Asia Pulse
ASIA_SESSION · --:--:-- KST · --

The market signal Western agents miss while Asia trades.

Kimchi premium, funding-rate regime, and Korea equity-chain analysis — structured for AI agents, priced per call, settled on-chain. Built on a live 24/7 trading system that runs on this same data.

Use on OKX AI → Use on Agentic Market → View products
7
Live endpoints
7
Licensed sources
24/7
Asia-hours coverage
BTC kimp +0.05% ETH kimp +0.02% DOGE kimp −0.70% SOL kimp +0.02% REGIME neutral SEMIS deviation +1.39pp DEFENSE deviation −2.40pp FUNDING BTC +0.008% BTC kimp +0.05% ETH kimp +0.02% DOGE kimp −0.70% SOL kimp +0.02% REGIME neutral SEMIS deviation +1.39pp DEFENSE deviation −2.40pp FUNDING BTC +0.008%
01
how it works

From licensed source
to agent-ready JSON.

Every call runs the same disciplined pipeline. No scraped feeds enter it, and nothing leaves it without an as-of timestamp.

01
Collect

Collect

Licensed and public sources only — CoinGecko, US Treasury, and government customs and procurement records.

02
Validate

Validate

Freshness gates and market-calendar awareness flag stale or failed inputs before they ship.

03
Analyze

Analyze

Walkforward-validated coefficients turn raw inputs into transmission signals and deviation.

04
Serve

Serve

One call returns the full analysis as JSON — attributed, freshness-annotated, with a stated falsifier.

7
Live
endpoints
7
Licensed
data sources
3
Korea equity
chains
2×
Daily analyst
reports
24/7
Asia-session
coverage
02
why asia pulse

Coverage, honesty,
and a real book behind it.

Three things that separate this from a raw price feed.

01

Asia-hours edge

Korean retail flow, kimchi premium, and Korea/Japan equity context — the window that opens while US and EU desks are dark.

02

Honest by design

Every analyst brief states its own falsifier — the condition under which the read is wrong. Every response carries a data-freshness block.

03

Skin in the game

The same data drives a live 24/7 trading system. We consume what we sell, so a broken layer costs us first.

03
products

Seven endpoints.
Pay per call.

No parameters required — each call returns the full analysis as JSON. Paid through the OKX AI marketplace, settled on-chain (XLayer).

01
Asia Session Regime Report
GET /report
Per-coin kimchi premium, BTC/ETH funding rates, regime classification, and where today’s reading sits in our own archive — the core Asia-session snapshot.
0.03USDT / call▸
analyst_brief · samplevalues masked
global priceBTCETHSOLDOGEWIFPEPESHIBBONKKorea richer ▲Korea cheaper ▼
Korea vs global crypto price
The Asia session opens in a neutral regime, [xx.x] hours in with no change over 24 hours — no directional edge from Korean flow. The Korea-vs-global price gap (the “kimchi premium”) sits flat-to-negative across majors and funding is neutral. Korean retail appetite reads muted, two-way.↪ full snapshot as JSON via API
live · 24h rollingtimestamped per blockpercentile ranked
02
Asia Session Premium Intel
GET /report/premium
Everything in the regime report, plus 24h premium stats, regime stability metrics, macro context, an events calendar, and a twice-daily AI-analyst report.
0.1USDT / call▸
analyst_brief · samplevalues masked
US Treasury curve — the macro backdrop we read Korea against2Y5Y10Y20Y30Yupward = no inversion
US macro backdrop (Treasury curve)
We read Korea against a backdrop you already know. US long yields are elevated with a positive [xx]bp spread — no inversion, but a rising long end tightens global liquidity. Asian currencies weaken faster than developed peers, compressing the Korea price gap. The calendar flags [x] high-impact catalysts around month-end.↪ full analyst report + calendar via API
2× dailymacro cross-checkedobserved minimum
03
Korea Semiconductor Chain
GET /chain/semis
US-to-Korea transmission signal, sector deviation vs. market, and an analyst brief.
0.1USDT / call▸
analyst_brief · samplevalues masked
US semis overnightKorea semis next openwhat US move impliesKorea overshootsthen converges →
US semis move → Korea response
While US desks slept, US semiconductors moved [x.xx]% overnight. On our walkforward-validated transmission ratio, Korean semis should pick up a fraction of that at the Korea open. Today they overshot the implied move by [x.xx]pp — a gap that tends to converge. A measured relationship, not a forecast: if it keeps outrunning the US signal, that reading fails.↪ full brief + coefficients via API
walkforward-validatedout-of-sample · n [x,xxx]R² [0.xxx]claim on threshold
04
Korea Defense Chain
GET /chain/defense
Shipbuilding & defense sector deviation, US federal procurement cross-check, and a brief separating confirmed awards from expectations.
0.1USDT / call▸
analyst_brief · samplevalues masked
US federal awardsKorea order bookupkeepnew-buildexportsbacklogpipelinerevenue is Korea-side
where the revenue actually sits
The demand driving this sector is Korea-side, not US-side. A major Korean shipbuilder was named preferred bidder for a multi-destroyer programme — a selection, not a signed contract. Every confirmed US federal award here is upkeep, not new-build, and Korean ship exports to the US read [x]. So the revenue lives in Korea’s order book — that is what you are buying exposure to. This reading ships with the condition that would invalidate it: an award described as a new build.↪ full brief + procurement ledger via API
ledger-verifiedcontract-stage precisefalsifier open
05
Korea Power Chain
GET /chain/power
Nuclear vs. grid-equipment divergence, US power-equipment transmission signal, and transformer export ledger with annualized run-rates.
0.1USDT / call▸
analyst_brief · samplevalues masked
US power-equipment signalKorea names that respondnuclear — no US linkgrid gear — US-linkeddashed = expected · solid = actual
US equipment signal → Korea makers
Not every Korean power name reacts to the US. US power-equipment moves (think grid-gear and transformer demand) transmit to Korean equipment makers — but Korean nuclear-engineering names do not track the same US signal. The nuclear names lagged their expected move by [x.x] and [x.x] points while the grid-gear names tracked. Buy the US link where it actually carries.↪ full brief + export ledger via API
sector-level βexport ledger annualizedclaim on threshold
06
Korea Equity Chain Bundle
GET /chain
All three axes plus a cross-axis narrative on how US signals transmit through Korean sectors — freshness-annotated per block.
0.2USDT / call▸
analyst_brief · samplevalues masked
How strongly each Korea sector tracks the US signalsemisdefensepowerUS-drivenKorea-only
US-signal strength by Korea sector
One map, three Korean sectors, and how strongly each tracks the US signal you already watch. Semiconductors carry the cleanest US link; defense is Korea-driven; power splits — its grid-gear names follow the US while its nuclear names run on Korea’s own clock. Each axis ships with its own falsifier and freshness annotation.↪ full three-axis narrative via API
3 axes · cross-linkedfalsifier enforced
07
Japan Market Ledger
GET /japan
Customs export ledger by destination, JGB yield curve, and the BOJ & earnings calendar — confirmed facts only, no predictions.
0.1USDT / call▸
trade_ledger · samplevalues masked
Where Japan's chip exports actually ship (share, latest yr)TaiwanChinaHKKoreaUScustoms-confirmed · 6 years
Greater China takes the majority — the US barely registers
Japan’s chip exports don’t go where the tape assumes. Greater China took 56–63% in every completed year since 2021, and 2026 is running above that band; the US takes ~2%. Ships: [xx]% clear through Panama, Liberia and the Marshalls — the flag-state structure, confirmed. JGB 30Y up [x.xx]pp in a year. Every block cites its official source and its next release date. The band and the year that left it ship together, with the figure that would put it back inside.↪ full ledger + JGB curve + BOJ calendar via API
customs ledger 6yrBOJ calendar D-dayband break flagged
04
data & licensing

Every source
is accounted for.

No scraped feeds. Every input is a licensed or public-domain source — disclosed here by category, verified internally.

✓CoinGecko (commercial) ✓US government fixed-income data ✓US federal dollar indexes ✓Government open-data portals ✓Asian government trade statistics ✓Government procurement records ✓US regulatory registers

What we sell is derived intelligence — coefficients, deviations, and narrative — never a redistribution of raw source data. Specific providers are held in reserve; each response ships with an as-of timestamp on every block, and CoinGecko is credited per its commercial terms.

Beyond timestamps, every response states what it measured and what would prove it wrong. When a reading crosses a published threshold — the top or bottom 5% of 250 sessions, or a break from the band a product’s completed years have held — a written claim carries the number, the window and the sample size. Claims that are not yet settled carry a falsifier: the condition that would invalidate them and where to check it. Every product states when its judgment log began — nothing is backfilled.

Give your agent the Asia-hours read.

Discover Asia Pulse on the OKX AI marketplace and pay only for the calls you make.

Use on OKX AI → Use on Agentic Market →
contact

Building an agent
on Asia data?

Questions about the endpoints, high-volume access, or a custom chain (Japan, Hong Kong, and beyond)? Partnerships and press welcome too. Drop a line and we’ll get back to you.

or email contact@asia-pulse.com
No spam. We reply from contact@asia-pulse.com.