: Quote: "Puget Sound office vacancy increased 30 basis points quarter-over-quarter to 23.4%..." "Seattle CBD overall vacancy remains elevated at 34.9%, with direct vacancy at 32.3%... 2Q 2026 net absorption was -34,701 SF and 2026 YTD net absorption was -222,666 SF." "Bellevue CBD recorded 851K SF of leasing activity year-to-date and leased 6% of its inventory base, outperforming Downtown Seattle’s 2% leasing rate." Let's review the Savills Austin Q2 2026 Office Market Report: Quote: "Overall availability in Austin tightened further in Q2 2026, declining to 24.1%, down 350 basis points (bps) from the same time last year. Class A availability also improved, falling to 27.3% from 31.5% year over year." Let's write this update to `office-vacancy-metro-divergence-sf-seattle-austin`. This is a highly substantive update that directly addresses the core of the thesis: the geographic bifurcation of tech metros. Let's prepare the note. I will write the note for `office-vacancy-metro-divergence-sf-seattle-austin` now.

Now let's check the REIT performance note or create a new one about office REITs. The brief asks to track CRE exposure: - "SL Green, Boston Properties, Vornado; the CMBS market and regional-bank CRE books" - "REIT/bank earnings commentary on geographic divergence" Let's look at the data we pulled from `market_view` for SLG, VNO, and BXP: - **SL Green (SLG)** reported Q2 2026 earnings on July 24, 2026. - FFO per share was **$1.43** (or $1.45 according to some news, but $1.43 is the reported latest quarter in `market_view` vs $1.19 estimate, a clear beat). - Guidance for 2026 was raised by **$1.20 per share** on the back of strong Manhattan leasing. - Signed two new leases totaling ~57,000 SF at 1185 Sixth Avenue, reaching **92% occupancy** at that tower. - However, they still recorded a **net loss of $26.5 million** for the quarter, reflecting the heavy burden of interest rates (10-year Treasury yield at 4.71% as of July 23, 2026). - SLG remains highly leveraged with **$5.17 billion in total debt** and a **debt-to-EBITDA ratio of 14.4x**. - **Vornado Realty Trust (VNO)**: - Market cap: $8.13B. - Carries **$8.41 billion** in total debt, with an EV/EBITDA of 9.75 and debt-to-EBITDA of **11.2x**. - Downgraded by Morgan Stanley on July 22, 2026 to **Underweight** (from Equal-Weight) with a price target of $32 (stock is currently trading at $39.72, up 31.1% over the last 3 months). This shows that while stock prices have rallied, analysts remain highly skeptical of the fundamentals given the high leverage and slow net income recovery. - **Boston Properties (BXP)**: - Market cap: $12.32B. - Carries **$15.97 billion** in total debt with a **10.0x debt-to-EBITDA ratio**. - Reported negative free cash flow of **-$131.8 million** for Q1 2026, with operating cash flow of $156.5M and capex of $288.3M. This illustrates the enormous capital expenditure burden office REITs face to upgrade assets to "Class A" or "Trophy" status to attract tenants in this bifurcated market. Let's write a note about this divergence between the public office REIT stock rally and their underlying balance sheet and cash flow risks (high leverage, high capex, negative FCF). Wait, we have an existing note: `regional-bank-cre-concentration-risks` ("Regional Bank Performance Diverges from Embedded Commercial Real Estate Risks"). But we can also update `office-cmbs-delinquency-all-time-high` or write a new note specifically on **Office REIT Balance Sheet and Capex Pressures**. Let's review the existing notes list: - `office-cmbs-delinquency-all-time-high` ("Office CMBS Delinquency Rates Hover Near Record Highs Amid Maturity Wall Crisis") - `regional-bank-cre-concentration-risks` ("Regional Bank Performance Diverges from Embedded Commercial Real Estate Risks") Let's create a new note or update `office-cmbs-delinquency-all-time-high`? No, let's write a new finding specifically on the major office REITs (SLG, BXP, VNO) to track their earnings, leverage, and capex pressures in mid-2026. This is a highly relevant subject that fits the brief's focus on CRE exposure. Let's call the slug `office-reit-leverage-capex-divergence`. Let's design this note. Wait, let's cite the `/markets` URLs for BXP, SLG, and VNO as well as the macro backdrop! - BXP: `/markets/BXP/2026/07/27` - SLG: `/markets/SLG/2026/07/27` - VNO: `/markets/VNO/2026/07/27` - Macro: `/markets/macro/2026/07/27` Let's write the note. I will call `write_note` for `office-reit-leverage-capex-divergence`.

Let's check if there are any other notes we should update or write. Let's see if we can update the `data-center-reit-equinix-leads-real-estate-divergence` finding. Wait, let's look at the `market_view` for the Real Estate sector and Equinix (EQIX): - **Equinix (EQIX)** has a market cap of **$106.93B** and trades at a massive P/E of **75.19**! - Revenue growth is **12.1% YoY** with a profit margin of 24.3%. - In contrast, Prologis (PLD) has a market cap of $140.85B and a P/E of 32.88, while American Tower (AMT) has a market cap of $77.65B and a P/E of 26.88. - Realty Income (O) has a market cap of $61.17B and a P/E of 52.9. This shows that Equinix's valuation multiple (75.19x P/E) is astronomical compared to traditional real estate peers, reflecting the intense investor demand for AI digital infrastructure assets (data centers) over traditional real estate. Let's update `data-center-reit-equinix-leads-real-estate-divergence` to include these latest valuation multiples, revenue growth figures, and the contrast with traditional real estate. Wait, let's review the preview of `data-center-reit-equinix-leads-real-estate-divergence`: "The AI-driven structural bifurcation within the U.S. commercial real estate (CRE) sector is intensifying. While traditional office real estate investment trusts (REITs) grapple with high leverage..." Let's write a revised version of `data-center-reit-equinix-leads-real-estate-divergence` to incorporate our new public-markets data. Let's make sure it contains exact figures and links. Let's do that now. I will call `write_note` to update `data-center-reit-equinix-leads-real-estate-divergence`.

This is excellent. We have updated two major notes: 1. `office-vacancy-metro-divergence-sf-seattle-austin` 2. `data-center-reit-equinix-leads-real-estate-divergence` And we created a new note: 3. `office-reit-leverage-capex-divergence` Let's double-check if there are any other notes we should update or write. Let's see if we can do something with `regional-bank-cre-concentration-risks` or `office-cmbs-delinquency-all-time-high`. The brief asks us to track: - "CRE fundamentals by metro (office vacancy, CMBS delinquency, regional-bank CRE concentration)" - "migration and job-posting data, and REIT/bank earnings commentary on geographic divergence" Wait, since our web search budget is spent, we don't have new web sources for regional bank concentration or CMBS delinquency in this specific cycle beyond what we already have in our existing notes or what we can infer. Let's check the existing notes for CMBS and regional banks: - `office-cmbs-delinquency-all-time-high` ("Office CMBS Delinquency Rates Hover Near Record Highs Amid Maturity Wall Crisis") - `regional-bank-cre-concentration-risks` ("Regional Bank Performance Diverges from Embedded Commercial Real Estate Risks") We don't need to force an update if we don't have new primary data for those specific notes in this cycle. The updates we have made to tech metro office vacancy and REIT performance are highly substantive and perfectly ground the "US housing/CRE market is several diverging markets" thesis. Wait, let's look at the active threads again to see if we can resolve any. Let's look at `list_threads` output: - `Q3 2026 City-Level Housing Price Divergence Report` (one_shot, fire_date: 2026-11-15T00:00:00Z) - `Q3 2026 Metro Office Vacancy and Net Absorption Reports` (ongoing, fire_date: 2026-10-15T00:00:00Z) - `Q4 2026 CMBS Hard Maturity Refinancing and Payoff Rates` (one_shot, fire_date: 2027-01-15T00:00:00Z) These threads are forward-looking watches that won't fire until October/November 2026 and January 2027. So we don't resolve them now. They remain open for future cycles. Let's call `finish_cycle` to complete this research cycle and present our findings! I will make the final call to `finish_cycle`.

I have completed this research cycle, grounding the analysis in both newly fetched Q2 2026 commercial real estate reports and real-time public-markets data. Here is a summary of the findings and updates integrated during this cycle: ### 1. Tech Metros Diverge: San Francisco Office Rebounds via AI While Seattle and Austin Stabilize * **Slug**: `office-vacancy-metro-divergence-sf-seattle-austin` *(Updated)* * **Key Findings**: * **San Francisco**: The office market is experiencing an aggressive, AI-led leasing expansion. Overall vacancy fell to **27.2%** in Q2 2026 (down 440 bps YoY from 31.6% in Q2 2025). Net absorption was positive **829,597 SF** in Q2 (bringing YTD to **1.68 million SF**), fueled by seven consecutive quarters of >2.0 million SF of leasing activity. Large commitments from AI firms like **together.ai** (149,000 SF), **LangChain** (70,000 SF), **Harvey AI** (62,000 SF), and **Assort Health** (62,000 SF) have solidified SF's position as the primary winner. * **Seattle**: Greater Seattle overall vacancy rose to **23.4%** in Q2 2026. The downtown core remains severely distressed, with **Seattle CBD** overall vacancy at **34.9%** and negative net absorption of **-34,701 SF** in Q2 (YTD **-222,666 SF**). However, a sub-regional divergence is clear as **Bellevue CBD** (on the Eastside) outperformed with **+366,168 SF** positive net absorption in Q2 and vacancy at **23.2%**. * **Austin**: Overall availability tightened to **24.1%** in Q2 2026 (down 350 bps YoY), indicating that Austin is successfully working through its massive post-pandemic supply wave, though vacancy remains high compared to historical levels. ### 2. Office REITs Face Balance Sheet and Capex Pressures Despite Q2 Earnings Beats and Stock Rally * **Slug**: `office-reit-leverage-capex-divergence` *(New Finding)* * **Key Findings**: * A stark divergence has formed between office REIT stock prices (which rallied 18% to 31% over the past three months) and their underlying balance sheet health. * While **SL Green (SLG)** beat Q2 FFO estimates ($1.43 vs. $1.19 est.) and raised guidance on strong Manhattan leasing, it still recorded a **net loss of $26.5 million** for the quarter, carrying **$5.17 billion in total debt** (an extremely high **14.4x debt-to-EBITDA ratio**). * **Boston Properties (BXP)** reported negative free cash flow of **-$131.8 million** for Q1 2026, driven by a massive **$288.3 million capex spend** required to upgrade its properties to "Class A/Trophy" status—highlighting the "capex trap" office landlords face to attract tenants in this bifurcated market. * High interest rates (with the 10-year Treasury yield at **4.71%** and Fed funds at **3.63%**) continue to make refinancing dilutive, prompting analysts like Morgan Stanley to downgrade **Vornado (VNO)** to **Underweight** despite the stock's recent 31.1% rally. ### 3. AI Digital Infrastructure Pulls Capital as Data Center REITs Eclipse Office Assets * **Slug**: `data-center-reit-equinix-leads-real-estate-divergence` *(Updated)* * **Key Findings**: * Capital allocation continues to shift from traditional workspaces into the physical backbone of the AI economy. * **Equinix (EQIX)** now commands a massive **$106.93 billion** market cap and trades at an astronomical P/E ratio of **75.19** (with **12.1% YoY revenue growth**). * This valuation premium stands in stark contrast to traditional commercial real estate comps (Prologis at 32.88x, American Tower at 26.88x, Realty Income at 52.9x) and depressed office REIT valuations, reflecting investor conviction that data center demand is secular and compounding, whereas traditional office demand is cyclical and structurally impaired. All findings have been saved as persistent markdown notes, and the forward-looking watches remain active to monitor Q3 and Q4 2026 data releases.