Part I asked whether an AMM surface formed. Part II opened the pool. This piece looks at the other state object: posted depth in a CLOB.

Hyperliquid became hard to ignore for more than one reason.

HYPE’s price action helped draw attention. I do not treat token performance as market-structure evidence. Fees are the cleaner hook: in a 2026 H1 DeFiLlama pull (Jan 1–Jun 30 UTC), Hyperliquid logged about $419 million, ahead of Aave and Uniswap in the same peer set.

Reuters, citing CryptoQuant, also named Hyperliquid among major offshore perp venues in a report on 2025 perp volume: $61.7 trillion globally, against $18.6 trillion in spot. I am not merging that headline figure with DeFiLlama daily fees; the scopes differ. It only explains why this is not a random venue pick.

The context is noisy. If marginal capital is crowding into AI, a few high-attention names, or leveraged venues, weak spot activity can look like an AMM problem before it is one. Fees make Hyperliquid worth inspecting. They do not decide the pool-versus-book question.

When the market forms without a pool, what state object replaces it?

Hyperliquid is the case study because its public info API exposes enough book state to make that question answerable.

This is not an AMM death claim

“AMM is dying” bundles too many claims-volume share, LP economics, launch patterns, trader attention. I am not adjudicating all of them.

The useful question is narrower: what market object fits the product?

Callable spot still maps to pools. Perps need margin, funding, liquidation, open interest, and position accounting-state a passive curve does not carry on its own. That split is old in TradFi. What is newer on-chain is inspectability: fee flow large enough to notice, plus public APIs that return book, funding, and OI fields.

For an AMM, the audit starts with reserves and swaps. For a CLOB, it starts with bids, asks, fills, funding, open interest, and whatever slices of the risk engine leak out. Fee rank does not settle that split.

When a non-AMM venue both collects real fees and publishes l2Book, the book stops being only a UI widget. It becomes a partial market object-imperfect, capped at 20 levels per side in Hyperliquid’s case, but callable from outside.

Callable inventory vs posted depth

An AMM and a CLOB fail differently.

An AMM exposes callable inventory governed by a pricing curve. A pool holds reserves and quotes through a rule such as x * y = k. A trader does not choose a counterparty. The pool is the counterparty. That is why Part II could measure execution by running pool state forward-change trade size and execution drag changes; change reserves and executable depth changes; move the external price and arbitrage pushes the pool until the remaining gap is fee-limited.

A CLOB exposes persistent book depth through price-level commitments. There is no single reserve pair that defines the market. There are bids, asks, sizes, spreads, and depth around the mid. A taker consumes posted orders. A maker can cancel, repost, or move inventory elsewhere.

AMM liquidity is committed into a contract. CLOB liquidity is posted into a queue.

The advantage is not that a book clears what a pool cannot. The book’s native fit is conditional liquidity: at this price, for this size, for now. A maker can tighten, widen, cancel, or hedge as inventory and risk change. Queue position and timing become part of the market.

A pool can make its defense layer more active through ranges, fees, and hooks, but it still exposes inventory through a rule. It does not become a native price-time queue.

In the AMM panel, the visible state is compact: a reserve point plus a pricing rule. In the CLOB panel, the visible state is a ladder of price levels. Each level can contain one or more posted orders ( n in the API response is order count at that level, not queue position). The book can change without a trade, because makers can cancel or replace orders.

A small Hyperliquid snapshot

On 2026–06–30T02:09:33.700Z, I pulled l2Book snapshots for four perps on the primary Hyperliquid dex and computed mid, spread, and visible notional depth within ±1% of mid. At snapshot time, spreads were tight- 0.17 bps on BTC, 0.63 on ETH, 0.13 on SOL, 0.15 on HYPE-and visible depth within ±1% of mid ranged from sub-million on SOL to roughly $18-23M on ETH (see Fig. 3).

The public l2Book endpoint returns at most 20 price levels per side. For all four markets in this snapshot, every visible level fell inside the ±1% band-so ±5% depth equals ±1% depth in the processed file. That is a measurement cap, not evidence that depth is flat beyond 1%.

The snapshot is a point-in-time view of what the public book exposed when the probe ran-not a claim about permanent liquidity.

Depth here comes from posted levels around the mid, not from reserves on a curve. In Part II, depth came from the pool state. Here it comes from the book.

The same probe window also pulled funding and open interest from metaAndAssetCtxs—fields with no direct AMM analog. BTC at snapshot time: funding 0.0000125, open interest 35,637 contracts (base units), 24h notional volume $3.12B; mark 60,652, mid 60,652.50, oracle 60,654 (premium −0.000016).

What the book exposes — and what it does not

Each mechanism leaves the outsider a different audit problem.

A public CLOB feed can expose more immediate structure than an AMM pool:

• best bid and best ask (l2Book)
• spread and visible depth around mid (derived from l2Book)
• recent fills (recentTrades – short window; 10 trades returned for BTC in this probe)
• funding and open interest (metaAndAssetCtxs)
• 24h notional volume (dayNtlVlm)
• fee schedule (documented in API; not re-derived here)

The missing side is still important.

The book shows posted depth, not why it was posted, whether a maker intends to stay, or queue priority. On a perp venue, margin and risk-engine state sit alongside the ladder. Liquidation decision paths are only partly visible; the rest stays outside the artifact.

CLOB data is granular and incomplete at the same time.

Dynamic fees change the AMM defense layer

AMMs are not standing still. Uniswap v4-style hooks and dynamic-fee pools make that clear. The old pool design treated the fee as a mostly fixed parameter. The newer design space lets the pool react: a hook can update fees before a swap, add custom accounting, or alter behavior around the swap path. Official docs position v4 hooks as a way to customize pool behavior-not as a path to turn the pool into an order book.

That matters because many AMM failures land on the LP side. The pool offers inventory through a deterministic rule. If the outside price moves first, informed flow can trade against stale inventory. If liquidity is concentrated in the wrong range, the LP absorbs the repricing. If routing or MEV makes the flow toxic, the fee may not compensate the inventory risk. Part II measured that tension through execution drag, arbitrage residual, and fee-vs.-LVR.

Dynamic fees and hooks are one response-Part IV will run that defense layer in simulation. Here the point is narrower.

Dynamic fees are not an AMM victory lap. They are evidence that the old fixed-fee pool was under-defended against toxic flow.

They do not make the pool into a CLOB. They make the pool’s defense layer more active.

On a CLOB, liquidity is already active-makers quote, cancel, repost, hedge, or leave. Failure shows up in stale quotes, queue games, and liquidation parameters more than in a passive curve being picked off.

Where loss lands depends on the mechanism. In an AMM, defense increasingly sits in fees, hooks, and LP strategy. In a CLOB, it sits in maker behavior, margin rules, and liquidation logic-and much of that is only partly visible from outside.

The risk surface changes too

Part II stayed inside AMM mechanics: whether fees cover passive inventory being repriced by arbitrage.

AMM failure tends to land on LP inventory -stale prices, LVR, range placement, toxic flow, sandwich and routing effects, and whether hooks or dynamic fees actually defend the pool.

CLOB failure tends to land on active market making -queue priority, cancellation latency, bad quotes, liquidation cascades, and risk-engine parameters. A perp CLOB adds funding, open interest, margin, and liquidation on top of the book.

Fixed-fee pools are fine for long-tail spot until volatility or toxic flow arrives. Dynamic fees help only if the oracle or proxy is right and the hook is not gamed. Books work when makers stay; they break when makers leave or the risk engine misfires.

AMM and CLOB expose different slices of the market and lose money in different places. Both leave something outside the artifact.

Closing

A market can form without an AMM pool if there is posted depth, matching, and risk infrastructure behind the book. I would not read this probe as industry-wide migration. It shows what becomes visible when the forming object is book-shaped.

The book-shaped side is inspectable enough to audit-incomplete, but not hypothetical. That changes where loss lands and which failures stay inside the artifact.

The question I keep returning to: where does the mechanism fail, who absorbs the loss, and what can an outsider verify?

For an AMM, start with reserves and swaps, then fees and hooks. For a CLOB, start with the book and whatever funding/OI the API returns; assume maker intent and liquidation paths are partly missing.

The pool is not always the object. The audit problem remains.

Appendix: source

• Part I: The token appeared twice. The AMM market formed once.
• Part II: Before MEV, I Built a Rust AMM Lab to Measure Pool State.
• Reproducibility gist — DeFiLlama fees and Hyperliquid info probe.

This post was originally published on my personal blog: https://egpivo.github.io/2026/07/19/when-the-market-forms-without-an-amm-pool.html.

Hyperliquid Shows What a Market Looks Like Without an AMM Pool was originally published in The Capital on Medium, where people are continuing the conversation by highlighting and responding to this story.

Der Beitrag Hyperliquid Shows What a Market Looks Like Without an AMM Pool erschien zuerst auf thanksmia.

Key Takeaways:

• Foundry Digital created a Zcash mining pool which soon had a hashrate of about 30% of the network. 
• Institutional miners are accelerating the compliant, purpose built ZEC infrastructure adoption. 
• A new block explorer, Zcashinfo.com offers real-time information on network activity and mining data. 

Foundry Digital has officially tapped into the Zcash mining ecosystem with excellent promise. With institutional support, its recently launched pool has already taken a good portion of network hashrate.

Foundry Zcash Pool Gains Instant Traction

Foundry announced the opening of its Zcash mining pool, welcoming onboard a number of institutional customers on the first day. Within weeks, the pool scaled rapidly to control around 30% of the network’s total hashrate.

Foundry Zcash Pool is officially live! Since our announcement last month, we’ve seen rapid hashrate growth reaching ~30% of network hashrate. Institutional miners have been looking for compliant, purpose-built $ZEC infrastructure, and we’re proud to deliver it.

Additionally,… pic.twitter.com/GOXyKrqhhH

— Foundry (@FoundryServices) April 13, 2026

The amount of growth is not typical of a new pool. It is an indicator of high demand among large-scale miners who want to have a trusted and compliant infrastructure.

Bitcoin mining services have enabled the company to establish a reputation and it is now transferring the same model to Zcash. The infrastructure of it is oriented towards transparency, auditability, and performance.

Read More: US Senators Push ‘Mined in America Act’ to Secure Bitcoin Mining and Reserve

Institutional Demand Reshapes Zcash Mining

The emergence of the Foundry pool reflects a new trend in the evolution of Zcash mining. Institutional miners are no longer just exploring, they are actively reallocating hashrate.

Compliance and Transparency Drive Adoption

Foundry designed the pool with compliance at its core. They include auditable payouts, real-time reporting, and a well-developed structure of operation.

The aspects are relevant to the regulated entities and to the public companies. They minimize counterparty risk as well as ensuring transparency about earnings and operations. Mike Colyer, the CEO, stated that it aims to provide infrastructure which institutions can count on and help develop the overall Zcash ecosystem.

Zcashinfo.com Brings Real-Time Network Data

Foundry also rolled out Zcashinfo.com, a community block explorer that was created along with the launch of the pool. The platform will offer real-time information about:

• Mining pool rankings 
• Hashrate distribution 
• Block production 
• Network difficulty trends 

Such a tool will provide a better picture of network dynamics to both miners and investors in real time.

Privacy Tech Meets Institutional Infrastructure

Zcash is also one of the most popular privacy-oriented currencies. It employs zero-knowledge proofs in order to confirm transactions without exposing sensitive information.

Its ability to integrate mining infrastructure of institutional grade would enhance its market stand. It demonstrates that even privacy-oriented assets may be in conformity with compliance demands.

Foundry has been received by Zooko Wilcox, who stated that the mining pool as well as the new explorer are both critical infrastructure to the network.

Hashrate Shift Signals Competitive Pressure

The fast emergence of Foundry is already transforming in terms of competition. Other pools might also lose dominance as its share increases resulting in a more spread hashrate.

A more even distribution would make the networks more resilient. At the same time, it introduces new competition among mining operators.

The early success of Foundry Zcash Pool shows that institutional capital and infrastructure are playing a larger role in proof-of-work ecosystems.

Read More: Morgan Stanley Unveils Bitcoin ETP With 0.14% Fee

Source link

Der Beitrag Foundry Zcash Pool Hits 30% Hashrate in Weeks, Signals Massive Mining Shift erschien zuerst auf thanksmia.

Key Takeaways

• Kraken has halted Monero deposits after a single mining pool seized more than 50% of the network’s hash rate.
• XMR last traded near $257, down about 6% over the past week.

Share this article

Kraken has suspended Monero (XMR) deposits after detecting that a mining pool had gained control of more than half of the network’s total hashrate, raising concerns about centralization and network security, according to its status page.

Kraken is actively monitoring the situation and will resume deposits once it determines conditions are safe. Trading and withdrawals for Monero remain fully operational on the platform.

The pause came after Qubic mining pool, led by IOTA co-founder Sergey Ivancheglo, claimed to have achieved control over 51% of Monero’s hashrate, which could theoretically enable block reorganization, transaction censorship, or double-spending attacks.

Ivancheglo said in a statement that the move was intended to expose vulnerabilities and prepare the Monero community for future threats.

Looks like #Qubic has achieved 51% over #Monero, we are waiting for independent confirmations. In the meanwhile #Monero team is polishing details of their 51% attack protection.
Many accused us of being sponsored by 3-letter agencies to attack this anon coin. What do you think…

— Come-from-Beyond (@c___f___b) August 12, 2025

XMR fell earlier this week following the reported attack. The token was trading near $257 at the time of writing, down approximately 6% over the past week, per TradingView.

Share this article

Source link

Der Beitrag Kraken halts Monero deposits after single pool takes over 50% hashrate control erschien zuerst auf thanksmia.