Research Advisor @titanbuilderxyz | @blockspaceforum | fmr. head of research @ChorusOne
Switzerland
- Tweets 569
- Following 206
- Followers 743
- Likes 1,745
SBC impressions ->
- Good energy from researchers; VCs stuck in infra<>token limbo & jaded.
- MCP likely has a timing game that will be much more competitive than PBS, probably lower impact for users, TBD.
- Maybe also possible for validators to outperform on tx fee capture under MCP, optimization vector is the same as for the timing game.
- Excited about @monad_xyz and not the only one. MEV infra for Monad will be competitive and has a lot of brainpower behind it.
- Preconfs took a backseat to the MCP narrative. MCP probably kills proposer execution preconfs, proposer inclusion may get easier, builder preconfs TBD (# concurrent proposers probably the deciding factor).
- Decentralised AI currently deeply unimpressive.
- Always meet the same people at @aori_io events, that’s a feature, enjoy talking to them. Nice event.
- Post-Covid NYC better than its reputation.
1
46
4,299
Searcher.wtf was great.
Props to the @aori_io team, and hopefully the first iteration of many.
3
2
45
4,157
Wasn’t robbed of my enthusiasm in Brussels ->
- Events were good. @aori_io @titanbuilderxyz @SorellaLabs stick out.
- Preconfs are an unlock with clear problems; fun.
- Many narrative plays, sure. Ignore & they’ll (infl-) fade away.
- Plenty opps. to build disruptive tech still. Eg personal currencies & other unexplored classics.
- Too many side events, community fragmentation etc: proxy points to say, there’s VC money. Not a negative.
1
6
45
5,106
Started looking at @berachain / PoL. It's a cool concept.
Best way for validators to distribute incentives will be via a public algorithm, computing on public inputs.
We'll build it.
Goal -> Maximize user welfare, minimize shitstorms, minimize slippage
2
2
36
7,567
1/ Our view of the mev-boost auction (v.01) -> adagio.chorus.one/
Tracks the PBS auction to simulate the payoffs with- and without timing games for our setup.
The long-term upside settles around ~10%, which is a comfy risk-adjusted spot.
Observations below ->
1
3
31
4,380
Quick(est) status quo of @monad_xyz MEV by popular interest.
Challenge: Monad decouples execution from consensus.
Implication: As validators reach consensus on the tx in block n, they are still executing block n-1.
Upshot: The available state could be at n-2 still.
Two ways to play it on the infra-side ->
1. Via an off-chain auction done in a probabilistic manner. @aprMEV is building this. More: nitter.app/aprMEV/status/18077710…
2. Via an on-chain auction through smart contracts, also probabilistic. @0xFastLane is building this. More: nitter.app/ThogardPvP/status/1822…
Probabilistic just means that transactions may revert given that the state is still in flux, and have to be priced accordingly.
Both great teams. Current intuition is that these approaches could co-exist, time will tell.
On the searcher-side, probably more top teams have started exploring than for any other new ecosystem.
On the validator-side, positive there'll be a competitive edge around latency. Maybe initially also around hardware, but that should reach parity quickly. Other idiosyncracies to be seen. Keen for the testnet.
Multiple Concurrent Proposers (MCP) is a fascinating spec that I fully support. That being said, some MEV strategies may interact suboptimally with it.
I'd like to share some searcher alpha, some monad alpha, my thoughts on MCP, and what they all have to do with FastLane.
1/n
2
3
30
3,521
Enjoyed my conversation with @DeGatchi
-> Timing Games, Solana MEV, DEX orderbook MEV mitigation…
.. and an aside on fish tanks & lava lamps as a source of randomness.
2
3
28
4,516
Why delegated preconfs vibe like a bad idea -
Goal / heuristic -> Preconf implementations should not create a centralization bottleneck that exceeds the current builder dominance.
Claim -> Delegated gateway designs will fail this goal.
Basic reasons and an alternative below.
#1 Multi-block MEV
- Multi-block MEV is ~superlinear.
- Gateways compete on proposer revenue.
- This defaults to, which gateway can extract most value from information.
- Sophisticated gateways will get large.
- Large gateways will be able to share more revenue with proposers.
-> It's a winner take all dynamic. OFACs exacerbate it.
#2 Switching Costs
- The PBS tx pipeline fundamentally is an auction.
- A delegated gateway design is not a spot market; it's more akin to a contract.
- The minimum switching cost is something like, interacting with an on-chain registry for each validator.
-> Gateway integrations will be sticky.
-> It's a BD game, think private order flow.
-> Sophisticated gateways will outperform.
Alternative: Proposers as preconf providers
The ability of a proposer to capture multi-block MEV is capped by it's voting power, which is capped by access to capital.
Proposers are not set up for MMEV - organizationally, legally, flow-wise, model-wise..
Stake is sticky, and this is unlikely to change. Case: Timing games haven't meaningfully manifested centralization.
The catch to the alternative
Sophisticated proposers will still outperform, but they won't earn superlinear returns to scale.
There'll be accessible ways of out-earning peers. One simple example could be just not preconfing tx-types that increase in value over time, e.g. stat. arbs.
@umb_nat will have more on this soon - stay tuned, exciting stuff.
Excited to see where it goes.
A great place to follow the cutting edge is the Bolt preconf repo run by the chads at @chainbound_, @fra_mosterts @mempirate @merklefruit et al.: github.com/chainbound/bolt/d…
More soon.
In summary -
1. Large gateways would be able to outbid others as they'd realize superlinear returns on multi-block MEV.
2. Large gateways would benefit from switching costs, proposer signups would ~feel like sticky private order flow.
3. Proposers as preconf have their returns to scale capped by voting power, which is capped by access to capital.
1
6
26
3,094
Heading to NYC for SBC on Monday.
Things I’m working on atm. -
- Preconfs & role of the proposer / information games.
- Berachain PoL / optimal allocation
- Timing games etc
- Personal currencies / pet interest
…and anything infra / MEV.
If interesting, DM to connect.
1
22
3,756
EthCC plan tomorrow:
> Do a panel on restaking risk / cost management at the @gauntlet_xyz restaking day from 15.45 - 16.20
> Sprint across the street / context switch.
> Do a panel on block sizing incentives at the @titanbuilderxyz event from 16.30 - 17.00
..Lfg
1
23
1,809
Bullish @symbioticfi
Felix is a great guy and true OG..:
Excited to share that I’ve joined @symbioticfi as the Head of Ecosystem to help build out a maximally permissionless and flexible shared security primitive.
4
22
1,126
Great to be on-board ⚔️
Two years ago, we open sourced our fork of the Solana client, capturing MEV in-validator.
Excited to see @uriklarman & the team develop and support this approach, which will be a net benefit to the ecosystem.
Throwback -
- Our repo: github.com/ChorusOne/solana-…
- Our paper: chorus.one/reports-research/…
Paladin⚔️ - an ENTIRELY new way to address MEV
here's how Solana validators can make more $ by NOT front-running
1/6
medium.com/@uri_61495/introd…
2
3
18
1,775
Spent the last 3 days talking preconfirms at @ZuBerlinCity.
Glad we proposed some of the first-ever preconfs. Very cool..!
But mostly excited about what's still cooking.. 🕶️
based preconfs are live on devnet 🕶️
→ preconf modules by @chainbound_, @primev_xyz
→ commitment sidebar by @Commit_Boost
→ gateway operators @EspressoSys, @titanbuilderxyz
→ relay operators @titanbuilderxyz, @ultrasoundmoney
→ staking operators @ChorusOne, @Kiln_finance
→ devnet devops @BarnabasBusa, @parithosh_j
→ dashboard viz @ascii_g1rl, @jacobdcastro
→ coordinators @DrewVdW, @kubimensah, @tzhen
→ ZuBerlin crew @cshg0x, @marc_nitzsche, @pascalstichler
→ special guests @AmirOnchain, @lvdanil, @macrosam
based teamwork 🕶️
1
3
16
1,925
Started to explore / build out infra for our @berachain PoL algo.
Drafted up a toy dashboard to visualize things - check it out.
Early Observations -
-> Infra seems performant. Can query the RPC via multiple threads & it stays responsive.
-> BEX is a fork of crocswap, now @ambient_finance, TIL! That’s nice.
-> Still early. Eg seeing signs of what may be user agent tests in the LP-token naming. Maybe you can spot it in the image.
Dashboard is not final in any way, just play at this point. It runs a live fetch, screenshot is a snapshot after startup.
-->Overall, early days still, but feels exciting.
Started looking at @berachain / PoL. It's a cool concept.
Best way for validators to distribute incentives will be via a public algorithm, computing on public inputs.
We'll build it.
Goal -> Maximize user welfare, minimize shitstorms, minimize slippage
2
2
17
3,008
Convinced preconfirms will massively impact node operators and Ethereum.
The chain with the biggest NO set and most value at stake is the natural clearinghouse to get NOs to execute actions -> ~ wherever.
Very early, but that's a plus for fun & profit.
2
1
15
701
Very much enjoyed working on this with @umb_nat, and feel great about the future of @dYdX.
To our knowledge, the first deep study of validator-driven MEV & mitigation in a truly decentralized CLOB DEX.
Chorus One proudly introduces our latest research endeavor, powered by a grant from @dYdX. This effort delves into the implications of MEV within the context of dYdX v4 from a validator's perspective. 🚀
As dYdX's imminent v4 heralds a transformative phase in the protocol's evolution, it marks the industry's first genuinely decentralized perpetual futures exchange.
Our paper accomplishes the following:
️⚡️It examines the MEV dynamics on v4, highlighting the critical need for objective MEV quantification to enable effective mitigation.
⚡️It pioneers the first-ever exploration of mitigating negative MEV outcomes within a fully decentralized, validator-driven order book.
⚡️It explores the uncharted territory of cross-domain arbitrage involving a fully decentralized in-validator order book and other trading venues.
⚡️It proposes concrete strategies aimed at levelling the playing field in derivative trading by addressing undesirable MEV behavior.
Download and dive in here: chorus.one/reports-research/…
1
3
14
1,647
We've been very active on Solana MEV as well.
@ethalita_ with the definite history of our open source Solana client fork enabling fair, backrun in-validator MEV extraction.
Pre-packaged with a triangular arbitrage module.
A brief story of MEV on Solana
- I subscribed to twitter to share this :D
I've joined @ChorusOne 2 years ago to specifically work on a @solana MEV research with @umb_nat and @vmulps. Our main goal: measure impact of sandwich attacks and suggest a healthy/sustainable approach to MEV given the specificities of the protocol.
It was early days, @jito_sol was close to launch a mempool model, and we could never imagine that @FTX_Official episode was about to happen...
During @SolanaConf 2022 in Lisbon we presented the result of our research:
The whitepaper "Breaking Bots: MEV on Solana and how to prevent frontrunning, spam attacks and centralization" and the solana-mev client - an open source modified validator client for "good" MEV extraction - basically arbs. Our main conclusion:
"[...] a block building marketplace would introduce latency and weaken Solana’s unique position as the go-to network for low-latency applications. Furthermore, as a central component the block building marketplace poses major centralization and censorship risks. [...]"
Of course, this was a MVP, and we stopped maintaining the client after some time to focus on our expertise - improve performance of our validator infrastructure.
March'24 @jito_sol discontinued the mempool - I assume they arrived to a similar conclusion based on real data. - Thanks Meme Gods! Being a fan of both the protocols and teams I couldn't avoid the feeling.
@ChorusOne is known by advocating and putting efforts into researching the mitigation of negative externalities of MEV. And now we could - finally - start running the @jito_sol client, knowing that voting power wouldn't be used to exploit users.
In March we were contacted by this team of MEV searchers who was looking for validators to run their fork of jito to extract MEV (enabling back sandwich attacks). "5MM in Stake from small validators" they said was already committed to it. I was devastated. Of course, we didn't join.
For us - node operators - profit is not always guaranteed, and I understand it can be tempting to engage in such practices. But...
🍄 monetary value can't be the only thing to be maximized. What should be your operators' values?
My favorite quote is from @toly: "do not stake with validators you don’t trust". Couldn't agree more.
A validator's voting power is given by the tokens delegated to it. So the whole ecosystem is responsible for setting up a culture - Foundation is only one of them.
🍄 what's the incentives to run a Solana node, the most expensive chain to run, with a 0% fee?
Complete the sentence: "If you're not charged for the product, ________".
Tbh, I've never been so thrilled about Solana. The fact that the @SolanaFndn is taking action against this, proves maturity... it raises awareness to the whole ecosystem, triggers a form of social consensus, and generates a cascade effect in the different pieces of this complex economic system.
Yes, I'm an optimistic! - I know that's what you're thinking :)
2
1
14
918
After cancelling my initial flight, @lufthansa has now waitlisted me for the third (!) alternate over the span of two days.
There are probably ~15 ppl in the same situation.
Epically bad performance for passengers and the ground staff forced to deal with the fallout.
Airline delayed my flight for a day, and I’m waitlisted.
Maybe this is an applied lesson on preconfs.
12
5
1,099
1/ Some quick alpha on the timing games & Solana discussion.
We ran the numbers on node geographic distribution in ~Feb.
Results did indicate that there was concentration around ~BeNeLux.
Specifically, in one NL datacenter.
Nodes in this datacenter seemed to do well -> but..
1
2
12
1,169
Live dashboard showcasing our in-validator view of the mev-boost auction.
Compares our Adagio mev-boost fork to a standard setup.
Landing soon alongside the methodology.
Will cover longer time-horizons.
Last week was a very strong week.
1
2
12
1,451
Great panel on MEV, LVR, and all things incentive alignment with @0xMagmar @tkstanczak @ThogardPvP @maxresnick today.
Thanks organizing @a41_allforone - cool event!
3
12
807
Even more on Solana soon.
Meanwhile, throwbacks to ->
- Some alpha on Solana geolocation: nitter.app/mostlyblocks/status/17…
- Our paper on in-validator MEV capture: chorus.one/reports-research/…
"If all validators performed like Chorus One, Solana’s overall transaction capacity could increase by 11.4%."
@ChorusOne @solana chorus.one/articles/metrics-…
1
12
1,240
Cool stuff from a great team.
We’ve started looking into pricing out preconf txes.. early days but very interesting..
1/ We are thrilled to soft-launch Bolt, a trustless protocol that allows Ethereum proposers to offer credible commitments about their block contents. Bolt V1 unlocks trustless inclusion preconfirmations, a new primitive that aims to enhance Ethereum’s UX and applicability.
11
831
1/13 Adagio, our mev-boost fork, has been publicly available for 100 days now.
We have been testing it on mainnet since early 2023.
A quick thread on timing games and how we found them, rational competition, and resilient decentralization 🧵
🥳Adagio, @ChorusOne's #Ethereum MEV-client has crossed its 100 day milestone!
🎖️During this period, Adagio extracted a total of 7.84% extra MEV vs. a "vanilla" client.
1
3
11
2,650
Excited about preconfs, and Commit-Boost as a way of making these accessible.
@drakefjustin had great framing -> "Docker for proposer commitments"
1) Two months ago, I left my job to contribute to Ethereum.
Today, I was excited to present further details around Commit-Boost, a validator platform for proposer commitments.
piped.video/watch?v=jrm4ZUoj…
1
1
11
1,796
Enjoyed bouncing thoughts on OEV with @capitalgrug!
Very cool design space, early still, now is a good time to dive in.
For details, check out our intro to OEV below. Props to @gabriellassh @0xPranav.
-> chorus.one/articles/an-intro…
Was great to speak with @mostlyblocks about OEV and the state of oracles today. Do give this one a read when you get the chance
1
10
1,294
Will present our current perspective & paper on timing games at EthCC -
-> Thursday, 11.40. Silver Hall, Core Research Track.
-> The Paper: chorus.one/reports-research/…
Mayb also a first perspective on preconfs.
Catch you there.
1
10
1,702
commit-boost becoming widely adopted software is one of my high conviction preconf predictions.
At a minimum, it's at feature parity with mev-boost, at better performance.
The vision of faciliating preconfs / last mile communication is ofc much larger and feels on-point.
1/ First Commit-Boost community call took place yesterday. Strong attendance from across the Ethereum community and great conversation!
- Notes: docs.google.com/document/d/1…
- Video: piped.video/watch?v=jM3uL00U…
- Agenda: github.com/Commit-Boost/pm/b…
1
10
1,181
Solvers compete on three things -
- Private order flow.
- Private liquidity.
- Routing.
Routing is the least sustainable edge.
Private order flow starts with gated order flow (i.e. has all OFACs) and from there best path is likely a private RFQ system for tail assets that don't trade on CEXes.
Some very clever teams are working on this.
1
10
1,061
Enjoyed this panel!
Good discussion beyond shallow takes.
Thanks all @tonydenyer @mcutler @terencechain @MuratLite, and @titanbuilderxyz for hosting.
4
10
1,799
The basket risk of the AVSs a node operator runs is not a linear combination of individual risk scores.
We're describing a first model for selecting an optimal combination of AVSs by combining modern portfolio theory with an MCMC model.
Props to @gabriellassh @umb_nat
💯 @eigencloud is the hottest ticket in crypto right now. With over $14B ETH in TVL, delegations & LST restaking enabled, it's time for a deeper conversation about its potential risks from a node operator's perspective.
Our latest report "Optimal Risk & Reward of EigenLayer: A First Look" aims to spark this vital discussion through data-driven analysis. 🔬
LINK: chorus.one/reports-research/…
🧵👇
3
9
1,012
Delivering shortly ->
BeraBoost v1 is our algorithmic approach to emission allocation on @berachain
Purpose is to remove the guesswork of discretionary allocation.
Goals -> runs on public data, lean, imposes a cost on curve-type manipulation, maximizes delegator revenue.
Approach -> maximize revenue over BGT emissions to delegator gauges and incentives.
Intuition -> delegators could receive one of two reward baskets:
1. BGT emissions in a way fully congruent with their gauge positions + incentives from these gauges.
2. BGT emissions in a way not fully congruent with their gauge positions + incentives from all gauges.
One key challenge is how to meaningfully compare different incentive assets.
Another key challenge is meaningfully differentiating organic delegators from Curve-type "mercenary delegators".
-> Landing soon.
Started looking at @berachain / PoL. It's a cool concept.
Best way for validators to distribute incentives will be via a public algorithm, computing on public inputs.
We'll build it.
Goal -> Maximize user welfare, minimize shitstorms, minimize slippage
2
1
9
1,422
Enjoyed this panel!
Beyond to the discussion on risk quantification, one shared spicy take was that many AVSs so far have been lackluster; we collectively look forward to more powerful applications of re-staking.
🎥 Managing Costs & Maximizing Returns for Operators
@a41_john, @aimxhaisse, and @mostlyblocks discussed strategies for maximizing returns, methods for controlling costs and evaluating the expenses in restaking operations on a panel moderated by
@CarsonMBrown at Restaking Day during @EthCC.
Watch the full panel here:
1
9
766
My talk from EthCC is up
Goal -> concise overview of proposer timing games, and an encore on what initial preconf games could look like.
Full paper: chorus.one/reports-research/…
One of my favourite talk, very interesting
ethcc.io/archive/The-cost-of…
Thanks @mostlyblocks
2
7
813
We'll do our best to roll our @berachain testnet performance over to mainnet, but no promises.
6
423
Cool talk and result!
- Self-built blocks do not feature private order flow and artificially depress the base fee.
- These are typically followed by mev-boost “recovery blocks” that subsequently significantly raise the gas fee.
-> Decoupling between the base fee & spot demand for blockspace if vanilla & mev-boost blocks are mixed.
-> Divergence expected to increase as the share of private order flow increases.
-> Another reason is that builders have more time to provide a block vs. a typical vanilla proposal.
Keen to read the study.
A weaker result would also extend to proposer timing games, where builders have more transactions available as the slot progresses.
Graph from our timing games study below.
Looking forward to contributing to Metagame Day at @modular_summit. My session on “Gas Matters: New Sources of Base Fee Volatility” starts at 3:20pm and promises to be 🌶️🌶️🌶️ collective.flashbots.net/t/m…
1
1
8
3,053
The TEE builder discussion boils down to, is this a good way of attracting order flow.
The narrative is self-fulfilling, the more people talk about TEEs, the more likely wallets / bots are to sell their flow to a TEE builder, making them feasible.
Not a bad strategy.
1
6
365
Lufthansa needs more skin in the game. Passengers get a single compensation claim for the first bricked flight, subsequent fails are freerolls
Should be an AVS ;)
6
2
241
Our Adagio mev-boost fork captured an additional 17% MEV rewards vs. a vanilla setup yesterday.
Volatile prices pay disproportionately for timing games, incl. due to stat. arb. inventory risk management.
When there's a lot of risk, trading late is even more attractive.
1
1
7
616
We're hiring for a quant. research analyst to contribute to deep research on MEV and staking dynamics more broadly.
Check out the job posting for examples of our past work.
Contacts, introductions, questions - and applications - highly appreciated: careers.chorus.one/o/quantit…
7
7
1,853
Glad to see our study on the risk / return aspects and externalities of timing games cited ;)
Cool analysis!
New analysis on the impact of timing games on Ethereum's consensus online!
ethresear.ch/t/on-attestatio…
The TL;DR is:
-> timing games cause missed attestations
-> the proposers of Kiln cause the biggest problems
-> attestation timing varies across Lido, Coinbase, and Kiln
6
479
@scrapingbits podcast in the books.
Fun conversation with @DeGatchi ranging from MEV mitigation and infra-optimization to lava lamps as a source of randomness. Coming soon.
1
2
7
1,708
Key lesson from weekend projects:
The implementation shapes the theory way more than vice versa.
Just start building straight away and see where things go.
1
6
248
Iterated our mev-boost fork to version 2.0. Turns out the mev-boost auction is not always timed consistently.
We now carry razor-sharp parameters into the auction every time by automatically adjusting for this. Less risk for the us and the network, at a higher payoff.
From here on out, would expect the most interesting further optimizations to hit client software / hardware / networking.
6/ Here's a sneak-peak of the results so far:
ELI16: Adagio 2.0 narrows the window in which we solicit blocks, allowing us to be much more precise about latency parameters, which reflects as less risk and higher rewards.
6
995
Featuring stats on our ETH mainnet validator-driven MEV pilot.
Our Q3 Quarterly Insights is out! Arguably our most expansive yet, this edition offers a deep dive into the key concepts shaping the crypto and staking world today.🔥
Dive in below! chorus.one/reports-research/…
5
194
We've iterated on timing games for a long time, in public.
Would be wary of introducing new complexity to offset a centralization risk that hasn't really manifested.
Everybody connected to Ultrasound or Titan (relay-side timing games) already has a baseline setup.
A sophisticated setup does much better, but there is a lot variance, and the upside takes time to crystallize on a chart.
Especially if you look at it through a risk / reward lens.
-> It's one dimension validators compete on, but does not make a client case by itself.
-> Restaking yield may actually be a larger factor overall.
-> Returns on sophistication also exist for attestations.
-> Solo validators are passion projects.
Agree with the call on shorter slot times increasing the relative upside; it's mostly a networking optimization.
Here's our current view on which optimizations pay off - nitter.app/mostlyblocks/status/18…
1/ Our view of the mev-boost auction (v.01) -> adagio.chorus.one/
Tracks the PBS auction to simulate the payoffs with- and without timing games for our setup.
The long-term upside settles around ~10%, which is a comfy risk-adjusted spot.
Observations below ->
5
647
Many AVSs vibe like rushed marketing plays.
Things like abandoned to-do lists in the source code, et cetera.
Ethereum's economic security comes in at ~$100bn.
Let's not sell it short.
@ChorusOne is charting a unique path in @eigencloud AVS onboarding.
Instead of following the crowd, we're prioritizing a strategic and very selective approach. 🚀
Our CEO, @crainbf, elaborated on this forward-thinking strategy on the 'Future of Liquid Staking and Restaking' panel at @token2049 Dubai (📹👇🏽).
In the coming days, we’re publishing our full AVS Selection Framework, covering the economic + engineering criteria we evaluate to provide a great risk/reward profile. Stay tuned 👀
#Eigenlayer #Restaking
1
5
273
Recommended listen.
You won't find anybody more knowledgeable on BTC blockspace than @Leozayaat
sat down with @Leozayaat of @alkimiya_io for @OnTheBrinkCIV to cover bitcoin blockspace markets and how users can take long and short views on BTC fees, and its effect on bitcoin
onthebrink-podcast.com/alkim…
5
440
This is V1 of our mev-boost fork. V2 will land ~soon🚀
In the past 7 days, the @ChorusOne Vault on StakeWise has produced TWICE the average Ethereum staking yield🤯
This is the power of a custom MEV strategy in the right hands: 100% more rewards than usual.
Don't miss out - available only on StakeWise 💎
app.stakewise.io
5
274
Excited to present our work on validator timing games & externalities at ETH Denver today: 5pm, Neptune Stage.
For a sneak peek, check out @umb_nat's excellent presentation from a recent Lido community call: piped.video/watch?v=J_N13erD…
3
5
976
2/ We're outperforming our original estimate by a lot.
Timing games are mostly an optimization over the hardware, networking, kernel layer.
The mev-boost fork itself is generally downstream.
Though some moated, non-obvious additions to mev-boost are possible:
nitter.app/ChorusOne/status/17882…
6/ Here's a sneak-peak of the results so far:
ELI16: Adagio 2.0 narrows the window in which we solicit blocks, allowing us to be much more precise about latency parameters, which reflects as less risk and higher rewards.
1
4
194
Also, my take on solvers aged quite well -
Solvers compete on three things -
- Private order flow.
- Private liquidity.
- Routing.
Routing is the least sustainable edge.
Private order flow starts with gated order flow (i.e. has all OFACs) and from there best path is likely a private RFQ system for tail assets that don't trade on CEXes.
Some very clever teams are working on this.
4
417
5/ The future is multi-chain.
Solana is a great network that we're committed to spending more time on.
Ethereum is a great network that we'll continue to spend a lot of time on.
Preferences vary and competition is healthy.
1
3
189
We’ll be live with an algorithm maximising delegator incentives over their gauge positions and (any) liquidity-adjusted incentives on day 1.
Convinced algorithmic is the most informed, least biased way of tackling it.
V1 Methodology is done & will be public soon..
4
286
Not a fan of delegated preconfirmations.
The most sophisticated party offering this to proposers will capture ~all flow and be able to trade multi-block MEV seamlessly by subsidising.
Even sophisticated proposers are capped by their stake and rarely propose twice in a row; operators proposers delegate to have no cap.
4
318
Live in ~2 hours. Tune in!
Don't miss Wednesday's Youtube Live featuring @ChorusOne, @SkipProtocol, @dYdX Research and dYdX Foundation 👀
3
160
Highly recommended reading ->
Initial discussion about auction dynamics for preconfs, through our node operator lens by @umb_nat.
Node operators can play for-profit games around timing & selection, or potentially capture on-chain MEV in-house.
More coming soon..
As promised, we delve into some dynamics may arise in the MEV landscape if validators are the providers of preconfirmations
ethresear.ch/t/preconfirmati…
4
452
Bunch of commercial AVS risk frameworks coming out right now.
Having an independent model will pay off.
Popular frameworks will tend to over-adjust, independent of their "correctness", just as a function of popularity.
Maybe an AVS' risk / reward profile is skewed at a certain security budget.
If a popular model then pulls the AVSs, the security budget can reduce excessively, giving a reasonable r/r deal to the remaining stake.
Can "counter-trade" by running an independent model.
On the flipside, for the AVS, having several independent risk lenses helps stabilize stake.
Separately, rolling internal parameters to an external parameters is likely not possible to an optimal extent. For a proprietary model, whenever something meaningfully changes, run it ad hoc.
Overall, expect the best approach to be a combination of an internal model and whatever external best practices are floating around.
Sanity checking each other and accounting for correlated behavior.
We welcome diverse perspectives & aim to foster an inclusive discussion.
Join us in exploring the optimal risks & rewards of EigenLayer, ensuring a secure & transparent future for Ethereum staking.
Read the full report: chorus.one/reports-research/…
1
1
4
854
18 6-hop arbs in this dataset today.
That's much more than the average (e.g. 0 on the last three days).
This address seems to be the only active searcher capable of executing these:
osmo123kgn90rmdcrhas5fu2anfvnyrhvgurlk7j4es
1
1
3
Excellent article on based sequencing, a form of shared sequencing, and preconfirmations.
Expect to hear much more from @mempirate and the @chainbound_ team
Very ready to become a "sophisticated active monopolist", wherever on the activity spectrum the final design may land..
(1/3) We’ve been doing a lot of research into based rollups, which has lead us to examine the tradeoffs around utilizing the l1 proposer for sequencing and preconfirmations. While based sequencing is beautifully simple and decentralized, introducing preconfs complicates things👨🍳
2
254
3/ It's fair to say that validator returns are on a risk<> return axis.
Centralizing optimizations like co-location increase your individual return, while the additional risk is effectively socialized across the network.
1
3
113
Very cool read on MEV infra in a parallel execution future.
Probabilistic block building is incoming, and @aprMEV will lead the way.
Excited we're backing @sing_me_a and the team, and for what's still to come.
1/ 🧵 how will the MEV landscape evolve in the parallel execution era under Monad?
0xapr.substack.com/p/mev-lan…
thank you to @0xsudogm @buffalu__ @_charlienoyes @cmspeary @lstephanian @odysseus0z @mostlyblocks @segfaultdoctor @sui414 for the review and discussion
1
3
642
We stand ready with prop. data on validator cost structure, rev. impact from timing games & other returns on sophistication, and generally anything else that helps model out an optimal validator set.
3
159
We did run an approach like this for a while via a forked Solana client & backrun arbitrage tx.
Here’s the paper: chorus.one/reports-research/…
Here’s the repo: github.com/ChorusOne/solana-…
Bullish on the direction and where @uriklarman will take it.
4
65
First Eigenlayer tail risk may have materialized two weeks ago, when the roof of the building hosting the Restaking event in Dubai collapsed during the networking reception.
3
163
A range is fair, aligns expectations while leaving room for negotiation, usually saves everyone time.
1
3
140
Very cool collaboration- CoW trades at zero cost in whitelisted balancer pools, which then earn the surplus from the CoW solver competition.
Offsets LVR for LPs, fees for CoW, and works out at a net win<>win when it attracts new liquidity to both.
An important step in making LVR-resistant AMMs the new dominant AMM design: CoW protocol and Balancer collaborating to develop CoW AMMs! Check out the proposal here: forum.cow.fi/t/cip-draft-bal… @CoWSwap @Balancer
1
2
295
Great paper applicable to liquid crypto, titled "Ponzi Funds".
Buying tail assets pushes prices up, inflating fund returns, leading to inflows into the fund, causing a Ponzi-like upside for early fund investors.
Empirical case that flows chasing self-inflated returns predict bubbles.
Model-wise, linear regression on fund inflows and past returns. Decomposes into self-inflated returns and fundamental returns.
Ponzi flows are the product of the coefficient of the flow-performance relationship and the price impact of the fund's buying.
Highly recommended:
arxiv.org/pdf/2405.12768
1
3
245
4/ Heuristic is that shorter block times lead to less MEV but higher returns on timing games given the same absolute latency advantage.
Significant absolute latency advantages are possible by optimizing networking.
Where this lands in terms of absolute extraction is noisy.
1
1
3
184
2/ The datacenter was also one of the few locations that still had bottlenecked high end hardware available.
You would expect validators that seek out this level of optimization to outperform.
Delineating this upside from returns on co-location takes time due to variance.
But the immediate course of action is clear.
1
2
121
Airline delayed my flight for a day, and I’m waitlisted.
Maybe this is an applied lesson on preconfs.
Heading to NYC for SBC on Monday.
Things I’m working on atm. -
- Preconfs & role of the proposer / information games.
- Berachain PoL / optimal allocation
- Timing games etc
- Personal currencies / pet interest
…and anything infra / MEV.
If interesting, DM to connect.
1
1
3
1,733
We (Chorus One) were a launch partner for the Titan relay, ie have tested it for a while.
It’s just a very modern, very efficient implementation.
Deployed in quite a few locations as well.
If you were running equidistant to all relays, I’d bet it’d be the fastest currently.
1
3
387
6/ If you want to repeat the experiment:
- Calling getClusterNodes gives you all validators on the network and their network address.
- Run the network addresses through an IP profiler.
- Here's a great script that we forked for this analysis: github.com/jnaulty/solana-va…
1
2
165
8/ We are aware that timing games carry negative externalities but believe that ignoring a prisoners dilemma is not solution-oriented.
Decentralization has to be incentive-compatible or turn into a LARP.
1
1
83
5/ Learn more:
Our study on the PBS auction, latency games, externalities ->
chorus.one/reports-research/…
Longer thread on how we found timing games in early 2023 ->
nitter.app/mostlyblocks/status/17…
1/13 Adagio, our mev-boost fork, has been publicly available for 100 days now.
We have been testing it on mainnet since early 2023.
A quick thread on timing games and how we found them, rational competition, and resilient decentralization 🧵
2
296
Stand-out result.
Glancing at the data, the ATOM pump is prob. the reason for this.
MEV swaps featuring LUNC are down ~50% vs. yesterday.
Goes to show the importance of liquidity (ATOM/OSMO pool leads the pack by far).
1
1
2
How will validators be incentivised? Specifically, is there some kind of performance-based incentive like eg transaction fee capture for the (common) txs a validators brings?
Just broadly, how is good validator performance measured & incentivised?
2
119
If I catch your drift correctly-
These last 100ms are likely worth a maximum if arbitrageurs know they're incoming, due to inventory risk.
If arbitrageurs don't know, then I'd say value tends to cluster at the expected block submission time.
Attaching some empirics from our timing games study: chorus.one/reports-research/…
1
1
2
420
One toy idea at EthCC was to run a “relay reputation” scheme in mev-boost.
Define desirable attributes & their weights, derive a preference factor for each relay, and weigh its bids by it.
More of a thought experiment ofc but I like the approach of imposing margins instead of a binary decision (..which is just a sub-case with max. weighing anyways).
cc @danielkivanov who has thought about this type of thing.
1
2
229
Cool read on an underrated corner of cryptoland
1/ Thrilled to release a first empirical study about Proof of Personhood: “Compressed to 0: The Silent Strings of Personhood" with my esteemed co-authors Mikhail, @cryptoidentity, founder of Idena Protocol, and @_paulaberman.
papers.ssrn.com/sol3/papers.…
Why does PoP matter?
2
219