Running volume activity on OpenBook

What OpenBook is

OpenBook is a central limit order book running on Solana. Participants place bids and asks at chosen prices, those orders rest on the book, and a trade happens when a new order crosses an existing one.

This is how conventional exchanges have always worked and it is unlike the pooled model that dominates decentralised trading on Solana. There is no reserve, no constant product formula and no automatic price. Price is simply the level at which somebody is currently willing to transact, and if nobody is willing, there is no price at all.

The distinction matters here because volume tooling is almost universally written against the pooled model. A tool that assumes a swap is one instruction with a deterministic outcome will behave badly on a venue where an order can sit unfilled for hours.

What differs from an AMM

The last row is the one that catches people. On a pooled venue a failed operation is loud: an error, a rejected transaction, an obvious signal. On an order book the equivalent is silence, because an order that never fills produced no error and no result, and a tool not designed to notice will keep going as though everything worked.

The broader conceptual comparison, including why each model suits different kinds of markets, is set out in the AMM and order book explainer.

Fills, partial fills and cancellations

Three behaviours have no AMM equivalent and all three need explicit handling.

Resting orders. An order placed away from the current level sits on the book publicly. Anyone can see it, including anyone forming a view about what is producing activity on the token. A pattern of identical orders appearing at regular intervals is considerably more legible on a book than the same pattern on a pool.

Partial fills. An order can execute against part of the available size and leave a remainder outstanding. Any accounting that assumes orders are binary will drift, and the drift compounds across a campaign until the reported and actual positions diverge noticeably.

Cancellation and settlement. Unfilled orders need cancelling and filled ones need settling, both of which are transactions with their own costs and their own failure modes. This is ongoing state management that a pooled venue simply does not have.

None of this is difficult. It is simply a different problem, and tooling written for the pooled model will not have solved it.

What it costs

The cost structure is genuinely different rather than merely differently sized, which is why comparisons that quote a single percentage are misleading.

There is no pool fee taken as a percentage of every swap in the way the DEX fee comparison describes for AMM venues. Instead there are transaction costs for placing, cancelling and settling, plus account rent for the open orders account, plus the spread paid when crossing the book.

The spread is the cost that matters most and the one that varies most. On a market with tight two-sided quoting, crossing costs very little. On a thin market with a wide spread, crossing costs far more than any AMM fee would have, and a strategy that places many small crossing orders on a thin book pays that spread repeatedly.

The practical implication is that order book economics reward patience and punish urgency, which is close to the opposite of the pooled model where every swap costs the same percentage regardless of timing.

There is a subtler cost that does not appear in any fee table and is worth naming. On a pooled venue the counterparty is a formula that never changes its mind. On a book the counterparty is a person or a program that can withdraw, and depth visible a second ago may not be there when your order arrives. Quoted depth on a book is therefore an indication rather than a guarantee, and strategies sized against it will occasionally find far less available than the screen showed. That gap widens precisely when conditions are volatile, which is when it is least convenient.

Configuring for an order book

1. Confirm the market exists and has two-sided depth. A market with orders on only one side is not tradeable in both directions regardless of what the listing suggests.
2. Size against visible depth, not against a percentage. The book shows exactly what is available at each level, which is more information than a pool provides and should be used.
3. Handle partial fills explicitly. Any process that treats an order as filled or unfilled will misreport, and the error accumulates.
4. Budget for cancellation and settlement. These are real transactions with real costs and they are frequently omitted from cost models built for pooled venues.
5. Vary placement. Identical orders at identical intervals are visible to everyone reading the book, and a book is easier to read than a transaction feed.
6. Monitor unfilled orders as a first-class outcome. On a pool, nothing happening means an error occurred. Here it means the market did not come to you, which requires a different response.

When it is the right venue

Honestly, for most new Solana tokens it is not, and the reason is simple: no market exists. Creating one is a deliberate act that most launches never take, and a venue with no market cannot be used at any price.

Where OpenBook genuinely suits is tokens with an established two-sided market, participants who want price control rather than immediate execution, and strategies where resting orders are the point rather than an inconvenience. Market making in the proper sense belongs here and belongs poorly on a pooled venue.

Where it suits poorly is anything that needs guaranteed execution, anything that depends on screener visibility given the inconsistent indexing coverage, and anything where the operator wants to configure a rate and stop thinking about it.

The general rule, which our own venue measurements published on the measured data page support across the board, is that activity should go where liquidity already is. A token whose real trading happens on a pooled venue gains nothing from activity on a book nobody is watching, and the effort is better spent on the venue where the buyers already are.