Abstract
In financial applications, latency advantages - the ability to make decisions later than others, even without the ability to see what others have done - can provide individual participants with an edge by allowing them to gather additional relevant information. For example, a trader who is able to act even milliseconds after another trader may receive information about changing prices on other exchanges that lets them make a profit at the expense of the latter. To better understand the economics of latency advantages, we consider a common-value auction with a reserve price in which some bidders may have more information about the value of the item than others, e.g., by bidding later. We provide a characterization of the equilibrium strategies and study the welfare and auctioneer revenue implications of the last-mover advantage. We show that the auction does not degenerate completely and that the seller is still able to capture some value. We study comparative statics of the equilibrium under different assumptions about the nature of the latency advantage. Under the assumptions of the Black-Scholes model, we derive formulas for the last mover's expected profit, as well as for the sensitivity of that profit to their timing advantage. We apply our results to the design of blockchain protocols that aim to run auctions for financial assets on-chain, where incentives to increase timing advantages can put pressure on the decentralization of the system.
Introduction
In high-frequency trading and blockchain environments, latency is often viewed as a technical hurdle. However, from an economic perspective, latency represents a significant informational advantage. A participant who can submit a bid even milliseconds after their competitors gains a clearer view of the asset's "true" value, particularly when that value is derived from external, highly liquid exchanges. This paper explores the "last-mover advantage" in common-value auctions, providing a mathematical framework to quantify how timing affects equilibrium strategies, auctioneer revenue, and system decentralization.
The Auction Model: Alice vs. Bob
The study centers on a game-theoretic model featuring two bidders: an early mover (Alice) and a last mover (Bob). The asset in question has a "common value" $v$, meaning it has the same intrinsic value to all participants (like a share of ETH or a Treasury bond), but that value is volatile and follows a stochastic process, such as Geometric Brownian Motion (GBM).
The interaction unfolds in two distinct periods:
- Period 1: Alice submits her bid based on the initial distribution of the asset's value.
- Period 2: Bob observes the realized value $v$ at a later timestamp. Crucially, while Bob has superior information about the asset's value, he cannot see Alice's bid before submitting his own.
To prevent Bob from perfectly exploiting her, Alice cannot rely on a single deterministic bid. Instead, the equilibrium requires Alice to employ a mixed strategy. By randomizing her bids across a specific range, she introduces uncertainty that prevents Bob from simply bidding a fraction higher whenever the asset value exceeds her bid.
Equilibrium and the 50/50 Win Rate
A surprising result of the authors' Perfect Bayesian Nash Equilibrium analysis is the distribution of wins. Despite Bob’s significant informational edge, he does not win the auction more frequently than Alice; in equilibrium, both participants win with a 50% probability.
However, the nature of their wins differs significantly. Alice wins when the realized value $v$ is lower than expected, often paying a price that reflects the "winner's curse." Bob, conversely, wins "better." His informational advantage allows him to capture "informational rent," winning primarily when the asset's value is favorable relative to the bidding range, effectively turning his timing advantage into a portfolio of financial options.
Bridging Auction Theory and Option Pricing
One of the paper's most significant contributions is the mathematical mapping of auction profits to classical derivative pricing models. The authors demonstrate that Bob’s expected profit is functionally equivalent to a financial option.
Specifically, when the auction has a zero reserve price ($L=0$), Bob’s profit $\Pi_B$ mirrors the payoff of an exchange option, calculated using Margrabe’s formula: $$\Pi_B = \mathbb{E}[\max(v_1 - v_2, 0)]$$ where $v_1$ and $v_2$ represent independent draws from the value distribution. This connection allows researchers to apply the Black-Scholes model to quantify the exact value of latency in dollars and cents.
Revenue Implications and Timing Pressure
A common fear in mechanism design is that informational asymmetries will cause an auction to "unravel," leading to zero revenue for the seller (the "lemons problem"). The authors show that the inclusion of a reserve price (or limit price) $L$ protects the auctioneer. While the last-mover advantage does extract value from the seller, the auction does not degenerate completely, and the seller still captures significant value.
The paper introduces the concept of "Timing Pressure," defined as the sensitivity of the last mover's profit to their timing advantage (the "Theta" of the latency). The analysis reveals that:
- In a competitive environment, the incentive to delay a bid is significantly higher than for a monopolist.
- Timing pressure peaks as the reserve price $L$ approaches the initial market price $p_0$.
- As the latency advantage $T$ increases, Bob’s profit grows at the direct expense of the auctioneer’s revenue.
Conclusion and Impact on Decentralization
The findings have profound implications for the design of blockchain protocols and decentralized finance (DeFi). The high "timing pressure" identified in the model creates a natural centralizing force. Because milliseconds translate directly into profit, participants are incentivized to co-locate their infrastructure or consolidate power within validator sets to capture last-mover advantages (often referred to as Maximal Extractable Value, or MEV).
To maintain decentralization, future protocol designs must account for these timing games. Potential solutions include "leaderless" auctions or privacy-preserving "mempools" that hide bid amounts, effectively neutralizing the informational edge of the last mover and protecting the integrity of the price discovery process.
Summary updated