A mathematical framework for modelling CLMM dynamics in continuous time
Shen-Ning Tung, Tai-Ho Wang · 2024
Read A mathematical framework for modelling CLMM dynamics in continuous time by Shen-Ning Tung, Tai-Ho Wang (2024) 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-2412-18580.txt.
Introduction
Decentralized finance uses automated market makers that let traders swap tokens within set price ranges, a design called concentrated liquidity. Shen-Ning Tung and Tai-Ho Wang build a strict mathematical framework for these systems in continuous time, in a 2024 paper. They model how the full liquidity profile shifts as trading goes on, and study traders who exploit price gaps.
They solve for the best gap-trading strategy under three outlooks: greedy, fixed-horizon, and endless play. One striking result is that trading fees rule out whole classes of price paths, since wiggly prices would mint endless fees. The authors say this shapes how such markets should be designed.
Catalog record
| CREATOR | Shen-Ning Tung, Tai-Ho Wang |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-18580
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A mathematical framework for modelling CLMM dynamics in continuous time Shen-Ning Tung, Tai-Ho Wang Abstract . This paper develops a rigorous mathematical framework for analyzing Concentrated Liquidity Market Makers (CLMMs) in Decentralized Finance (DeFi) within a continuous-time setting. We model the evolution of liquidity profiles as measure-valued processes and characterize their dynamics under continuous trading. Our analysis encompasses two critical aspects of CLMMs: the mechanics of concentrated liquidity provision and the strategic behavior of arbitrageurs. We examine three distinct arbitrage models—myopic, finite-horizon, and infinite-horizon with discounted and ergodic controls—and derive closed-form solutions for optimal arbitrage strategies under each scenario. Importantly, we demonstrate that the presence of trading fees fundamentally constrains the admissible price processes, as the inclusion of fees precludes the existence of diffusion terms in the price process to avoid infinite fee generation. This finding has significant implications for CLMM design and market efficiency. 1. Introduction Decentralized Finance ( DeFi ) has revolutionized the financial landscape by enabling trustless, peer-to-peer transactions without intermediaries [ HRS+21 , GM23 ] . At the heart of this revolution are Automated Market Makers ( AMMs ) [ CJ21 , Loe22 ] , which facilitate efficient and permissionless token exchanges. Unlike traditional limit order books ( LOBs ) that match discrete bids and asks [ GPW+13 ] , AMMs employ mathematical formulas to determine prices based on asset quantities within liquidity pools, ensuring continuous liquidity availability and instant trade execution. The introduction of Concentrated Liquidity Market Makers ( CLMMs ) by Uniswap V3 [ AZS+21 ] marked a paradigm shift in AMM design. Unlike traditional Constant Function Market Makers ( CFMMs ) [ AC20 , ACE22 , ACD+23 ] , where liquidity is distributed uniformly across the entire price range according to a fixed bonding curve, CLMMs empower liquidity providers ( LPs ) to concentrate their capital within specific price intervals. This fundamental change shifts control of the market’s bonding curve from a rigid, protocol-defined function to an emergent structure determined collectively by LP decisions. By enabling market participants to actively shape the liquidity distribution, CLMMs address a key inefficiency of CFMMs, where substantial capital may sit idle in rarely-traded price ranges. The resulting dynamic liquidity profile more closely resembles traditional limit order books [ MMR23b ] , combining the capital efficiency of order book markets with the automation and permissionless nature of AMMs. Recent literature has made significant strides in understanding CLMM dynamics. Heimbach et al. [ HSW23 ] provide a comprehensive analysis of the risks and returns associated with concentrated liquidity provision. Echenim et al. [ EGM23 ] develop precise models of CLMM mechanics using “liquidity curves” to analyze strategies and calculate LP fee earnings. Urusov et al. [ UBY24 ] contribute specialized backtesting frameworks for evaluating CLMM performance. Milionis et al. [ MMRZ22a , MMRZ22b ] explored the theoretical foundations of AMM dynamics under myopic arbitrage, deriving a Black-Scholes-type formula, termed “loss-versus-rebalancing” ( LVR ), to quantify LPs’ adverse selection costs. Their subsequent work [ MMR23a ] extended this analysis to incorporate trading fees and discrete block generation times. Fukasawa et al. [ FMW23b , FMW23a ] discussed model-free hedging strategies in the context of Geometric Mean Market Makers ( G3Ms ). Further research has broadened the analytical framework, with Bergault et al. [ BBB+24 ] and Cartea et al. [ CDSB+23 ] exploring innovative AMM designs based on stochastic control theory. Despite these advances in understanding CLMMs, a unified mathematical framework that fully captures the dynamic interplay between concentrated liquidity provision, price movements, and trading fees, comparable to those developed for LOBs [ CST10 , CdL13 , CM21 , CDX23 ] , has remained elusive. This paper addresses this gap through three main contributions: (1) We introduce a framework for modeling liquidity profiles as measure-valued processes (see Equations ( 16 ) and ( 19 )). This formulation precisely characterizes how concentrated liquidity affects market behavior and trading outcomes. (2) Building upon the joint work with C.-Y. Lee [ LTW24 ] , we employ stochastic analysis and control theory to analyze liquidity profile dynamics under arbitrage. This analysis reveals that the presence of trading fees imposes fundamental constraints on admissible price processes, specifically precluding diffusion terms to avoid infinite fee generation. (3) We derive closed-form solutions for optimal arbitrage strategies under three distinct scenarios: myopic arbitrage, finite-horizon optimization, and infinite-horizon optimization with discounted or ergodic control. These solutions provide valuable insights into how rational actors interact with CLMMs and influence price discovery and market efficiency. Our findings have significant implications for CLMM design and management, offering guidance on optimizing fee structures and liquidity incentives to enhance market efficiency. The framework we develop bridges classical financial theory with emerging DeFi mechanisms, providing valuable insights for both researchers and practitioners in the field. Outline The remainder of this paper is structured as follows: Section 2 examines the fundamentals of CFMMs, establishing the mathematical groundwork. Section 3 introduces liquidity profiles in CLMMs and develops the measure-theoretic framework for analyzing their properties. Section 4 presents our main results on CLMM dynamics, including the characterization of optimal arbitrage strategies and their implications for price processes. Finally, Section 5 concludes with a discussion of practical implications and directions for future research. 2. Automated Market Maker Fundamentals This section provides a concise review of Constant Function Market Makers. Readers already familiar with this topic can proceed to the next section. 2.1. Constant Function Market Makers Constant Function Market Makers ( CFMMs ) [ AC20 , ACE22 , ACD+23 ] is a fundamental class of AMMs that operate based on a bonding function. These systems facilitate decentralized trading by enabling interactions between two primary categories of participants: • Liquidity Providers (LPs): Users who deposit assets into the CFMM and receive liquidity tokens representing their pool share. • Liquidity Takers (Traders): Users who exchange one asset for another using the liquidity provided in the pool. 2.1.1. Trading Mechanism The core mechanism of a CFMM relies on its bonding function, which represents a level set: f(x,y)=\ell, where • x,y\in\mathbb{R}_{+} represent the quantities of the risk asset and numéraire, respectively; • \ell\in\mathbb{R}_{+} denotes the liquidity amount of the pool; • f:\mathbb{R}_{+}\times\mathbb{R}_{+}\to\mathbb{R}_{+} is a twice continuously differentiable function. A trade, represented by quantity changes (\Delta x,\Delta y) , is considered valid if and only if it preserves the bonding function: f(x+\Delta x,y+\Delta y)=f(x,y). This constraint ensures that the liquidity amount remains constant throughout each trade, maintaining the fundamental relationship between asset quantities. For a CFMM to support viable trading, the bonding function must satisfy several key properties: • Monotonicity: The implicit function y(x) derived from f(x,y)=\ell must be strictly decreasing. • Convexity: The trading curve must be convex (see Equation ( 2 )) to ensure price impact increases with trade size. • Scaling: For any \lambda>0 , f(\lambda x,\lambda y)=\lambda\ell . This ensures reserves (x,y) scale linearly with the liquidity amount \ell . 2.1.2. Price Formation The instantaneous price (or spot price) of the risk asset in terms of the numéraire is derived using the implicit function theorem: (1) P=-\frac{dy}{dx}=\frac{f_{x}}{f_{y}}, where f_{x} and f_{y} are partial derivatives of f with respect to x and y . This ratio of partial derivatives represents the slope of the tangent line to the trading curve at