📈 economics

A Smart-Contract to Resolve Multiple Equilibrium in Intermediated Trade

This paper proposes a privacy-preserving smart contract that resolves multiple equilibria in intermediated repo trades by enabling broker-dealers to securely report their constraints and select a joint profit-maximizing outcome, thereby preventing trade collapse without requiring trust or complex strategic anticipation.

Original authors: Daniel Aronoff, Robert M. Townsend

Published 2026-04-21

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Daniel Aronoff, Robert M. Townsend

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

The Big Picture: A Traffic Jam in the Financial Highway

Imagine the financial world as a massive highway where money and assets (like government bonds) are constantly being swapped. Usually, this happens through a "middleman" system. You have a Borrower (who needs cash) and a Lender (who has extra cash). But they don't talk directly; they go through two Broker-Dealers (let's call them Broker A and Broker B).

Broker A talks to the Borrower.
Broker B talks to the Lender.
Broker A and Broker B talk to each other to make the deal happen.

The problem? This system is prone to a traffic jam caused by confusion.

The Problem: The "Too Scared to Move" Dilemma

The paper argues that this trading system has a flaw called "Multiple Equilibrium." In plain English, this means there are many different ways the trade could work, but the brokers don't know which one everyone else is thinking of.

The Analogy: The Two-Driver Intersection
Imagine two drivers, Driver A and Driver B, approaching a narrow bridge at the same time.

Scenario 1 (The Good Outcome): Driver A thinks, "Driver B will wait for me," so Driver A speeds up. Driver B thinks, "Driver A will wait for me," so Driver B speeds up. They both cross safely and make a profit.
Scenario 2 (The Bad Outcome): Driver A thinks, "Driver B is going to be aggressive and rush me," so Driver A slows down to be safe. Driver B thinks, "Driver A is going to be aggressive," so Driver B also slows down.
The Result: Both drivers slow down so much that nobody crosses the bridge. The trade collapses.

In the real financial world, this happens because brokers have to pay a "tax" (called balance-sheet costs) to hold assets. If the profit from the trade is too low (because they are scared of the other broker), they decide it's not worth the risk. Even though a profitable trade exists, they get stuck in the "slow down" scenario and trade stops completely. This is a coordination failure.

The Solution: The "Smart Contract" Traffic Cop

The authors propose a solution: a Smart Contract. Think of this not as a human judge, but as a robotic traffic cop that runs on a computer.

Here is how this robot solves the traffic jam:

1. The "Secret Menu" (Privacy)

Usually, brokers are afraid to tell each other their prices because they don't want to get ripped off or have their secrets stolen.

The Fix: The Smart Contract uses magic encryption (like a high-tech locked box). Broker A puts their "secret menu" (what they are willing to charge) into the box. Broker B does the same. The robot reads both menus, but nobody else sees them. It's like putting your secret recipe in a safe deposit box that only the robot has the key to.

2. The "Minimum Wage" Rule (The Hurdle)

Brokers have a rule: "I will only do this deal if I make at least $X profit." If the deal pays less, they walk away.

The Fix: The robot asks both brokers, "What is your minimum profit?" It takes the higher of the two numbers. This becomes the "Minimum Wage" for the deal.

3. The "Best Deal" Calculator

The robot looks at all the possible deals that pay at least that Minimum Wage.

It ignores the "slow down" scenarios where everyone is scared.
It ignores the deals that don't pay enough.
It finds the single best deal that maximizes the total profit for both brokers combined.

4. The "No-Brainer" Strategy

The coolest part is how the brokers behave.

Old Way: Brokers had to guess what the other guy was thinking. "If I say $10, will he say $12? Or will he say $5?" This is hard and stressful.
New Way: The robot makes it a dominant strategy to just tell the truth.
- If Broker A lies and says they need a higher profit than they really do, the robot might reject the deal entirely.
- If Broker A lies and says they need less, they just lose money.
- The Result: The smartest, easiest thing for a broker to do is to simply say, "Here is my real price, and here is my real minimum." They don't need to guess or strategize. They just hit "send."

Why This Matters: Saving the Highway

The paper was written by economists at MIT, and it addresses a very real problem that happened during financial crises (like March 2020). When banks are scared, they stop lending, and the whole economy slows down.

The Old Way: Banks freeze up because they are worried about which "equilibrium" (scenario) will happen. They might choose the "no trade" option just to be safe.
The New Way: The Smart Contract guarantees that if a profitable deal exists, it will happen. It removes the fear of the other guy.

The "Magic" Tech Behind It

The paper mentions Zero-Knowledge Proofs and Trusted Execution Environments.

Zero-Knowledge Proof: Imagine you want to prove you are over 21 to buy beer without showing your ID or telling the bartender your name. You just prove "Yes, I am old enough." The Smart Contract proves the math is right without revealing the secret numbers.
Trusted Execution: Imagine a glass-walled room where the robot does the math. Everyone can see the robot is working, but no one can peek inside to see the secret ingredients.

Summary

The paper proposes a digital referee for financial trades.

The Problem: Brokers get stuck in a "fear loop" where they don't trade because they can't agree on what the other guy will do.
The Fix: A computer program that secretly collects their prices, finds the best possible deal that pays everyone enough, and forces that deal to happen.
The Benefit: It stops the market from freezing up, keeps privacy safe, and makes it easy for brokers to just tell the truth.

It's like replacing a chaotic, shouting intersection with a perfectly synchronized, automated traffic light system that ensures traffic keeps moving, even when the drivers are nervous.

1. Problem Statement

The paper addresses a critical coordination failure in the U.S. Treasury repo market, specifically within chains of trade intermediated by two broker-dealers (BDs).

The Market Structure: A repo chain involves a cash lender (Money Market Fund, MM), a cash borrower (Hedge Fund, RM), and two intermediary broker-dealers ( $BD_{mm}$ and $BD_{rm}$ ). The BDs trade with their respective clients and then with each other in an interdealer market.
The Core Friction: The interaction between client-facing schedules and interdealer terms creates a game with multiple equilibria. Due to the negative externality where one BD's client volume dictates the inventory risk for the other, the market can settle into a continuum of self-consistent outcomes ranging from high-volume/high-profit trades to low-volume/low-profit trades, or even a complete collapse to no trade.
The Regulatory Context: Post-2008 regulations (e.g., Supplementary Leverage Ratio) have increased the "shadow cost" of balance-sheet usage. BDs require a minimum return spread (a "hurdle rate") to justify intermediation.
The Failure Mechanism: Even if a high-profit equilibrium exists that satisfies both BDs' hurdle rates, the market may collapse to a "no trade" equilibrium. This occurs because BDs must allocate balance-sheet capacity ex-ante without knowing which equilibrium will be selected. If the expected return (averaging over possible equilibria) falls below the hurdle rate, BDs withdraw capacity, causing a liquidity failure despite feasible trade opportunities existing.

2. Methodology

The authors employ a combination of Game Theory, Mechanism Design, and Cryptography to model and solve the problem.

A. Theoretical Model

Structure: A two-stage game. First, clients and BDs agree on trade schedules (repo rates vs. volumes). Second, BDs play a simultaneous-move game in the interdealer market to determine the final trade volume and split the surplus.
Assumptions:
- BDs are price-setters (monopolists) in client markets.
- Client demand/supply functions are concave and twice differentiable.
- BDs face inventory penalties for unmatched positions and a minimum spread constraint ( $\kappa$ ) driven by balance-sheet costs.
- Surplus is split equally between BDs in the interdealer market.
Equilibrium Analysis: The authors prove the existence of a continuum of pure-strategy Nash equilibria. They demonstrate that under asymmetric information (where BDs do not know the counterparty's client shock), multiple equilibria persist.

B. Mechanism Design (The Smart Contract)

To resolve the multiplicity, the authors propose a Smart Contract acting as a trusted, self-executing planner.

Protocol:
1. Reporting: Each BD reports its client trade schedule (price-quantity pairs) and its minimum hurdle spread ( $\kappa$ ).
2. Constraint Selection: The protocol identifies the binding hurdle constraint ( $\hat{\kappa} = \max(\kappa_{mm}, \kappa_{rm})$ ).
3. Optimization: The contract computes the constrained joint-profit-maximizing trade volume ( $T^*$ ) that satisfies the hurdle constraint.
4. Execution: The contract automatically executes the trade at the rates corresponding to $T^*$ .
Incentive Compatibility: The authors prove that truthful reporting of client schedules and hurdle spreads is a dominant strategy (or weakly dominant) for the BDs. This allows BDs to act myopically without needing to form complex beliefs about the counterparty's strategy.

C. Cryptographic Implementation

To address the privacy concerns of BDs (who cannot reveal sensitive client schedules to competitors), the paper outlines two implementation architectures:

Trusted Execution Environments (TEE): Computations occur in hardware-isolated enclaves.
Zero-Knowledge Proofs (ZKPs): Specifically, succinct non-interactive arguments of knowledge (zk-SNARKs).

Privacy: The BDs commit to their schedules using vector commitments (e.g., Merkle/Pedersen roots). The smart contract proves that the selected trade $T^*$ is the profit maximizer subject to the hurdle constraint without revealing the underlying schedules.
Auditability: The correctness of the computation is publicly verifiable via the ZKP, ensuring integrity without compromising privacy.

3. Key Contributions

Theoretical Characterization of Multiplicity: The paper formally models how intermediated repo chains generate a continuum of equilibria and proves that this multiplicity persists even under asymmetric information. It highlights how equilibrium selection risk interacts with binding balance-sheet constraints to cause market failures.
Constrained Joint Profit Maximization: The authors identify a specific equilibrium (the joint profit maximizer subject to hurdle constraints) that is Pareto superior to other feasible equilibria and proves it can be implemented as a Nash equilibrium in a reporting game.
Trust-Minimized Mechanism: The paper bridges the gap between theoretical mechanism design and practical deployment by proposing a smart contract that:
- Eliminates the need for a trusted central planner.
- Resolves equilibrium selection risk.
- Preserves the privacy of proprietary client data.
- Enables auditability via cryptography.
Myopic Implementation: A significant contribution is showing that the mechanism works with myopic strategies. BDs do not need to simulate counterparty responses; they simply report their true data, and the protocol handles the coordination.

4. Key Results

Existence of Multiple Equilibria: Theorem 2 establishes that for a range of parameters, there exists a continuum of equilibrium repo-rate pairs.
Equilibrium Selection Failure: Section 7 demonstrates that without a coordination mechanism, the risk of selecting a sub-optimal equilibrium can lead to a "no trade" outcome even when profitable trades are theoretically possible.
Truthful Reporting Equilibrium: Theorem 5 proves that in the smart-contract game, truthful reporting of client schedules and minimum spreads constitutes a subgame-perfect equilibrium.
Feasibility Guarantee: The protocol guarantees that if any hurdle-feasible equilibrium exists, the smart contract will select the joint-profit-maximizing one, thereby preventing the collapse to no trade.
Privacy Preservation: The cryptographic protocols (Section 8.4) ensure that the only information revealed is the final trade outcome and a proof of correctness; the underlying client schedules remain hidden.

5. Significance

Market Stability: The proposed mechanism offers a solution to liquidity crises in the Treasury repo market (like the March 2020 turmoil) caused by dealer balance-sheet constraints and coordination failures. By guaranteeing that feasible trades occur, it enhances market resilience.
Regulatory Efficiency: It acts as a "second-best" infrastructure improvement. It does not require deregulation or subsidies but increases the effective capacity of the market by removing equilibrium-selection risk.
Technological Integration: The paper provides a concrete blueprint for integrating blockchain and cryptography into traditional financial markets. It moves beyond theoretical "smart contracts" to a specific application where privacy (via ZKPs) and trust-minimization are essential for market design.
Policy Relevance: It suggests that regulators should focus on reducing equilibrium-selection risk in intermediated markets, potentially by encouraging or mandating such coordination mechanisms, rather than solely focusing on capital requirements.

In summary, Aronoff and Townsend demonstrate that a mathematically rigorous, cryptographically secure smart contract can solve a classic game-theoretic coordination problem in high-stakes financial markets, ensuring liquidity and efficiency while protecting sensitive commercial data.