Competing Auctions in Intermediated Markets
summary
The gist
We analyze competing auctions in intermediated markets where a seller selects among parallel mechanisms for the sale of a single good…
In short
The study examines competing auction systems used by a seller choosing between different mechanisms for selling a single good, specifically focusing on Ethereum's proposer-builder separation. It finds that if an in-protocol first-price bidding channel exists, it severely restricts the design possibilities for out-of-protocol intermediaries like relays. This suggests that relying on integrated protocols limits the flexibility of alternative auction designs.
Key concepts
- Relay and Protocol Architecture
- This refers to Ethereum's structure where a seller (proposer) separates the process of proposing an action from the process of building or executing it. The paper analyzes how this separation impacts how different intermediaries, like relays, can conduct auctions for a single good.
- Auction Unraveling
- This concept describes how competing auction formats behave when one option becomes available. The study shows that under certain conditions (like single-homing or multi-plexing), one format—the sealed first-price auction—can completely eliminate the others, forcing them to collapse into that structure in equilibrium.
- Single-Homing vs. Multi-plexing
- These describe different market environments for bidders. Single-homing means bidders are restricted to one path, while multi-plexing means they can use multiple paths simultaneously. The results show how the availability of an in-protocol channel affects these different market structures.
- Leakage Resistance
- This relates to whether a seller can safely share sensitive information, like bids. The paper finds that a first-price auction is resistant to leaks from one fast bidder but not others. If the seller can credibly promise not to leak, it is optimal for them to do so.
Terminology used across episodes
This episode discusses
The paper
Competing Auctions in Intermediated Markets · Read on arXiv
Bruno Mazorra, Minghao Pan, Christoph Schlegel
Flashbots · California Institute of Technology
Transcript
Introduction to the show: ident: Computer Science. The day in computer science: software, systems and theory on arXiv.
Tom: We're looking at the paper: "Competing Auctions in Intermediated Markets".
Jane: Here's the gist of the paper. We analyze competing auctions in intermediated markets where a seller selects among parallel mechanisms for the sale of a single good, most prominently the relayand-protocol architecture of proposer-builder separation in Ethereum.
Tom: This analysis suggests that the availability of an in-protocol (first-price) bidding channel severely limits the design space available to out-of-protocol auctions by relays and other intermediaries.
Jane: Market Structure and Competition The paper investigates competing auctions in intermediated markets, where a seller chooses among parallel mechanisms for selling a single good, focusing on the relayand-protocol architecture of proposer-builder separation in Ethereum.
Tom: The context involves competition between intermediaries for a single good, contrasting with classical literature where competition is usually between sellers.
Paper summary: Tom: So, we’re looking at this paper called "Competing Auctions in Intermediated Markets," and it’s about how sellers choose between different ways to sell one single thing, especially in these Ethereum relay-and-protocol setups. It focuses on proposer-builder separation.
Jane: Right, so the main idea is that when you look at these competing auction structures, there are some very specific outcomes depending on whether the intermediary can enforce single-homing on its bidders or if it’s an open bidding format.
Lu: Basically, they find that if the intermediary can force everyone to use only one path, like a sealed second-price auction, it completely unravels into a sealed first-price principal auction. But with an open bidding format, it only partially collapses into a first-price equilibrium when latency is symmetric or when the intermediary sorts fast bidders to themselves when latency is asymmetric.
Meng: That makes sense for production systems. If you can enforce a single path, the complexity drops significantly because you know exactly what structure you’re dealing with. But what about those multi-plexing environments where there’s no enforcement by the intermediary?
Lalam: The paper extends these results to those multi-plexing environments too, suggesting that the availability of an in-protocol first-price bidding channel pushes the entire market toward that same auction structure regardless of whether there is enforcement or not.
Tom: That leads into a big question for us: this in-protocol first-price channel, this thing they're talking about, it severely limits the design space available for out-of-protocol auctions run by relays and other intermediaries.
Jane: Exactly. The paper suggests that if you have that sealed first-price channel available within the protocol itself, it makes designing different auction mechanisms for external relays much harder.
Lu: I think the key is in how they handle information sharing during the auction period. They look at what happens when the seller or intermediary chooses to share bids received during that time to orchestrate a kind of partial open-bid auction.
Meng: What’s that about bid sharing? How does that actually affect the economics for the seller? Is it just some theoretical exercise, or is it something we have to worry about at scale?
Tom: Well, they find that sharing information with just one fast bidder isn't optimal for the seller. But if there are two or more fast bidders, sharing those bids becomes a best response for the seller, and those fast bidders actually benefit from that bid disclosure.
Jane: It sounds like the incentive structure shifts depending on how many fast bidders you have in that specific auction setup. This connects back to some literature on information disclosure in auctions.
Lu: They also look at last-look advantages, which is when a proposer has an incentive to leak the bids it has received, and they find that if there are at least two fast bidders, the proposer would have an incentive to leak those bids.
Meng: Okay, so if we’re building a system where we want maximum efficiency in this bidding process, and we know there are multiple fast actors, should we be worried about leakage or is it something to design around?
Paper summary: Lalam: If the seller can credibly commit not to leak bids, then it’s actually optimal for them to do that anyway. It makes the situation more predictable if you have that commitment in place.
Tom: That commitment is what they call a credible sealed bidding channel, which is essentially an in-protocol first-price channel. And it’s this availability of that thing pushing the market toward a specific structure, even when we don't have enforcement.
Jane: It seems the whole point is that this in-protocol mechanism fundamentally changes what kind of auction designs we can even consider for off-protocol intermediaries like relays.
Lu: The institutional reading suggests that proposer-builder separation will probably reshape the relay market by limiting the design space for relays running second-price or open bidding mechanisms.
Meng: So, if I’m looking at deploying a relay service, this paper implies that relying on an open-bid mechanism might be less viable if we can't guarantee that sealed first-price channel availability.
Tom: It points toward a potential role for relays as a credible intermediary and for services like fast bid and block propagation because of this structural constraint.
Jane: And the viability of maintaining an open-bid relaying service really depends on how much competition exists in the builder market itself. That’s where the real uncertainty lies.
Lu: The paper does have some limitations, though, they mention not modeling collusion among builders or the searcher-builder pipeline upstream of the auction. Also, they treat each slot in isolation rather than as part of a repeated game.
Meng: So it stops short on things like how builders might cooperate to affect the outcome over time, and it doesn't look at that whole chain of events leading up to the auction itself.
Tom: It also assumes valuations are drawn independently and identically distributed from a distribution F that is absolutely continuous and strictly increasing on a compact interval. That’s a pretty strong assumption about the underlying bidder behavior.
Jane: So, for someone just listening, what does this mean practically? It means when we design these systems, we have to be very careful about whether we can guarantee that sealed first-price channel is available in protocol or not.
Lu: If you can’t guarantee it, you’re likely stuck with those partially collapsed open-bid auction results under asymmetric latency.
Meng: For deployment purposes, I guess this means we need to focus our engineering effort on building systems that can credibly commit to not leaking bids if we want the seller to get the best outcome from that in-protocol structure.
Tom: Exactly. The core message of "Competing Auctions in Intermediated Markets" is about how this specific architectural choice—proposer-builder separation—tightly constrains the kinds of competitive auction designs we can actually build on top of it.
Jane: It’s a warning about the design space, showing that an in-protocol sealed-bid first-price channel is a very powerful constraint on what external mechanisms can do.
Conclusion: Tom: So we've been looking at these competing auctions in intermediated markets, specifically focusing on how Ethereum’s relay and protocol separation sets up sellers to choose between different selling mechanisms.
Jane: It boils down to this paper, "Competing Auctions in Intermediated Markets," by Armstrong and others. The main thing they show is that the availability of a sealed first-price channel within the protocol really restricts what kinds of external auction designs relays can use.
Lu: They looked at how things unravel under different conditions. For instance, if you force single-homing on bidders, a sealed second-price intermediary auction completely disappears and turns into a sealed first-price principal auction.
Meng: That sounds clean, but what about the rest of the cases? Like when things are multi-plexed and there’s no enforcement by the intermediary? Does that structure still hold up?
Lalam: Yes, they extend those results to multi-plexing environments too. The availability of that in-protocol first-price bidding channel pushes everything toward that same auction structure, regardless of whether there’s enforcement or not.
Tom: That leads us to the core idea here. If you have this sealed first-price channel built into the protocol, it severely limits the design space for off-protocol intermediaries like relays running second-price or open bidding mechanisms.
Jane: And that means if you're designing a relay service, relying on an open bid might be less viable if you can't guarantee that sealed first-price channel is available.
Lu: The institutional reading suggests this architecture will reshape the relay market by limiting the design space for those external auction designs. Proposers might forward bids to an open-bid feed, but there’s value in credibly committing not to leak bids through reputation mechanisms or some secure execution environments.
Meng: So if we want that best outcome from the in-protocol structure, we need a way for the seller to actually commit not to leaking information. That's a big engineering hurdle.
Tom: Exactly. The paper concludes that second-price intermediary auctions unravel fully against a sealed first-price principal auction under both single-homing and multi-plexing conditions.
Jane: It really paints a picture of how the internal protocol structure dictates the external competition we can even think about designing for. What this means practically is that we need to be very careful about whether we can guarantee that sealed first-price channel is available in protocol or not.
More episodes
- 2505.16254-Human-AI Writing as Distributed Cognition: Reassessing Collaborative Writing Theories in the Age of LLMs
- 2506.10913-Choreographic Quick Changes: First-Class Location (Set) Polymorphism
- 2507.14812-A Black-Box Approach for Exogenous Replenishment in Online Resource Allocation
- 2507.15448-Frames over finite fields with respect to the Galois inner product and their connections to self-orthogonal and LCD codes
- 2508.21572-NewsReX: An Open-Source Multi-Framework for Neural News Recommendation
- 2509.17028-Impact of Packetization on Network Calculus Analysis
- 2510.17342-AoA Services in 5G Networks: A Framework for Real-World Implementation and Systematic Testing
- 2511.04088-Efficient and rate-optimal list-decoding in the presence of minimal feedback
- 2511.17264-Notes on Stack Machines and Quantum Stack Machines
- 2601.15318-On the closest balanced game