ZN Partners · Published Work

The real, the raw and the dirty

What you never knew you didn't know about market data.

Every price on every screen looks like a fact. It's actually a verdict — the end of a long chain of decisions about what counts, whose clock rules, and which version of history survives. Most people who work with market data every day have never seen the whole chain.

Market Data: A Structural Guide — the new book by Dan Solak

Twenty-three chapters from venue mechanics to machine consumption: why datasets legitimately disagree, why precision is not truth, and why the cleaner data looks, the more interpretation has already occurred. Grounded in the events practitioners remember — negative oil, the franc de-peg, busted trades, the day Berkshire broke Nasdaq's integer.

Cover of Market Data: A Structural Guide by Dan Solak

Why this book exists

The confusion is structural, not personal

Market data powers trading systems, risk models, analytics platforms, compliance workflows, and research pipelines. It is bought, sold, normalized, enriched, corrected, and archived at enormous scale. Entire industries exist to produce it.

And yet, among professionals who work with it daily, there is persistent confusion about what it actually represents. That confusion is not incompetence. It is structural.

Market data is presented as a set of facts — prices, volumes, timestamps, identifiers. Discrete fields that can be stored, queried, compared. In practice those fields are the end product of processes involving human intent, system design, regulatory compromise, and irreversible choice. The further data travels from the interaction that created it, the more objective it appears, and the less it resembles the event it claims to record.

This book does not propose a correct version of market data. No such version exists. Its purpose is narrower and more useful: to make the hidden decisions visible again, so they can be examined rather than inherited.

Truth in market data is not a property of the data alone. It is a function of when — and under which regime — that data is interpreted.

What's inside

Twenty-three chapters

The book follows data from the moment of interaction to the moment it is treated as settled fact — and examines what is decided at each step.

Part One

Structure precedes data

What market data actually is, how venues and interaction models determine what can be observed, and how data changes meaning as it flows from creation to consumption.

Chapters 1–3

Part Two

The control stack

Capture, timestamp choice, normalization, eligibility, corrections, versioning, aggregation — the sequence through which interpretive range narrows and commitment becomes irreversible.

Chapter 4

Part Three

Time, motion, and meaning

Latency as distribution rather than delay; condition codes and the fragility of meaning; auctions, halts, and market states; and what it costs to decide when "when" actually happened.

Chapters 5–8

Part Four

Identity and the boundary of truth

Symbols and referential stability, reference data as moving static, and the boundary where real-time observation gives way to historical definition.

Chapters 10–12

Part Five

Analytics, vendors, and infrastructure

From ticks to analytics, versioned truth as a commercial product, pipeline quality, bi-temporal integrity, and how source strategy shapes what can later be known.

Chapters 13–17

Part Six

Entitlements, governance, and scale

Permissioned reality and the partitioning of access, why governance frameworks arrive after control has already accumulated, and what markets might look like if designed today.

Chapters 18–21

Part Seven

Machines and inherited assumptions

AI as a visibility amplifier early on, and later as structural pressure — including what it means for machine participants to encode, and remove, interpretation they never examined.

Chapters 9, 22–23

Who it's for

Written for people who already sense the problem

Chief data officers and heads of data accountable for something they cannot fully inspect
Quantitative researchers whose results depend on decisions made upstream of them
Architects and engineers building capture, storage, and reconstruction systems
Vendors and venues who must explain what their data does and does not represent
Regulators, auditors, and governance teams working with data whose provenance is opaque

It is not an introduction to markets, and it is not a technical manual. It assumes you already work with this material and have noticed that the standard explanations do not quite hold.

The author

Dan Solak

Dan has spent more than two decades working with financial market data — in content and distribution, platform architecture, and the design of large-scale capture, mastering, and reconstruction systems. He co-founded ZN Partners, where he writes on data and market structure.

Market Data: A Structural Guide is his first book. It is published independently through ZN Partners.