A city that banks other people’s booms
San Francisco has repeatedly been the place where wealth produced elsewhere was financed, managed and spent on property. The gold of 1848 was mined in the Sierra foothills, not in the city. Yet San Francisco grew from fewer than 1,000 residents before the Gold Rush to about 25,000 by 1849 (National Park Service), and a state census counted 36,151 in 1852 (Annals of San Francisco). The city had the port, the merchants and, soon, the banks. When Nevada’s Comstock Lode began producing silver, the Bank of California, opened in San Francisco in 1864, and its “Bank Crowd” dominated the district’s economic life from 1867 to 1875 (Online Nevada Encyclopedia).
The 1906 earthquake and fire destroyed about 28,000 buildings across 4.7 square miles (USGS). Much of the rebuilding was paid for with outside money: of about $235 million in insured losses, roughly $180 million was paid, more than a fifth of it by British underwriters (Insurance Journal). That was a transfer to replace lost capital, not new wealth. Economic historians estimate that the disaster immediately reduced U.S. GNP by 1.5 to 1.8 percentage points, and that British insurers’ claim payments helped set the stage for the Panic of 1907 (Odell and Weidenmier).
The technology cycles followed the same arc faster. Office vacancy was 1.6 percent in early 2000 (FRBSF). After the dot-com collapse, the Bay Area lost 350,000 jobs, 9.5 percent of the total, between December 2000 and August 2003 (FRBSF), and CBRE’s vacancy rate reached 19.1 percent in 2003. The next expansion brought it down to 3.7 percent by late 2019 (CBRE via WealthManagement).
The recurring mechanism is not that the city itself produces gold, silver or software. San Francisco concentrates the institutions, firms and workers that turn distant or intangible value into demand for local buildings, on land scarce enough for that demand to reach prices. Each cycle was also followed, sooner or later, by a contraction.
LedgeX Research · Figure 1
Capital in, capital out: four cycles in one city
Selected verified events, 1846–2026. Open any event for its figures and sources. The point is the pattern: inflows from booms elsewhere, then a sharp contraction.
- Resource wealth
- Contraction
- Technology cycle
- AI era
1846–52Resource wealthGold Rush: the city grows on wealth mined elsewhere
Fewer than 1,000 residents in 1846; about 25,000 by 1849; 36,151 counted by the 1852 state census. Mining was in the foothills; the port, merchants and banks were in San Francisco.
1864–75Resource wealthComstock silver is financed from San Francisco
The Bank of California opens in 1864 and comes to dominate Comstock mining and milling; the profits flow back to San Francisco.
1906ContractionEarthquake and fire
About 28,000 buildings destroyed across 4.7 square miles. Of about $235 million in insured losses, roughly $180 million was paid, more than a fifth by British underwriters. U.S. GNP fell an estimated 1.5–1.8 points.
2000Technology cycleDot-com peak: office vacancy 1.6%
Vacancy was 1.6% in 2000:Q1 and 7.6% a year later, more than half of the increase from tech sublease space.
2000–03ContractionBust: 350,000 Bay Area jobs lost
Bay Area employment fell 9.5% between Dec. 2000 and Aug. 2003; CBRE office vacancy reached 19.1% in 2003.
2019Technology cycleSecond tech expansion: vacancy 3.7%
CBRE vacancy at the end of 2019, just before the pandemic.
2024 Q3ContractionPandemic trough: vacancy 36.9%
The highest in CBRE’s series for the city.
2025AI eraAI capital: $152.5 billion of venture funding
San Francisco metro companies raised 44.9% of all U.S. venture dollars, led by OpenAI’s $40 billion round.
2026 Q3AI eraVacancy 27.7% after eight positive quarters
AI firms leased about 4.4 million sq ft in the first three quarters of 2026 (Cushman & Wakefield).
Events are illustrative, not a complete economic history. Vacancy figures before 2023 come from different brokers and are not a single consistent series. Compilation: LedgeX Research.
From the trough to 2026
In the third quarter of 2024, 36.9 percent of San Francisco’s office space stood vacant, the highest rate in CBRE’s series for the city (San Francisco Examiner, CBRE data). The housing market had also stalled. About 4,130 houses and condominiums sold in 2023, the fewest since at least 2005 (Compass via the Chronicle), as the 30-year mortgage rate reached 7.79 percent in October 2023 (Freddie Mac). Census estimates put the city’s population at 816,926 in 2023, down from 878,826 in 2019 (Census via FRED).
By the third quarter of 2026, CBRE measured vacancy at 27.7 percent after eight consecutive quarters of positive net absorption (Examiner). The county’s median single-family sale price reached $2.15 million in March 2026 (California Association of Realtors), a record by Compass’s count (Bloomberg). Zumper’s median one-bedroom asking rent was $4,400 in September 2026, up 25.4 percent in a year, the fastest growth among the large cities it tracks; nationally the figure was flat (Zumper).
These changes coincided with San Francisco’s emergence as the center of generative-AI investment. Coincidence in time is not proof of cause. This article examines how AI capital can reach the city’s real estate, where that effect is visible in the data, and where it is not.
Four channels from AI capital to property
AI investment can reach real estate through four channels. They differ in timing, and in how much money actually changes hands locally.
Office leasing. This is the most direct channel and the best documented. Cushman & Wakefield counts about 4.4 million square feet leased by AI firms in the first three quarters of 2026, up from 2.1 million in all of 2025, in a market of 87 million square feet (Cushman & Wakefield). OpenAI’s footprint passed 1 million square feet in March 2026, including three large Mission Bay buildings (The Registry). In February 2026 Anthropic leased all of 300 Howard Street, about 466,000 square feet, with move-in reported for 2027 (Business Wire; SFGate). A signed lease is a commitment, not occupancy: some of this space will stay empty for another year.
Employment and pay. AI companies employ relatively few people. OpenAI and Anthropic together had fewer than 10,000 employees in early 2026 (SF Standard). Job postings by generative-AI companies in San Francisco rose from an average of 85 a month in 2023 to 410 in 2025, according to Lightcast data compiled by the city’s chief economist (Examiner). Yet information-sector employment in the San Francisco–San Mateo division fell 10 percent between August 2023 and August 2026, and total employment fell 2 percent (BLS; BLS). If AI is moving housing demand, the likely channel is not the number of jobs but the pay: a small number of very high-income households bidding for a fixed stock of homes.
Venture capital and liquidity. Venture funding goes to companies, which spend much of it on computing and chips, often outside the city. Employees gain spendable wealth only when their equity becomes cash. Companies in the San Francisco metro raised $152.5 billion in 2025, 44.9 percent of the U.S. total, led by OpenAI’s $40 billion round (PitchBook-NVCA via the Examiner). Far smaller sums reached households: OpenAI’s employee share sales were reported at about $6.6 billion in October 2025 and $7 billion in August 2026 (CNBC). Research on California IPOs found that house prices near a new public company’s headquarters rose 0.7 to 0.9 percent around filing and issuance (Hartman-Glaser, Thibodeau and Yoshida, 2023). The effect is measurable but modest, and the design measures prices near headquarters, not who the buyers were.
Constrained supply. Extra demand raises prices most where supply cannot respond. Saiz estimated that 73 percent of the land around San Francisco is undevelopable and ranked the metro fourth-least responsive in the country (Saiz, 2010). The city added 2,669 net units in 2025, below its 10-year average of 3,622 (SF Planning), and a 2026 Controller’s study found that new apartment buildings do not pencil at current costs, even with no affordable-housing requirement (SF Controller).
LedgeX Research · Figure 2
How technology wealth reaches real estate
The path from a funding announcement to a home price runs through several steps. Each one shrinks the amount and delays the timing. The chips show what this article could and could not measure.
1
Venture capital raised
Money invested in companies. Mostly spent on computing, chips and salaries, much of it outside the city.
Measured: $152.5B, SF metro, 2025
2
Company expansion
Offices leased, then occupied. The most direct link to real estate.
Measured: ~4.4M sf of AI leasing, 2026 Q1–Q3
3
Salaries
Recurring income that supports rent and mortgage payments. Concentrated in few, highly paid workers.
Partly measured: postings rising, information jobs down 10%
4
Equity becomes cash
Tender offers, secondary sales, IPOs. Only realized equity becomes spendable wealth.
Partly measured: reported tender totals only
5
Household purchasing power
Down payments and bids for a nearly fixed stock of homes.
Not measured: no data links buyers to AI equity
6
Prices and rents
The observed outcome, which other forces also move.
Measured: rents +25%; repeat-sales indexes below 2022 peak
Supply sets the multiplier. San Francisco ranks among the four least supply-responsive U.S. metros, so a given rise in purchasing power moves prices more than it would in a city that can build quickly.
A conceptual diagram, not a quantified flow. Figures in the chips are cited in the text. Diagram: LedgeX Research.
What the market data show
Housing prices. The answer depends on the measure. The county’s median single-family price was $2.15 million in March 2026, a record by Compass’s count, and $2.2 million in May, according to C.A.R. (March; May and June). A median moves with the mix of homes that sell, and the mix moved toward the top: Redfin reported luxury closed sales in the city up 46 percent from a year earlier in mid-2026 (Real Estate News). The Case-Shiller indexes, which track repeat sales of the same homes across the San Francisco–Oakland metro area, show something more modest. In July 2026 the single-family index was 3.4 percent above December 2023 and still 5.4 percent below its May 2022 peak. The condominium index was 1.1 percent below its December 2023 level and 8.4 percent below its peak (S&P via FRED; condo index). Sales volume recovered only partly: third-quarter closings rose from 877 in 2023 to 1,107 in 2026, still below 1,157 in 2022 (SFAR MLS via Sotheby’s).
Rents. Rents moved more sharply. Zumper’s median one-bedroom asking rent rose from about $2,920 at the end of 2023 to $4,400 in September 2026 (Zumper via the Examiner; Zumper). Apartment List’s October 2026 report put the city’s rents 25.7 percent above a year earlier (Apartment List). Rents depend on current income and new arrivals rather than accumulated equity, so they are the price most consistent with an influx of well-paid workers. They are also consistent with scarcity: CoStar counted fewer than 1,000 market-rate units started in the preceding 12 months (SF Standard).
LedgeX Research · Figure 3
Record medians, partial recovery
City median prices hit records in 2026. Indexes that compare repeat sales of the same homes are still below their 2022 peaks, and condominiums lag houses.
A · Repeat-sales price indexes, San Francisco–Oakland metro
- Single-family
- Condominium
Data table: Case-Shiller index values
| Month | Single-family | % vs May 2022 peak | Condominium | % vs Mar 2022 peak |
|---|---|---|---|---|
| Dec 2022 | 339.57 | -11.5% | 286.26 | -7.8% |
| Mar 2023 | 336.06 | -12.4% | 282.08 | -9.1% |
| Jun 2023 | 342.48 | -10.7% | 282.89 | -8.9% |
| Sep 2023 | 351.08 | -8.5% | 290.03 | -6.6% |
| Dec 2023 | 351.11 | -8.5% | 287.57 | -7.4% |
| Mar 2024 | 352.82 | -8.0% | 282.97 | -8.9% |
| Jun 2024 | 357.90 | -6.7% | 286.24 | -7.8% |
| Sep 2024 | 358.13 | -6.6% | 285.46 | -8.1% |
| Dec 2024 | 360.95 | -5.9% | 282.38 | -9.1% |
| Mar 2025 | 358.27 | -6.6% | 280.36 | -9.7% |
| Jun 2025 | 350.37 | -8.7% | 274.45 | -11.6% |
| Sep 2025 | 355.12 | -7.4% | 272.36 | -12.3% |
| Dec 2025 | 360.60 | -6.0% | 281.69 | -9.3% |
| Mar 2026 | 359.49 | -6.3% | 285.45 | -8.1% |
| Jun 2026 | 361.41 | -5.8% | 287.43 | -7.4% |
| Jul 2026 | 362.87 | -5.4% | 284.32 | -8.4% |
| 2022 peak | 383.55 (May) | 0.0% | 310.48 (Mar) | 0.0% |
B · Same two years, two kinds of measure: June 2024 to June 2026
| Measure | Jun 2024 | Jun 2026 | Change | Source |
|---|---|---|---|---|
| Median single-family sale price, SF County (C.A.R.) | $1,650,000 | $2,128,000 | +29.0% | S8 S7 |
| Case-Shiller single-family index, SF metro (repeat sales) | 357.90 | 361.41 | +1.0% | S34 |
| Case-Shiller condominium index, SF metro (repeat sales) | 286.24 | 287.43 | +0.4% | S35 |
How to read this. A median is the middle sale in a period, so it rises when more expensive homes sell. A repeat-sales index compares the same homes over time, which removes most of that mix effect. Panel A covers the San Francisco–Oakland–Hayward metro area, not the city alone; panel B mixes a county median and metro indexes, so compare the size of the changes, not the levels. Monthly medians are not seasonally adjusted; June is compared with June.
Sources: S&P Cotality Case-Shiller indexes, seasonally adjusted, Jan. 2000 = 100, via FRED S34 S35; California Association of Realtors county medians S8 S7. Quarter-end months shown, plus the latest month (July 2026). Chart: LedgeX Research.
Offices. CBRE’s vacancy rate fell from 36.9 percent to 27.7 percent in two years, and 2025 leasing of about 10.2 million square feet was the highest in six years, roughly a quarter of it by AI firms (Chronicle, CBRE data). The recovery is uneven. Placer.ai measured September 2026 office visits still 33.8 percent below 2019 (Placer.ai), and weekday foot traffic in the Financial District was about half its 2019 level in 2025 (Downtown SF Partnership). Values remain far below their peak. The debt on 600 California sold in 2026 for $361 per square foot, against $897 when WeWork bought the tower in 2019 (The Real Deal). Blackstone and DivcoWest paid about $111 million for 300 Howard, about $239 per square foot, before leasing it to Anthropic (SFGate). More than a quarter of the city’s office space is still empty.
LedgeX Research · Figure 4
AI leasing and the office market
CBRE’s quarterly vacancy rate for the City of San Francisco, with three AI lease milestones. Vacancy has fallen nine points from its peak, from a very high level.
- Vacancy, as reported
- Derived from a reported change
- AI lease milestone
- OpenAI: ~486,600 sf sublease, Mission Bay (Oct 2023) S57
- OpenAI: 550 Terry A. Francois Blvd. (Sep 2024) S58
- Anthropic: all of 300 Howard, ~466,000 sf (Feb 2026); OpenAI passes 1M sf (Mar 2026) S23
Scale: AI leasing against the whole market (Cushman & Wakefield)
4.4M sf leased by AI firms, 2026 Q1–Q3, against an inventory of 87.1M sf (≈5%). Leases are commitments; occupancy can lag by a year or more. S21
Data table: quarterly vacancy and status
| Quarter | Vacancy | Status | Source |
|---|---|---|---|
| 2023 Q4 | 35.6% | Reported | S62 |
| 2024 Q1 | 36.7% | Reported | S63 |
| 2024 Q2 | — | Not shown: no traceable final figure found | |
| 2024 Q3 | 36.9% | Reported | S59 |
| 2024 Q4 | 36.5% | Reported | S64 |
| 2025 Q1 | 35.8% | Reported | S64 |
| 2025 Q2 | — | Not shown: no traceable final figure found | |
| 2025 Q3 | 34.4% | Reported | S41 |
| 2025 Q4 | 32.8% | Reported | S61 |
| 2026 Q1 | 30.4% | Reported | S60 |
| 2026 Q2 | 29.2% | Derived: Q3 2026 value plus the reported 1.5-point quarterly decline | S1 |
| 2026 Q3 | 27.7% | Reported | S1 |
How to read this. Vacancy is CBRE’s overall rate (direct plus sublease) for the city office market. Quarterly values were compiled from CBRE’s own quarterly figures and contemporaneous reports of CBRE data. CBRE revises preliminary readings, so two quarters without a traceable final figure are left blank rather than estimated. Other brokers define the market differently (Cushman & Wakefield measured 29.3% for 2026 Q3), so series from different brokers should not be mixed. The lease markers show timing only; the chart does not attribute vacancy changes to them.
Sources: CBRE via S62 S63 S59 S64 S41 S1; CBRE S61 S60; leases S57 S58 S23 S22. Chart: LedgeX Research.
Public finances, policy and competing explanations
The city’s December 2025 financial plan projected deficits growing to $1.17 billion by fiscal 2029–30, while noting that San Francisco “benefits from its position at the center of the AI industry” (Joint Report). Both statements can be true. Under Proposition 13, property is taxed at 1 percent, plus voter-approved debt levies, of a value that rises at most 2 percent a year and is reset only on sale or new construction (Board of Equalization). Rising home prices raise revenue only as homes change hands; falling office values cut it quickly through appeals. Appeals rose from 1,417 in fiscal 2019–20 to 9,281 in 2024–25 (SF Standard), and General Fund property tax is projected to fall 1.2 percent in 2026–27 (Controller).
Business taxes respond faster. Budgeted business tax revenue rises 38 percent in 2026–27, to $1.57 billion, though the Controller attributes part of the increase to a one-time timing effect of Proposition M (Controller), and the December plan tied growth largely to companies’ worldwide sales rather than activity inside the city. The adopted budget closed a $643 million gap with about $696 million in one-time funding (SF Standard). A private valuation is not public revenue.
Other forces moved at the same time. Mortgage rates do not explain 2026: the 30-year rate rose from 5.98 percent in February to 7.40 percent in early October while prices climbed (Freddie Mac). Some buyers’ wealth comes from long-public companies such as Nvidia rather than private AI firms (Compass, via Real Estate News). Reported crime fell in 2025, with car thefts down 24 percent and larcenies 22 percent (SFPD data via SF Standard). Policy changes are real but mostly postdate the turn in office demand, which began in late 2024 (Examiner): Mayor Daniel Lurie’s four-day office requirement for city workers took effect in August 2025 (GrowSF), his Heart of the City directive followed that fall (City of SF), and the Family Zoning Plan (December 2025) and state transit upzoning under SB 79 (July 2026) will add housing over years, not quarters (GrowSF; Holland & Knight). Finally, part of any rebound is recovery from an unusually depressed 2023.
What can be established
Three findings are well supported. AI companies are the largest source of new office demand in San Francisco, concentrated in Mission Bay and SoMa. Rents have risen faster than in any of the large U.S. cities Zumper tracks. And the top of the housing market has reached record prices. The housing evidence is weaker than the headlines: repeat-sales indexes describe a partial recovery, condominiums remain below their peak, and nothing yet isolates the share of price growth attributable to AI rather than to scarcity, equity markets or recovery from a trough.
The indicators worth watching are those that separate these explanations: whether the Case-Shiller indexes pass their 2022 peaks; whether 2027 occupancy matches 2026 leases; how house prices behave around large liquidity events, such as tender offers or IPOs; and whether office assessments stabilize. San Francisco has captured capital from booms before. Its recent history is equally clear that booms can give way to sharp repricing.
About the data and graphics
- Geography. County medians (C.A.R.) and city office figures cover the City and County of San Francisco. Case-Shiller indexes cover the San Francisco–Oakland–Hayward metro area. Employment covers the San Francisco–San Mateo–Redwood City metropolitan division. Venture totals cover the San Francisco metro area as defined by PitchBook-NVCA.
- Prices. Medians are not adjusted for the mix of homes sold. Repeat-sales indexes are, but they reflect the metro area and exclude new construction.
- Office data. Vacancy is a single broker’s series (CBRE). CBRE revises its preliminary readings; quarters without a traceable final figure are left blank, and one quarter derived from a reported change is labeled. Lease sizes are as reported at signing.
- Causation. Nothing in this article estimates the share of any price change caused by AI. Coincident timing is reported as timing.
- Currency. Data as published through Oct. 14, 2026. Monthly releases due after that date are not included.
- Ownership. Cited data and reports belong to their publishers. LedgeX Research’s contribution is the synthesis, the compilation and the graphic design.
Sources
- S1 SF office vacancy drops again as AI-fueled recovery shows legs (CBRE Q3 2026 data) — San Francisco Examiner, 2026-10-14. Reported (broker data).
- S2 Compass 2023 San Francisco sales totals — San Francisco Chronicle (via PressReader), 2024-02-12. Reported (broker data).
- S3 Freddie Mac Primary Mortgage Market Survey: 30-year fixed, Oct. 26, 2023 (7.79%) — Freddie Mac, 2023-10-26. Primary.
- S4 Freddie Mac Primary Mortgage Market Survey, 2026 archive — Freddie Mac, 2026-10-08. Primary.
- S5 Resident population of San Francisco County (CASANF0POP) — U.S. Census Bureau via FRED, vintage 2025. Primary.
- S6 March 2026 home sales and price report (county medians) — California Association of Realtors, 2026-04. Primary (MLS data).
- S7 June 2026 home sales and price report (county medians) — California Association of Realtors, 2026-07. Primary (MLS data).
- S8 June 2025 home sales and price report (county medians; June 2024 comparison) — California Association of Realtors, 2025-07. Primary (MLS data).
- S9 San Francisco house prices hit record $2.15 million on AI boom — Bloomberg, 2026-04-06. Reported (Compass data).
- S10 National Rent Report, September 2026 — Zumper, 2026-09. Primary (listings data).
- S11 Why San Francisco rent is no longer most expensive in U.S. (Zumper year-end 2023) — San Francisco Examiner, 2023-12-18. Reported (listings data).
- S12 Gold Rush transforms San Francisco — National Park Service, n.d.. Agency history.
- S13 The Annals of San Francisco (1855): 1852 state census — Soulé, Gihon & Nisbet, via SF Genealogy, 1855. Primary (contemporary account).
- S14 William C. Ralston and the Bank of California — Online Nevada Encyclopedia, n.d.. Secondary (reference).
- S15 The 1906 earthquake: casualties and damage — U.S. Geological Survey, n.d.. Primary (federal).
- S16 1906: insurance and the San Francisco earthquake and fire — Insurance Journal, 2006-05-08. Secondary.
- S17 Odell & Weidenmier, Real Shock, Monetary Aftershock: The 1906 San Francisco Earthquake and the Panic of 1907 (J. Economic History 64(4)) — NBER Working Paper 9176, 2004. Peer-reviewed.
- S18 Western Economic Developments, June 2001 (office vacancy 2000Q1–2001Q1) — Federal Reserve Bank of San Francisco, 2001-06. Primary.
- S19 Daly & Doms, Economic Letter 2003-33: Bay Area job losses — Federal Reserve Bank of San Francisco, 2003-11-07. Primary.
- S20 We fact-checked claims about San Francisco office vacancy (CBRE historical series) — WealthManagement.com, 2022-09-15. Reported (broker data).
- S21 San Francisco Office MarketBeat, Q3 2026 — Cushman & Wakefield, 2026-10. Primary (broker data).
- S22 OpenAI completes 280,000 sq ft sublease at 1800 Owens Street — The Registry, 2026-03-12. Reported.
- S23 DivcoWest and Blackstone secure full-building lease with Anthropic at 300 Howard Street — Business Wire, 2026-02-02. Primary (company release).
- S24 AI is booming, but tech jobs in San Francisco are not — The San Francisco Standard, 2026-01-28. Reported.
- S25 San Francisco AI boom: tech jobs plateau (Lightcast postings compiled by the city economist) — San Francisco Examiner, 2026-01-11. Reported (city data).
- S26 All employees: Information, San Francisco–San Mateo–Redwood City MD (SMU06418845000000001) — BLS / California EDD via FRED, 2026-08. Primary.
- S27 All employees: Total nonfarm, San Francisco–San Mateo–Redwood City MD (SMU06418840000000001) — BLS / California EDD via FRED, 2026-08. Primary.
- S28 Why venture capital couldn’t quit San Francisco in 2025 (PitchBook-NVCA data) — San Francisco Examiner, 2026-01-14. Reported (PitchBook-NVCA data).
- S29 OpenAI completes $7 billion employee share sale — CNBC via Yahoo Finance, 2026-08. Reported.
- S30 Hartman-Glaser, Thibodeau & Yoshida, Cash to spend: IPO wealth and house prices (Real Estate Economics 51(1), 68–102) — Real Estate Economics, 2023. Peer-reviewed.
- S31 Saiz, The Geographic Determinants of Housing Supply (Quarterly Journal of Economics 125(3)) — QJE, 2010. Peer-reviewed.
- S32 2025 Housing Inventory — San Francisco Planning, 2026-04. Primary.
- S33 Triennial Economic Feasibility Report (inclusionary housing) — San Francisco Controller’s Office, 2026-04-30 (corrected 2026-05-19). Primary.
- S34 S&P Cotality Case-Shiller San Francisco Home Price Index, single-family, SA (SFXRSA) — S&P Dow Jones Indices via FRED, 2026-07. Primary.
- S35 S&P Cotality Case-Shiller San Francisco Condo Price Index, SA (SFXRCSA) — S&P Dow Jones Indices via FRED, 2026-07. Primary.
- S37 Tech wealth reshaping San Francisco’s housing market (Redfin luxury data; Compass comment) — Real Estate News, 2026-07-06. Reported.
- S38 San Francisco market update (SFAR MLS quarterly closed sales) — Sotheby’s International Realty, 2026-10. Broker report (MLS data).
- S39 San Francisco rent report, October 2026 — Apartment List, 2026-09. Primary (listings data).
- S40 Rental records in San Francisco (CoStar supply data; Zumper neighborhoods) — The San Francisco Standard, 2026-07-02. Reported.
- S41 Tech boom fuels S.F.’s best office year since 2019 (CBRE 2025 leasing) — San Francisco Chronicle via Yahoo Finance, 2025-12-19. Reported (broker data).
- S42 Data reveals downtown San Francisco’s evolution beyond the 9-to-5 — Downtown SF Partnership, 2026-03-16. Primary (district data).
- S43 September 2026 Placer.ai Office Index — Placer.ai, 2026-10-13. Primary (mobility data).
- S44 Lone Star Funds snags office tower after auction falls flat (600 California) — The Real Deal, 2026-04-21. Reported.
- S45 Anthropic grabs $111M office downtown (300 Howard) — SFGate via Yahoo Finance, 2026-02. Reported.
- S46 Joint Report: Five-Year Financial Plan Update, FY 2026-27 through FY 2029-30 — SF Controller, Mayor’s Budget Office & Board Budget Analyst, 2025-12-19. Primary.
- S47 Property Taxes: Law Guide (Publication 29) — California State Board of Equalization, current. Primary.
- S48 San Francisco property tax appeals backlog — The San Francisco Standard, 2025-11-13. Reported.
- S49 Revenue Letter: FY 2026-27 and FY 2027-28 proposed budget — SF Controller’s Office, 2026-06-09. Primary.
- S51 San Francisco balanced its budget. Does the fiscal crisis remain? — The San Francisco Standard, 2026-08-06. Reported.
- S52 San Francisco 2025 crime statistics (SFPD data) — The San Francisco Standard, 2026-01-02. Reported (police data).
- S53 City workers’ return-to-office date delayed to Aug. 18 — GrowSF, 2025-04-22. Secondary.
- S54 Mayor Lurie unveils Heart of the City executive directive — City and County of San Francisco, 2025. Primary.
- S55 Family Zoning Plan approved — GrowSF, 2025-12-02. Secondary.
- S56 California Gov. Gavin Newsom signs SB 79 — Holland & Knight, 2025-10. Secondary (law-firm summary).
- S57 OpenAI expands in San Francisco with major sublease from Uber — Hoodline, 2023-10-30. Reported.
- S58 OpenAI expands in Mission Bay (550 Terry A. Francois Blvd.) — The San Francisco Standard, 2024-09-20. Reported.
- S59 SF office vacancy rate declines in promising end to 2024 (first positive absorption) — San Francisco Examiner, 2024-12-19. Reported (broker data).
- S60 San Francisco Office Figures, Q1 2026 — CBRE, 2026-04. Primary (broker data).
- S61 San Francisco Office Snapshot, Q4 2025 — CBRE, 2026-01. Primary (broker data).
- S62 San Francisco office vacancies reach record high to end 2023 (CBRE data) — San Francisco Examiner, 2024-01-10. Reported (broker data).
- S63 San Francisco office vacancy rate inches to new record high (CBRE data) — San Francisco Examiner, 2024-06. Reported (broker data).
- S64 SF office vacancies drop as companies grab more space (CBRE Q1 2025 data) — San Francisco Examiner, 2025-04. Reported (broker data).
About this research: LedgeX Research examines the public records, physical constraints and economic forces that shape land and property decisions. We separate findings from hypotheses and unresolved questions. This article and its figures are not investment, lending, appraisal or legal advice, and they do not assess any specific property. Corrections: team@ledgexproperties.com.