ChainFeeds Research used its latest PRO issue to cover a wide set of topics across market structure, Bitcoin protocol work, Ethereum research and new academic papers. The edition centers on PropAMMs, Lightning node assumptions, a new BIP340 aggregation draft, formal state-preservation models for cross-domain systems, and the question of whether positive-sum market design can move from theory into default usage.
PropAMM and Ethereum market structure
Ethereum Foundation researcher mike and Ritual mechanism design researcher maryam said PropAMM is an important direction if Ethereum wants to bring more trading activity back on-chain.
The report contrasts three models. Traditional AMMs remain composable and transparent, but their liquidity is passive and users can get worse execution because of LVR and stale pricing. RFQ systems bring in professional market makers and can quote better prices, yet execution sits off-chain, liquidity is less transparent and on-chain composability is weaker. Aggregators only mix routes between on-chain pools and off-chain quotes, so they do not remove that trade-off at the base layer.
In a PropAMM design, the contract stays on-chain and keeps DEX-style execution and composability, while professional market makers actively manage the pricing curve. They do that through low-cost, high-frequency parameter updates or oracle updates so the contract can trade as close to real-time prices as possible.
ChainFeeds said PropAMMs already account for a sizable share of trading volume on Solana. The report ties that to short block times and low update costs, which let market makers pay high per-unit fees to place price updates ahead of user trades and cut the risk of stale-price arbitrage.
Ethereum can run PropAMMs as well, but under tighter constraints. A 12-second slot raises the chance that prices go stale. Under proposer-builder separation, builders also have an incentive to censor a market maker’s update transaction and then sell the stale-price arbitrage opportunity to searchers.
The issue notes that builders such as Titan and Quasar currently offer trusted channels that can help market makers land update transactions in blocks. It also says that setup can produce an unstable experience and may deepen builder concentration. The report’s conclusion is narrow but clear: PropAMMs can improve on-chain execution quality on Ethereum, but that requires anti-censorship guarantees at the top of the block within the same slot. FOCIL may only partially relieve the problem, while in-protocol ACE may be a more fundamental fix.
Bitcoin protocol progress and Lightning node assumptions
In the Bitcoin section, Anony addresses what the report describes as a common misunderstanding: using the Lightning Network does not mean a user must personally operate a node that stays online around the clock, and a self-custodial mobile wallet is not automatically unsafe.
The reason, according to the piece, is that channel security comes from the penalty mechanism. If a counterparty tries to close a channel unilaterally with an old state, the other side can respond within the timelock window and claim the funds. In practice, the security requirement is not perpetual uptime. It is that offline periods must not exceed the reaction window set for the channel. Users can also adjust parameters or rely on watchtowers for monitoring.
The main drawback of a node that is not always online is the receiving experience. A payee usually needs to be online to update channel state and release the payment preimage, so a mobile wallet may require the user to open the app in time. ChainFeeds says LSP support combined with system notifications can help, and asynchronous payments could eventually reduce the need for the sender and receiver to be online at the same moment.
An always-online node is better suited to forwarding payments and routing, but it comes with remote connectivity challenges, especially in places where public IP addresses are scarce or network access is restricted. The report points to NWC, which uses Nostr relay as a more flexible remote-control method. Its takeaway is that ordinary users do not need to avoid mobile Lightning wallets over security fears, while always-online nodes and mobile nodes each involve trade-offs.
This part cites Bitcoin Optech Newsletter #415.
BIP340 full signature aggregation draft
The issue also highlights a new BIP draft from Fabian Jahr that would introduce the DahLIAS full aggregation signature standard for BIP340 Schnorr signatures.
Under the proposal, multiple signatures could be combined into one fixed 64-byte aggregate signature that does not grow with the number of signers. The protocol is interactive and uses an untrusted coordinator to reduce communication complexity.
The process runs in two rounds. First, participants collect and aggregate signer public-key nonces. Then they collect partial signatures and combine them into the final signature.
The report says one possible use case is cross-input signature aggregation, which could reduce the size of multi-input transactions and lower on-chain fees. Any consensus change required for that use case sits outside the scope of this BIP.
Ethereum research on cross-domain state preservation
In the Ethereum research section, Oraclizer CPO Jay Kim argues that cross-domain systems such as bridges, rollup exits, shared sequencers and permissioned settlement chains need a verifiable framework for state preservation rather than broad claims that state has simply been synchronized.
The core idea is to abstract state migration between different chains or systems into composable state-machine mappings and then prove the properties mechanically in Isabelle/HOL.
Each domain is modeled with state, actions, a transition function and terminal conditions. A cross-domain mapping has to preserve not only state but also actions. If a mapping from a rollup to L1 and another mapping from L1 to a settlement chain each satisfy state-preservation conditions, then composing the two mappings yields an end-to-end preservation guarantee.
The framework also satisfies identity mapping and associativity, letting more complex cross-domain systems be broken into single paths and verified step by step.
It adds synchronization levels as well. An asset can declare how many chains it needs to preserve state consistency across, while a system can declare the highest level it supports. A formal safety guarantee is only available if the system’s capability meets or exceeds the asset’s requirement. Otherwise, the system should not claim full cross-domain state preservation.
ChainFeeds says the value of the framework is that it turns cross-chain security from an experience-based judgment into a provable interface. The report says that is especially relevant when regulatory states such as freezes, restrictions or confiscation have to move across domains without being downgraded by mistake during transfer.
Positive-sum microstructure as an adoption problem
Developer william Glynn argues in the issue that over the past 15 years crypto has gradually pushed through bottlenecks in execution performance, zero-knowledge proving costs and cross-domain interoperability. In his view, the factor now limiting the sector’s ability to realize its distinct value is not throughput, proving or bridging, but incentive design.
His claim is direct: as long as default mechanisms still reward MEV, speculation, liquidity capture and attention farming, gains in technical efficiency will only make extraction faster rather than push markets toward positive-sum coordination.
The problem, he writes, is not that positive-sum mechanisms cannot be built. It is that more neutral mechanisms struggle to become the default path. He uses a commit-reveal uniform-price batch auction as the example. Users first submit hash(order || secret) to lock an order, then reveal the order and secret in a later phase. Ordering inside the batch is determined by a random seed generated from all revealed secrets, and trades clear at a uniform clearing price.
In that structure, searchers cannot buy adjacency to a victim and cannot exploit price changes inside the batch to sandwich a trade. The main source of sandwich profits is removed at the structural level.
Still, the report says these mechanisms have limits. They mainly remove within-batch ordering rent. They do not eliminate cross-batch MEV, oracle manipulation or delay games at outer layers. Honest behavior is also only an approximate equilibrium under large-market assumptions, not a dominant strategy at every scale.
Protocols such as CoW Protocol have already put similar designs into practice, which in the report’s view shows that the bottleneck is not the absence of solutions. Instead, liquidity and attention continue to flow to the underlying AMMs where extraction remains more profitable. It adds that many batch-auction protocols shield users from direct sandwiching while still relying on pools such as Uniswap for a large share of execution, leaving a protective wrapper around extractive infrastructure rather than replacing it.
The conclusion is that crypto’s final bottleneck is the adoption of positive-sum microstructure, not just its design. The report says what has to change is the default liquidity path, the metrics used to judge success and the way attention is allocated. Trading volume alone is not enough; the more relevant question is whether a mechanism reduces extraction and improves cooperative equilibrium. If positive-sum mechanisms gain traction in channels that care about rigor and practical results, that would help prove their value. If they still fail to advance after full testing, the claim should also be treated as falsifiable.
MEV section
The issue says there were no important updates this week in the MEV section.
Paper of the week: Duet
The paper section features Duet: Co-Optimizing P2P Message Propagation and Rotating-Leader Consensus.
The authors are from Shanghai Jiao Tong University, Shanghai Academy of Future Internet Technology, Beijing Academy of Blockchain and Edge Computing, and Fudan University.
The work studies the disconnect between the consensus layer and the P2P network in blockchain systems and proposes Duet, a design that treats network topology and node latency as coordinated state information to optimize rotating-leader consensus.
The approach includes optimizing proposer order, using tree-based propagation under normal conditions to improve efficiency, falling back to gossip when needed, and building propagation trees based on latency.
In experiments across 10 regions and as many as 300 nodes, throughput improved by up to 7.26 times compared with traditional gossip, according to the report.