# $SHELTER · New York City shelter census

People who slept in a New York City Department of Homeless Services shelter each night, adults and children, as the city reports it the next day. One city, the largest shelter system in the country.

| Field | Value |
| --- | --- |
| Newest complete period | Nov 10: 83,563 people, score 12, Quiet |
| Source | NYC Open Data, Department of Homeless Services, DHS Daily Report (https://data.cityofnewyork.us/Social-Services/DHS-Daily-Report/k46n-sa2m) |
| Terms | NYC Open Data. The city says there are no restrictions on the use of Open Data. |
| Cadence | Daily, one day late |
| Grade | reference |
| Rulebook | Rulebook v0.2, 2026-10-04 |

## Settlement

Why this grade: A city agency's own nightly count, published as open data. A venue can read it directly.
What it costs to move: Not practical. The count is of people in city shelters, taken by the agency that runs them.
When the source is late: The city posts a night the next day. Until it lands the day stays open. A night that never lands is void.
Precision: Whole people, the city's total individuals in shelter for the night.

## As first published, and the proof

Not on-chain yet. The newest complete period goes on-chain with the daily proof at 12:40 UTC.

## Last 30 complete periods

| Period | Level | Change | Band |
| --- | ---: | ---: | --- |
| Nov 10 | 83,563 people | -0.2% | Quiet |
| Nov 9 | 83,765 people | +0.1% | Quiet |
| Nov 8 | 83,699 people | +0.2% | Quiet |
| Nov 7 | 83,558 people | 0.0% | Quiet |
| Nov 6 | 83,533 people | +0.1% | Quiet |
| Nov 5 | 83,443 people | -0.3% | Quiet |
| Nov 4 | 83,684 people | -0.1% | Quiet |
| Nov 3 | 83,794 people | 0.0% | Quiet |
| Nov 2 | 83,822 people | +0.1% | Quiet |
| Nov 1 | 83,700 people | 0.0% | Quiet |
| Oct 31 | 83,697 people | +0.1% | Quiet |
| Oct 30 | 83,641 people | -0.1% | Quiet |
| Oct 29 | 83,716 people | -0.1% | Quiet |
| Oct 28 | 83,784 people | 0.0% | Quiet |
| Oct 27 | 83,813 people | 0.0% | Quiet |
| Oct 26 | 83,847 people | +0.1% | Quiet |
| Oct 25 | 83,802 people | 0.0% | Quiet |
| Oct 24 | 83,811 people | 0.0% | Quiet |
| Oct 23 | 83,829 people | +0.1% | Quiet |
| Oct 22 | 83,766 people | +0.1% | Quiet |
| Oct 21 | 83,682 people | -0.2% | Quiet |
| Oct 20 | 83,814 people | -0.5% | Normal |
| Oct 19 | 84,224 people | +0.2% | Quiet |
| Oct 18 | 84,075 people | +0.1% | Quiet |
| Oct 17 | 83,976 people | +0.1% | Quiet |
| Oct 16 | 83,886 people | +0.1% | Quiet |
| Oct 15 | 83,781 people | -0.3% | Quiet |
| Oct 14 | 84,037 people | -0.1% | Quiet |
| Oct 13 | 84,106 people | -0.1% | Quiet |
| Oct 12 | 84,172 people | +0.3% | Normal |

JSON: https://tickerz.com/api/indices/shelter

## Asked often

**How many people are in New York City homeless shelters?** 83,563 people slept in a New York City Department of Homeless Services shelter on the night of Nov 10 2026, adults and children, as the city reports it.

**Where does the $SHELTER number come from?** Tickerz reads it from NYC Open Data, Department of Homeless Services, DHS Daily Report. Daily, one day late. Each number is kept as it was first published, with a timestamped proof, and any later revision is shown beside it.

**Can I download the $SHELTER data?** Yes, free and with no key: the full history as JSON at https://tickerz.com/api/indices/shelter and every period as CSV at https://tickerz.com/api/indices/shelter/history.csv.

## Name a number

Name a number you want counted. Within 48 hours it is a ticker, or you get the plain reason it cannot be built. POST https://tickerz.com/api/indices/requests with {"counts": "...", "source_url": "..."}, or the MCP tool name_a_number. Answers: https://tickerz.com/x#x-request

## For agents

- JSON API, no key: https://tickerz.com/docs (OpenAPI: https://tickerz.com/openapi.json)
- MCP server, Streamable HTTP, no key: https://tickerz.com/mcp
- Site summary: https://tickerz.com/llms.txt (everything in one file: https://tickerz.com/llms-full.txt; agent instructions: https://tickerz.com/agents.md)
- Signed numbers, every feed and The Tape: https://oracle.tickerz.com (each signed number with its recount: https://tickerz.com/signed)
- Plain answers about prediction market data: https://tickerz.com/answers
- Any page here in Markdown: send Accept: text/markdown, or open /md/<path>.

The Activity score says how unusual trading is, not whether an asset is good. It is not a price signal: whether it predicts returns is measured live at https://tickerz.com/evidence. Not advice.