NYC Rent-Stabilized Building Price Index

Market-rate comparisonBoroughs (indicative)Monthly readings

Red: official quarterly index (dashed = provisional). Orange: monthly readings, which are noisier and published sooner. Hover for values. Borough series rest on 15–90 sales a quarter and are shown for context only.

How the index is built

Every saleAll sales of NYC buildings with 6+ apartments since 2016, from the Department of Finance: 5,896 qualifying sales, refreshed every month as the city publishes new records.

Only rent-stabilized buildingsMatched by tax lot to the State's 2024 registration list (47,198 buildings). 421-a tax-break buildings, co-ops and rooming houses are excluded.

Only real market salesPortfolio deals, single-apartment transfers and data errors are screened out. Every exclusion is published with its reason.

Like-for-likeA robust hedonic regression compares price per apartment while holding size, age, elevator, mixed use and neighborhood constant.

Never revisedPublished values are frozen in a ledger, and each new period is chained on. Late-recorded deeds never rewrite the past.

Hard to moveNo single sale has ever changed a quarter by more than 0.9%. The settlement value filters out measurement noise.

Full methodology →

Why it is built this way

Research behind the choices

Hedonic, not repeat-sales. Apartment buildings rarely sell twice in a decade, so a repeat-sales index (like Case-Shiller) would discard most sales. For indexes used to settle derivatives, hedonic indexes are much more stable under revision than repeat-sales ones (Clapham, Englund, Quigley & Redfearn 2006), and they signal turning points sooner (Shimizu, Nishimura & Watanabe 2010). Hill (2013) surveys the method.
Rolling window, chained. This is the rolling-window time-dummy method of the IMF/Eurostat Handbook on Residential Property Price Indices. Revision is a key risk for index derivatives (Deng & Quigley 2008), so published values are never changed.
Filtering a thin market. Indexes built from scarce transactions are noisy. Tradable commercial real-estate indexes handle this with signal extraction and frequency conversion (Bokhari & Geltner 2010; Bourassa, Cantoni & Hoesli 2017). The settlement value uses a Kalman filter for the same purpose.

Why rent-stabilized buildings

Rent rules change what a building can earn, so they change what it is worth. When rent control ended in Cambridge, MA, the value of formerly controlled buildings and their neighbors rose sharply (Autor, Palmer & Pathak 2014). When San Francisco expanded rent control, landlords reduced rental supply (Diamond, McQuade & Qian 2019).
New York's 2019 Housing Stability and Tenant Protection Act limited rent increases on stabilized apartments. The index shows stabilized buildings falling from that quarter on, while market-rate rental buildings did not.
Cross-check: Ariel Property Advisors reports stabilized building prices of $270K per unit in 2019 and $163K in 2025 (−40%). The index's levels are about $263K and $159K–$169K.

Quarterly values

Quarter	Status	Index	$ per apartment	Raw estimate	Raw change	95% interval	Sales	Max single-sale effect	Published
2026Q3	provisional	57.1	$150,096	57.4	+2.5%	-6.0% to +11.7%	129	0.51%	2026-10-15
2026Q2	official	56.0	$147,377	56.0	+2.0%	-5.3% to +9.8%	137	0.50%	2026-10-16
2026Q1	official	56.3	$148,014	54.9	-10.6%	-17.0% to -3.8%	196	0.51%	2026-10-16
2025Q4	official	61.9	$162,711	61.4	-3.7%	-11.1% to +4.4%	143	0.55%	2026-10-16
2025Q3	official	63.7	$167,644	63.7	+0.1%	-6.9% to +7.6%	139	0.46%	2026-10-16
2025Q2	official	63.8	$167,895	63.6	-1.7%	-8.8% to +5.9%	160	0.57%	2026-10-16
2025Q1	official	64.5	$169,765	64.8	+2.1%	-6.0% to +10.8%	114	0.65%	2026-10-16
2024Q4	official	63.8	$167,746	63.4	-1.6%	-9.3% to +6.6%	146	0.60%	2026-10-16
2024Q3	official	65.1	$171,097	64.5	-5.2%	-13.1% to +3.4%	111	0.57%	2026-10-16
2024Q2	official	67.1	$176,572	68.1	+10.4%	+0.9% to +20.7%	111	0.68%	2026-10-16
2024Q1	official	63.7	$167,600	61.7	-16.6%	-24.3% to -8.1%	78	0.74%	2026-10-16
2023Q4	official	72.0	$189,257	74.0	+16.7%	+6.4% to +28.1%	77	0.70%	2026-10-16
2023Q3	official	65.0	$170,835	63.4	-11.1%	-18.6% to -2.9%	107	0.61%	2026-10-16
2023Q2	official	71.3	$187,443	71.3	+1.4%	-7.0% to +10.5%	115	0.66%	2026-10-16
2023Q1	official	71.2	$187,384	70.3	-3.8%	-11.3% to +4.3%	96	0.75%	2026-10-16
2022Q4	official	74.9	$196,948	73.1	-11.1%	-17.5% to -4.3%	121	0.54%	2026-10-16
2022Q3	official	82.0	$215,725	82.2	+1.4%	-5.2% to +8.6%	130	0.46%	2026-10-16
2022Q2	official	81.3	$213,776	81.0	-2.6%	-8.6% to +3.8%	183	0.54%	2026-10-16
2022Q1	official	82.2	$216,116	83.2	+4.5%	-1.3% to +10.6%	147	0.44%	2026-10-16
2021Q4	official	78.4	$206,300	79.6	+7.8%	+1.8% to +14.1%	211	0.39%	2026-10-16
2021Q3	official	74.2	$195,089	73.9	-2.7%	-8.8% to +3.9%	119	0.58%	2026-10-16
2021Q2	official	75.3	$198,103	75.9	+3.6%	-3.6% to +11.3%	109	0.60%	2026-10-16
2021Q1	official	73.2	$192,578	73.3	+1.3%	-6.7% to +10.0%	80	0.61%	2026-10-16
2020Q4	official	73.1	$192,212	72.3	-3.4%	-11.5% to +5.4%	92	0.67%	2026-10-16
2020Q3	official	76.0	$200,009	74.8	-6.3%	-15.2% to +3.5%	75	0.89%	2026-10-16
2020Q2	official	80.7	$212,167	79.9	-4.4%	-13.5% to +5.7%	51	0.86%	2026-10-16
2020Q1	official	83.7	$220,155	83.5	+2.1%	-6.3% to +11.2%	92	0.64%	2026-10-16
2019Q4	official	84.3	$221,823	81.8	-11.6%	-18.8% to -3.9%	85	0.66%	2026-10-16
2019Q3	official	94.4	$248,214	92.6	-7.4%	-13.4% to -0.9%	73	0.64%	2026-10-16
2019Q2	official	101.2	$266,244	100.0	-6.5%	-11.6% to -1.0%	143	0.41%	2026-10-16
2019Q1	official	106.0	$278,694	106.9	+4.3%	-1.6% to +10.6%	105	0.43%	2026-10-16
2018Q4	official	102.6	$269,828	102.5	+0.0%	-5.2% to +5.6%	173	–	2026-10-16
2018Q3	official	102.9	$270,745	102.5	-2.2%	-7.1% to +3.0%	164	–	2026-10-16
2018Q2	official	104.7	$275,477	104.7	-1.4%	-6.2% to +3.6%	161	–	2026-10-16
2018Q1	official	104.8	$275,556	106.3	+6.7%	+1.0% to +12.7%	193	–	2026-10-16
2017Q4	official	99.4	$261,489	99.6	+0.8%	-4.8% to +6.6%	149	–	2026-10-16
2017Q3	official	98.7	$259,683	98.9	+0.9%	-4.8% to +7.0%	158	–	2026-10-16
2017Q2	official	98.2	$258,370	98.0	-1.1%	-6.9% to +5.1%	195	–	2026-10-16
2017Q1	official	99.2	$260,874	99.1	-1.4%	-6.7% to +4.3%	134	–	2026-10-16
2016Q4	official	99.7	$262,120	100.4	+3.7%	-1.1% to +8.7%	220	–	2026-10-16
2016Q3	official	96.9	$254,760	96.8	-1.5%	-6.1% to +3.2%	201	–	2026-10-16
2016Q2	official	96.9	$254,920	98.4	+7.5%	+2.7% to +12.6%	246	–	2026-10-16
2016Q1	official	91.5	$240,604	91.5	+0.0%	+0.0% to +0.0%	227	–	2026-10-16

Downloads

Citywide quarterly (CSV)Citywide monthly (CSV)Published-value ledger (JSON)All series + metadata (JSON) Market-rate comparison (CSV)Manhattan Bronx Brooklyn Queens Every sale used (CSV)Every sale excluded, with reason (CSV)Stabilized buildings list (CSV)

Release built 2026-10-15T23:36Z from code version 7ac319b. Input snapshots and their checksums are listed in rsi.json under release.

Governance

The index follows the structure of the IOSCO Principles for Financial Benchmarks (2013): governance, benchmark quality, methodology quality and accountability.

Transparent methodology: the full method and code are public, and every number can be recomputed from the published inputs.
Update schedule: an automated job checks the city's data daily and republishes when new records appear (the city updates monthly, mid-month). A period becomes official once the data runs 30 days past its end; until then it is marked provisional.
Data sufficiency: a quarter is published only with at least 40 qualifying sales. The minimum so far is 51.
Changes: method changes are announced at least one quarter ahead, and filter parameters are re-estimated at most once a year, with notice.

Corrections: published values are not revised. A data error is corrected with a dated notice, and the original value is kept for the record.
Audit trail: each release records its code version and input checksums.
Questions and complaints: open an issue on GitHub.

This is a research preview. It has not been independently audited, and no administrator has been formally designated under any benchmark regulation.

Limitations

Stabilization status comes from the 2024 registrations. Buildings that left stabilization before then, mostly before 2019, are missing from the early years.
Portfolio deals are excluded because the city records only their total price, not a price per building.
Sales are dated by deed, so the index trails deal-making by one to three months.
The model cannot see the share of apartments that are stabilized, the rent roll or building condition. These drive value, so quarterly estimates are noisy (about ±7%).

Rent-stabilized buildings in New York City have lost 44% of their value since 2019.