🔢 mathematics

Global regularity and free boundary geometry in the planar Choné-Rochet model

This paper establishes global $C^1$ and $C^{1,1}$ regularity for minimizers of the planar Choné-Rochet variational problem on convex domains, demonstrates the optimality of $C^1$ regularity by constructing counterexamples with flat boundaries, and proves that the tamed free boundary is a locally $C^1$ embedded curve.

Original authors: Shibing Chen, Alessio Figalli, Yi Ru-Ya Zhang

Published 2026-03-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Shibing Chen, Alessio Figalli, Yi Ru-Ya Zhang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a monopolist (a single seller) trying to sell products to a crowd of customers. Each customer has a different "taste" or preference, represented by a point on a map. You want to set prices to make the most profit, but you don't know exactly what each customer wants; you only know the general distribution of tastes.

This is the real-world problem behind the math in this paper. The authors are studying a specific mathematical model (the Choné–Rochet model) that describes how a smart seller should price their goods.

Here is the breakdown of their discovery, translated from "Mathematician" to "Human."

The Big Picture: The "Perfectly Smooth" Map

The authors are looking at the optimal price map. Think of this map as a landscape where the height of the terrain represents the "utility" or happiness a customer gets from a product.

The Goal: Find the shape of this landscape that maximizes profit.
The Rules: The landscape must be "convex" (like a bowl, never a dip) and cannot go below zero (you can't charge negative money).
The Mystery: Mathematicians knew this map was smooth inside the domain (the middle of the map), but they weren't sure if it stayed smooth all the way to the edges (the boundary of the market).

The Three Main Discoveries

1. The Edge is Smooth (The "No Rough Spots" Discovery)

The Problem: In many math problems, things get messy at the edges. You might expect the price map to have a sharp corner or a sudden jump right at the border of the market.
The Discovery: The authors proved that for a 2D market (a flat map), the price map is perfectly smooth everywhere, right up to the edge.

The Analogy: Imagine a sheet of rubber stretched over a frame. Sometimes, if the frame is weird, the rubber might get crinkled or sharp at the corners. The authors proved that for this specific problem, no matter how you shape the frame (as long as it's a convex shape), the rubber sheet remains perfectly smooth and bendable from the center all the way to the rim. There are no sharp kinks.

2. The "Flat Edge" Trap (When Smoothness Breaks)

The Problem: If the market boundary is strictly curved (like a circle), everything is great. But what if the boundary has a flat side (like a square or a rectangle)?
The Discovery: The authors found that if the market has a flat edge, the smoothness can break down. The gradient (the slope of the price map) can become infinitely jagged.

The Analogy: Imagine driving a car on a perfectly smooth road that suddenly hits a flat, straight wall. If you try to turn the steering wheel to follow the wall, the wheel might spin wildly or jerk.
The Result: They constructed a specific example where the "slope" of the price changes so violently near a flat edge that it doesn't follow any standard rule of smoothness. This proves that their first discovery (that it's smooth everywhere) is the best possible result; you can't demand it be even smoother without adding more rules.

3. The "Tamed" Border (The Invisible Line)

The Problem: Inside the market, there is a hidden line. On one side, the price map is strictly curved (like a deep bowl). On the other side, it's flat (like a table). This line is called the Free Boundary.

The Mystery: What does this invisible line look like? Is it a jagged scribble? A fractal? A smooth curve?
The Discovery: The authors proved that this line is locally a smooth curve.
The Analogy: Imagine a river (the flat region) meeting a mountain (the curved region). The shoreline between them might be expected to be rocky and chaotic. The authors proved that this shoreline is actually a gentle, winding path that you could walk along without tripping. It's not a jagged mess; it's a clean, smooth curve.

Why Does This Matter?

For Economists: It gives them confidence that the pricing models they use are mathematically sound. They know the "optimal price" won't suddenly explode or become undefined at the edges of their market.
For Mathematicians: It solves a 20-year-old puzzle. Previous work showed the map was smooth inside, but this paper closes the gap and proves it's smooth everywhere (under the right conditions).
The "Rigidity" Insight: The paper also shows a fascinating "rigid" behavior. If the price map is flat on a boundary segment, it forces the whole shape of the solution to be very specific (like a perfect parabola). It's as if the math says, "If you make it flat here, you must be this specific shape everywhere else."

Summary in One Sentence

The authors proved that the mathematical map for optimal pricing is perfectly smooth from the center to the edge (unless the edge is flat, in which case it gets messy), and the invisible line separating different pricing strategies is always a clean, smooth curve.

1. Problem Statement and Context

The paper investigates the regularity and geometric structure of the unique minimizer $u$ of the Choné–Rochet variational problem in two dimensions ( $\mathbb{R}^2$ ). This problem arises in mechanism design and nonlinear pricing, specifically modeling a monopolist maximizing profit under asymmetric information.

The variational problem is defined as:
$\min_{u \in K(\Omega)} L[u; \Omega], \quad L[u; \Omega] := \int_{\Omega} \left( \frac{1}{2}|Du - x|^2 + u \right) dx$
where $\Omega \subset \mathbb{R}^2$ is a bounded convex domain, and $K(\Omega)$ is the set of non-negative convex functions on $\Omega$ .

Key Mathematical Features:

The Euler-Lagrange equation leads to a free boundary problem where the domain decomposes into three regions:
- $\Omega_0$ : The contact set where $u=0$ .
- $\Omega_1$ : The region where $u>0$ but $u$ is not strictly convex (the "flat" region where the Hessian is degenerate).
- $\Omega_2$ : The region where $u$ is strictly convex ( $\Delta u = 3$ ).
The interface between $\Omega_1$ and $\Omega_2$ is the free boundary.
Previous work established interior $C^{1,1}$ regularity and $C^1$ regularity up to the boundary for specific domains, but global $C^1$ regularity for arbitrary convex domains and the precise geometry of the free boundary remained open questions.

2. Methodology

The authors employ a combination of variational methods, geometric analysis, and blow-up techniques:

Variational Perturbation: The authors utilize affine perturbations of the minimizer (similar to Caffarelli-Lions techniques) to derive energy estimates. They construct competitors by "tilting" the supporting affine functions to analyze the behavior near the boundary and within the contact set.
Geometric Decomposition: They analyze the structure of "leaves" (contact sets $[x]$ where $u$ agrees with its supporting plane). The proof distinguishes between "one-ended leaves" (one endpoint on $\partial\Omega$ ) and "two-ended leaves" (both endpoints on $\partial\Omega$ ).
Blow-up Analysis: To study the free boundary, the authors perform blow-up limits at points on the interface. They utilize Caffarelli's alternative for obstacle problems to classify points based on their density (0 or 1/2).
Rigidity Arguments: A crucial component is the proof of a rigidity theorem (Theorem 1.5) which rules out "cusp-type" singularities in the blow-up limits, forcing the solution to be a specific quadratic form in certain sectors.
Counterexample Construction: To demonstrate the optimality of their regularity results, they construct a specific family of smooth convex domains with flat boundary segments where the gradient fails to satisfy any prescribed modulus of continuity.

3. Key Contributions and Results

A. Global Regularity Estimates

The paper resolves the regularity of the minimizer $u$ on the entire domain $\Omega$ :

Global $C^1$ Regularity (Theorem 1.1):
- Result: For any bounded convex domain $\Omega \subset \mathbb{R}^2$ , the minimizer $u$ is globally $C^1(\Omega)$ .
- Significance: This fills a gap in previous literature (specifically correcting an incomplete argument in Carlier–Lachand-Robert [6]) by proving that the gradient extends continuously to the boundary, even for non-strictly convex domains.
Global $C^{1,1}$ Regularity (Theorem 1.2):
- Result: If $\Omega$ is strictly convex, or more generally if the zero set $\{u=0\}$ has positive measure, then $u \in C^{1,1}(\Omega)$ .
- Significance: This establishes a uniform bound on the Hessian $D^2u$ globally, provided the geometry prevents the formation of "flat" boundary segments that touch the zero set in a degenerate way.
Optimality of $C^1$ Regularity (Theorem 1.3):
- Result: The authors construct a smooth bounded convex domain with a flat boundary segment where the gradient $Du$ fails to satisfy any prescribed modulus of continuity $\omega$ .
- Significance: This proves that without additional geometric assumptions (like strict convexity), global $C^1$ regularity is the best possible result; global $C^{1,\alpha}$ or $C^{1,1}$ regularity cannot be guaranteed.

B. Free Boundary Geometry

The paper significantly advances the understanding of the "tamed free boundary" $T$ , defined as the interface between the strictly convex region $\Omega_2$ and the non-strictly convex region $\Omega_1$ where the leaves have exactly one endpoint on the boundary.

$C^1$ Regularity of the Free Boundary (Theorem 1.4):
- Result: If $\Omega$ is $C^{1,1}$ and $u \in C^{1,1}(\Omega)$ , the tamed free boundary $T$ is locally a $C^1$ embedded curve.
- Significance: Previous results (e.g., McCann, Rankin, Zhang [17]) only established that the free boundary has Hausdorff dimension strictly less than 2. This paper proves the boundary is actually a smooth curve with no singular points (cusps or corners) in the tamed region.
Rigidity Theorem (Theorem 1.5):
- Result: A local rigidity result showing that if a convex function $u$ satisfies $\Delta u = 1$ in a sector $D$ and $u=0$ on a segment of the boundary, then $u$ must be a specific quadratic function ( $x_2^2/2$ ) in that sector.
- Significance: This theorem is used to rule out 0-density points (cusps) in the blow-up analysis of the free boundary, ensuring the free boundary is regular everywhere.

4. Significance and Impact

Resolution of Open Problems: The paper settles long-standing questions regarding the global regularity of the Choné–Rochet model in 2D, moving from interior estimates to global bounds.
Sharpness of Results: By constructing the counterexample, the authors precisely delineate the limits of regularity, showing that the geometry of the domain (specifically the presence of flat boundary segments) dictates the regularity of the solution.
Free Boundary Smoothness: The proof that the free boundary is a $C^1$ curve (rather than just a set of finite dimension) is a major step forward in the regularity theory of free boundary problems with convexity constraints. It implies that the transition between pricing regimes (in the economic application) is smooth.
Methodological Innovation: The combination of variational perturbation arguments with a new rigidity theorem for blow-up limits provides a robust framework that could be applicable to other free boundary problems involving convexity constraints.

In summary, this paper provides a comprehensive and sharp analysis of the planar Choné–Rochet model, establishing global $C^1$ regularity for all convex domains, global $C^{1,1}$ for strictly convex domains, and proving that the associated free boundary is a smooth $C^1$ curve.