Running live, dailySteps 1 and 2

Options market analytics

Live tick data from a fragmented crypto options market, turned into analytics you can trace. Our own agents run the routine every day; a person watches them.

MarketCrypto options and derivatives
DataExchanges, ETF options, RFQ and OTC
RunsDaily, internal operations
IndicatorsAbout 40 live, daily
Full analytics[options site] · coming
The problem

One market, spread across venues that don't fit together.

Bitcoin options trade on crypto exchanges, on decentralised venues, as listed ETF options in the US, and in size through RFQ platforms and OTC desks. Each speaks its own language. A screen shows only one piece of it.

Shown on the analytics siteCollected for our own analysis
Crypto options exchangesDeribitCoincall
Decentralised venueParadex
Listed ETF options, USIBIT options (Nasdaq ISE)
RFQ and block tradingParadigm, SignalPlus [names to clear]
OTC desksQuotes under agreement
Reference dataBinance spot and perpetualsEquity and macro series
Data layerOne traced recordEvery quote keeps its venue, instrument, time and conventions, so venues can be compared like for like.

Same market, different shapes.

One month of implied volatility as three venues quote it, and as one surface once every quote is on common terms.

Illustrative · not market data
Venues are anonymised. The shapes follow a typical month in our data; the values are not market data. One colour scale for every venue and every day.
What we built
  1. 01Collectors for each venue, including a low-latency system for Deribit tick data (CVC '25).
  2. 02A market data layer that brings the venues onto common terms and keeps every source.
  3. 03About 40 market indicators, computed and checked every day by our own agents.
The agentic idea

Not one agent, but many, each with a defined job.

All of them work on the same traced data. The routine runs on the loop; what leaves the house passes a person in the loop.

On the loop: agent runs, a person sets limits, gets alerts, can stop itIn the loop: a person approves every result
MarketOptions across venuesQuotes and trades from exchanges, ETF options, RFQ platforms and OTC desks.
Data layerTraced market dataNormalised, timestamped, every record linked to its source. Results flow back as internal data.
Agents, each with one job
Observation watches the market and flags unusual moves.
Analytics builds volatility surfaces, skew and term structure.
Backtesting reruns models on fresh data, on a schedule.
Reporting drafts reports and controlling figures.
[Agents per process to confirm]
A person
On the loopSets limits, gets alerts, can stop any agent.
In the loopApproves every report before it goes out.
ReceiversInternal processingReaders of the analytics site
How we work: data first, then agents

Where each part stands

Market observation and analyticsStep 2Agents run it; a person watches.
Recurring backtestingStep 2Agents run it; a broken limit raises an alert.
Reporting and controllingStep 2Internal; agents run it, a person watches.
Research and evaluationStep 1AI-supported; a person approves every result.
Reports that go outStep 1A person approves every one.
Evidence

Published, and running.

All publications
IEEE Crypto Valley Conference (CVC '25) · 2025A Low-Latency System for Collecting Massive Crypto Option Tick Data from Deribitdoi.org/10.1109/CVC65719.2025.00013
In operationCollection, analytics and backtests run every day as part of our own operations.
Full analytics

The options analytics get their own site.

About 40 indicators on volatility, skew, open interest and market depth, each with its source, time and definition. Built on the same traced data, by GDTQ.

[options-site.tld]Coming [date]. Link goes live with the site.
Preview · three of about 40 indicatorsIllustrative · not market data
ATM implied volatility41.2%
BTC · 30d · Deribit · 17:00 UTC
7 days38.9%30 days41.2%60 days42.7%
1-year low1-year high
Implied minus realised+9.6
BTC · 30d IV vs 30d RV, vol points · daily
Implied, 30 days41.2%Realised, 30 days31.6%Realised, 7 days28.4%
1-year low1-year high
Options open interest, 24h change−$180M
BTC · notional · Deribit · 24h window
Calls−$260MPuts+$80MShare near the money39%
1-year low1-year high

Want this kind of data layer for your market?

We start with a research project and build the product beside your existing business.

Start a research projectNext project: QAVCM