Optimal Rebate Design: Incentives, Competition and Efficiency in Auction Markets

Thibaut Mastrolia and Tianrui Xu · 2025

Read Optimal Rebate Design: Incentives, Competition and Efficiency in Auction Markets by Thibaut Mastrolia and Tianrui Xu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-12591.txt.

Introduction

Thibaut Mastrolia and Tianrui Xu study how a stock exchange should pay the dealers who keep its auctions liquid, in a 2025 paper. In their model, the exchange charges fees to big investors who trade blocks of a risky asset. It then hands fees back as rebates to dealers who post standing orders. Dealers shape both the trade price and how often trades arrive.

The team frames the setup as a boss dealing with many agents. They give clear conditions for when the dealers settle into a stable balance of strategies. The exchange's own best plan is written as a hard equation, which they solve with a neural method. Their results show that well-set fees and rebates pull the auction price closer to the asset's true value. Both sides still gain. The work speaks to market design.

Catalog record

CREATORThibaut Mastrolia and Tianrui Xu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE111 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-12591

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Optimal Rebate Design: Incentives, Competition and Efficiency in Auction Markets Thibaut Mastrolia and Tianrui Xu Abstract This study explores the design of an efficient rebate policy in auction markets, focusing on a continuous-time setting with competition among market participants. In this model, a stock exchange collects transaction fees from auction investors executing block trades to buy or sell a risky asset, then redistributes these fees as rebates to competing market makers submitting limit orders. Market makers influence both the price at which the asset trades and their arrival intensity in the auction. We frame this problem as a principal-multi-agent problem and provide necessary and sufficient conditions to characterize the Nash equilibrium among market makers. The exchange’s optimization problem is formulated as a high-dimensional Hamilton-Jacobi-Bellman equation with Poisson jump processes, which we solve using a verification result. To numerically compute the optimal rebate and transaction fee policies, we apply the Deep BSDE method introduced in [ 33 , 31 , 17 ] . Our results show that optimal transaction fees and rebate structures improve market efficiency by narrowing the spread between the auction clearing price and the asset’s fundamental value, while ensuring a minimal gain for both market makers indexed to the asset’s price on a coexisting limit order book. Keywords: Microstructure, market design, auctions, Nash equilibrium, stochastic games, stochastic control, market making, HJB PDE. 1 Introduction Auction markets have diverse applications, ranging from online advertising and stock trading to electricity markets (see [ 50 ] for an overview of these applications). The profound theoretical and practical contributions to auction theory by 2020 Nobel laureates Robert Wilson and Paul Milgrom [ 49 ] highlight the vital role of market design. In their foundational works, [ 63 ] and [ 51 ] studied the conditions under which an auction market yields a final price that closely approximates the true value of the asset being auctioned. Later, they advised the U.S. Federal Communications Commission and contributed to the implementation of simultaneous multiple-round auctions for spectrum license allocation. This paper explores auction market design for stock exchanges, aiming to improve the efficiency of modern financial markets. The central limit order book (CLOB) operates as a continuous double auction, where participants submit their desired prices and quantities to the exchange. The term “double” refers to the dual participation of buyers bidding to purchase shares and sellers offering to sell them. A trade occurs the moment a buyer’s bid meets or exceeds a seller’s ask. In contrast, a batch auction is a trading mechanism where orders are collected over a fixed time interval and executed simultaneously at a single clearing price, rather than being matched continuously in real time. This process aggregates all buy and sell orders submitted during the batch period, determining a uniform price that maximizes the total executed volume based on supply and demand. These two mechanisms underpin the primary trading systems in modern markets: the continuous limit order book and the periodic auction system. The CLOB functions in real time, continuously accepting new orders and matching them instantly when a suitable counterparty is available, either at a specified price (limit orders) or at the best available price (market orders). For empirical evidence, price formation models, and stochastic control problems related to limit order books and market making, we refer the reader to [ 1 , 27 , 14 , 29 , 26 , 13 , 2 , 5 ] and the monograph [ 11 ] . By contrast, a periodic auction operates through a sequence of repeated batch auctions. When a trader submits an order, an auction phase is initiated and remains open for a fixed duration, ranging from a few milliseconds to several minutes depending on the exchange. During this period, market participants can submit, modify, or cancel orders. At the close of the phase, the exchange computes a clearing price that maximizes the executed volume. All executable orders are then filled at this single clearing price, regardless of their original submission times. A new auction begins once the previous one concludes. Unlike the continuous system, periodic auctions deliberately defer execution to the end of the interval. This approach has gained growing traction among traditional exchanges, with notable implementations including Cboe U.S. Equities Periodic Auctions, Turquoise (operated by the London Stock Exchange Group), and CME Globex. 1.1 Periodic auctions versus continuous limit order books Although the CLOB is the prevailing trading mechanism globally, its continuous nature introduces significant structural challenges. The influential work of [ 8 ] argues that continuous trading inherently creates transient arbitrage opportunities. These fleeting opportunities incentivize market participants to engage in a technological arms race, competing purely for speed to react faster to market updates and execute orders milliseconds ahead of rivals. Budish et al. further argue that this speed competition is economically wasteful, as it does not increase overall market efficiency but rather forces traders to heavily invest resources into latency reduction. Studies by [ 61 ] and [ 25 ] echo these concerns. For instance, [ 61 ] shows that high-frequency traders, acting as latency arbitrageurs, can widen bid-ask spreads and disrupt market price discovery. These structural defects raise a fundamental question of whether alternative trading systems can mitigate these issues. [ 8 ] strongly advocates for frequent batch auctions (conducted, for example, every tenth of a second) to eliminate the inherent flaws of the continuous LOB. Building on this, [ 41 ] determined the optimal duration of such auctions using empirical data from Euronext, concluding that auction mechanisms often outperform limit order books. Furthermore, [ 21 , 18 ] investigate hybrid markets where continuous trading sessions alternate with batch auctions. Collectively, these studies advocate for batch auctions as a robust solution to pool liquidity, neutralize the mechanical advantages of high-frequency trading, and ultimately enhance market efficiency. 1.2 Sniping, transaction fees, and rebates: a motivation from DeFi mechanisms In this paper, we investigate the specific market design features required to make a periodic auction market efficient, well-operated, and beneficial for all participants. While periodic auctions successfully curb speed-based competition and aggregate liquidity for better price discovery [ 41 ] , their baseline design is not without flaws. A toy two-player auction game studied in [ 19 ] demonstrates that, under imperfect information and without specific incentives, a naive auction market can be highly inefficient, failing to generate any trade. However, [ 19 ] proves that introducing transaction fees indexed on each player’s half-spread stimulates transactions, and provides a quantitative framework for the optimal fees an exchange should set to attract liquidity. Therefore, rebate and fee policies are fundamental to generating trade in auction markets with imperfect information. Furthermore, [ 48 ] highlights a critical vulnerability in periodic auctions: unregulated strategic traders have no incentive to reveal their intentions early, preferring to submit orders at the very last moment. This “sniping” behavior distorts the auction, increasing the distance between the clearing price and the fundamental efficient price of the asset. To counteract this effect, exchanges can either randomize the clearing time or optimally calibrate fees to penalize snipers. Consequently, designing optimal financial incentives is essential to monitor and regulate the behavior of strategic traders. “The majority of auctions worldwide require cash bids. Yet in many auctions,

Read the complete text