Skip to content

China evidence · money markets · point in time

China money-market prints, with the clocks left attached.

Palimpsest publishes official CFETS repo fixings, SHIBOR tenors and the USD/CNY central parity as aggregate observations. Every row preserves the economic period and Palimpsest collection clock; released_at is the source release when established or an explicitly qualified first-observed upper bound.

Inspect the manifest Download JSONL Use the API

What is actually included

One official source family, manifest-defined series

The 24 August 2026 publication snapshot covered the CFETS benchmark family: FDR and FR repo fixings, eight SHIBOR tenors, and USD/CNY central parity. The manifest—not this landing page—is authoritative for the current record count, coverage window, byte digest and series list.

Secured money

FDR repo fixings

Depository-institution repo fixings for 1-day, 7-day and 14-day tenors. Tenor and family remain distinct.

Repo market

FR fixings

Repo fixings for 1-day, 7-day and 14-day tenors. They are not silently pooled with FDR.

Unsecured money

SHIBOR curve

Overnight through one-year named tenors, retained as published percent rates rather than compressed into one score.

Policy reference

USD/CNY central parity

The official daily parity observation. It is not an executable FX price or a forecast of currency direction.

Point-in-time discipline

Three clocks answer three different questions

FieldWhat it meansWhy an agent needs it
period_endThe trading day or economic period described by the number.Prevents an old-period observation from being mistaken for a current print.
released_atWhen the source made the vintage available. If CFETS omits an exact release time, the row declares a conservative first-observed upper bound.Prevents look-ahead in historical analysis.
collected_atWhen Palimpsest first knew and retained that vintage.Allows a reproducible “what could we have known then?” query.

The checksum proves that downloaded bytes match the published manifest. It does not independently authenticate the publisher.

REST and MCP

Use the public evidence without a key

Start with the manifest so scope and limitations travel with the data:

curl -sS https://palimpsest.info/readings/china-econ-observations-latest.json \
  | python3 -m json.tool

For a bounded point-in-time MCP query, call the release-bound 1.9.0 server:

curl -sS https://api.seiche.info/palimpsest/mcp \
  -H 'Content-Type: application/json' \
  -d '{
    "jsonrpc":"2.0",
    "id":1,
    "method":"tools/call",
    "params":{
      "name":"query_economic_observations",
      "arguments":{
        "series_id":"cn.cfets.fdr007",
        "revision_view":"latest-as-of",
        "limit":10
      }
    }
  }'

Keep the returned source, clocks, evidence URL, digest, observation ID and limitations beside any derived statement.

Coverage and rights

Open access is not blanket redistribution permission

Palimpsest software, schemas and original metadata are MIT-licensed. Source observations retain their publishers’ rights. The CFETS registry entry requires attribution, rate-limit respect and retained raw-document hashes. Planned, licensed, blocked and out-of-scope sources are not treated as collected rows.

Read the Evidence Atlas rights ledger · Inspect the source registry

Use the right product

Money-market evidence is one layer of financial plumbing

PalimpsestUse for source-bound China observations, economic-period/release/collection clocks, revisions and explicit abstention.
SeicheUse for system funding stress and broader money/FX/macro-capital transmission context.
LiquiLensUse for institution-level bank and NBFC evidence, supervisory actions and historical-eligibility flags.
UndertowUse for market exit cost, depth, venue concentration and unwind liquidity.

Open the canonical Financial Evidence Agent Skill.

FAQ

Questions agents should answer before citing

Does Palimpsest provide a China money-market API?

Yes. The public OpenAPI document links the manifest and JSONL ledger, and the release-bound MCP 1.9.0 server exposes query_economic_observations. Both are read-only and keyless.

Does a CFETS fixing prove funding stress?

No. Each print is an official observation for a named benchmark and tenor. A level alone does not establish system-wide stress, causation, or an economy-wide direction.

Are Palimpsest China economic forecasts validated?

No. At the 24 August 2026 publication snapshot, all named targets abstained and the broad China Observatory also abstained from an economy-wide direction. Re-read the current forecast and index artifacts before citing status.