📈 economics

The Quantum Structure of Markets: Linking Hamiltonian-Jacobi-Bellman Dynamics to Schrodinger Equation through Feynman Action

This paper proposes a Euclidean path-integral control framework that bypasses the explicit construction of value functions in Hamilton-Jacobi-Bellman formulations to derive noncooperative feedback Nash equilibria for optimal firm behavior, offering a tractable alternative for complex nonlinear stochastic markets while drawing parallels to quantum mechanics and mean-field game theory.

Original authors: Paramahansa Pramanik

Published 2026-03-27

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Paramahansa Pramanik

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 Idea: Trading the Stock Market Like a Quantum Particle

Imagine you are trying to navigate a boat through a foggy, stormy ocean to reach a specific island. You want to get there using the least amount of fuel (cost) while maximizing your catch of fish (profit).

The Old Way (The "Map" Approach):
Traditional economics uses a method called Hamilton-Jacobi-Bellman (HJB). Think of this as trying to draw a perfect, detailed map of the entire ocean before you start sailing. You calculate every possible wave, every possible wind shift, and every possible route on a giant grid.

The Problem: If the ocean is huge (many companies, many variables), the map becomes so massive that no computer can draw it. It's like trying to map every single grain of sand on a beach to find the best spot to build a sandcastle. This is called the "curse of dimensionality."

The New Way (The "Quantum" Approach):
This paper suggests a different way, borrowed from quantum physics (the study of tiny particles like electrons). Instead of drawing a map, imagine you are a quantum particle.

In quantum physics, a particle doesn't take just one path to get from point A to point B. It takes every possible path at the same time. Some paths are through the storm, some are through calm water, some are loops.
The "Quantum" method in this paper says: "Let's simulate thousands of these possible boat journeys at once. We don't need a perfect map. We just need to see which paths look the most promising and give them more 'weight'."

The Core Metaphor: The "Path Integral"

The paper uses something called a Path Integral. Here is how to visualize it:

The Foggy Ocean: The market is full of uncertainty (random noise, sudden price changes, unexpected news).
The Ghost Boats: Instead of one boat, imagine thousands of "ghost boats" leaving the harbor at the same time. Each one takes a slightly different, random route based on the wind and waves.
The Scorecard: As each ghost boat travels, we calculate its score (profit minus cost).
- A boat that gets stuck in a storm or burns too much fuel gets a low score.
- A boat that catches a great school of fish and avoids the storm gets a high score.
The Magic Filter: We don't just pick the one boat that got the highest score. Instead, we look at all the boats. We give the high-scoring boats a "super boost" (mathematically, an exponential weight) and the low-scoring boats a "dimmer switch."
The Result: When you average out all these weighted boats, the "blur" of the fog clears up, and a clear, optimal path emerges. This path tells the real company what to do right now.

Why is this better than the old way?

The paper argues that the old "Map" method (HJB) breaks down when things get complicated or non-linear (when the rules of the game change suddenly).

Analogy: Imagine trying to solve a maze.
- Old Way: You try to calculate the exact distance from every single point in the maze to the exit. If the maze is huge, your brain explodes.
- New Way: You release a million ants into the maze. They all wander randomly. The ones that hit a dead end stop. The ones that find the exit keep going. You just follow the trail of the ants that found the exit. You didn't need to know the whole maze; you just needed to see where the ants went.

The Three Scenarios the Paper Tests

The author applies this "Quantum Ant" method to three different types of market situations:

The "Price Taker" (Walrasian Equilibrium):
- Analogy: A single small fish in a giant ocean. The fish can't change the water temperature; it just has to swim with the current.
- Result: The method finds the best swimming strategy for the fish without needing to know the exact temperature of the whole ocean.
The "Teamwork" (Pareto Optimality):
- Analogy: A school of fish swimming together. If one fish turns left, the others might follow to stay safe. They want to maximize the total fish caught by the whole school, not just one.
- Result: The method calculates how the whole school should move to ensure everyone gets fed, even if they have to coordinate their turns.
The "Competition" (Nash Equilibrium):
- Analogy: A game of poker. You want to win, but you know the other players are also trying to win. You have to guess what they will do and react.
- Result: The method finds a "stable" strategy where no player wants to change their move because they know the others are playing optimally too.

The Secret Ingredient: The "Magic Wand" (The $g$ function)

The paper mentions a tricky part: choosing a "magic wand" function (called $g$ ).

Analogy: Imagine you are trying to listen to a radio station, but there is static (noise). You need to tune the radio to the right frequency to hear the music clearly.
In this math, the "magic wand" is a specific mathematical adjustment that helps filter out the noise in the market data. If you pick the right "frequency" (the right $g$ function), the solution pops out clearly. If you pick the wrong one, you just hear static. The paper shows how to pick the right one for different market conditions.

The Bottom Line

This paper is a proposal to stop trying to draw perfect, impossible maps of the economy. Instead, it suggests using a simulation-based, "quantum" approach:

Simulate thousands of possible futures.
Weight the good ones heavily and the bad ones lightly.
Let the "best" path emerge naturally from the chaos.

It's like finding your way through a dark forest not by drawing a map of the whole forest, but by throwing a handful of glowing fireflies into the air and following the ones that fly toward the exit. It's a powerful new tool for solving complex economic problems that were previously too hard for computers to handle.

1. Problem Statement

The paper addresses the challenge of determining optimal firm behavior in complex economic environments characterized by:

Stochastic Dynamics: Market shares evolve according to non-linear stochastic differential equations (SDEs) driven by Brownian motion.
Forward-Looking Constraints: Decisions involve implicit commitments to future states, creating "forward-looking" dynamics that often violate the time-consistency required for standard recursive methods.
High Dimensionality: The presence of many interacting firms (large $k$ ) leads to the "curse of dimensionality," making traditional grid-based solutions intractable.
Equilibrium Complexity: The paper seeks to characterize three distinct market structures:
1. Walrasian Equilibrium: Price-taking firms maximizing individual profit.
2. Pareto Optimality: Cooperative firms maximizing a weighted sum of total profits.
3. Non-Cooperative Nash Equilibrium: Firms maximizing individual profit given the strategies of others (feedback Nash).

The Core Limitation: Standard approaches rely on the Hamilton-Jacobi-Bellman (HJB) equation and the Pontryagin Maximum Principle. These methods require the explicit construction of a value function $V(s, X)$ . In non-linear, high-dimensional, or forward-looking settings, the HJB equation often lacks smooth solutions, is computationally prohibitive, or fails to yield a unique global optimum.

2. Methodology

The author proposes a Euclidean Path-Integral Control framework, drawing an analogy between stochastic economic systems and quantum mechanical systems. The methodology bypasses the value function entirely.

A. Theoretical Framework

Stochastic Lagrangian Formulation: The firm's problem is recast as maximizing a dynamic profit functional subject to SDE constraints. The constraint is handled via a Lagrangian multiplier, creating a stochastic action functional.
Feynman Path Integral: Instead of solving a partial differential equation (PDE) backward in time, the approach aggregates contributions from all possible trajectories (paths) of the market share $X(s)$ from an initial state to a terminal state.
Wick Rotation and Schrödinger Equation:
- The author applies a Wick rotation (transforming time to imaginary time) to convert the backward HJB equation into a forward Schrödinger-type equation (or equivalently, a forward Fokker-Planck equation).
- This transformation relies on a specific integrating factor (a function $g(s, X)$ ) chosen to linearize the problem.
- The resulting equation governs the evolution of a "transition wave function" $\Psi(s, X)$ , which represents the probability amplitude of the system's state.

B. Solution Mechanism

No Value Function: The method does not require solving for $V(s, X)$ . Instead, optimal strategies are derived from the first-order conditions of the transformed Schrödinger-type equation.
Monte Carlo Sampling: The solution is computed via path-integral Monte Carlo simulation. The algorithm:
1. Simulates an ensemble of stochastic trajectories (paths) for the market share.
2. Assigns exponential importance weights ( $w_i \propto e^{-\epsilon A}$ ) to each path based on the action functional (profit).
3. Computes the optimal control as a weighted average of the controls along these paths.
Handling Non-Linearity: By using the path integral, the method accommodates non-linear drift and diffusion coefficients (e.g., Merton-Garman-Hamiltonian dynamics) where standard HJB methods fail or become numerically unstable.

3. Key Contributions

Derivation of Non-Cooperative Feedback Nash Equilibrium: The paper successfully derives a feedback Nash equilibrium within a path-integral setting for a large population of firms, contrasting it with Mean-Field Game formulations.
Alternative to HJB/Pontryagin: It provides a rigorous alternative to the Pontryagin Maximum Principle for problems where the value function is difficult to construct or where the HJB equation is non-linear and non-unique.
Handling Merton-Garman-Hamiltonian Dynamics: The framework is applied to stochastic volatility models (European call options) where the Hamiltonian is non-linear, a case where traditional analytical solutions are often impossible.
Existence Proofs: The paper establishes the existence of Nash equilibria under specific regularity conditions (compact convex strategy sets, quasi-concavity) using fixed-point theorems (Fan-Browder).
Computational Scalability: It demonstrates that the path-integral approach avoids the exponential memory growth ( $O(n^k)$ ) of grid-based HJB solvers, scaling linearly with the number of Monte Carlo paths ( $O(n)$ ), making it suitable for high-dimensional systems.

4. Results and Examples

The paper validates the methodology through several examples and numerical simulations:

Example 1 (Walrasian Firm): A linear-quadratic setup where the path integral yields a closed-form optimal strategy. The result is compared to the standard HJB solution, showing equivalence in linear cases but highlighting the path integral's flexibility.
Example 2 (Non-Linear Comparison): A comparison between the "Quantum Approach" (Path Integral) and the "Pontryagin Approach" (HJB).
- Result: The Path Integral approach yields a cubic polynomial for the optimal strategy, whereas the Pontryagin approach (assuming a linear value function ansatz) yields a linear feedback rule ($u = 2bX/3$). The paper argues the cubic solution captures non-linear dynamics better.
Example 3 (Cooperative Pareto): A multi-firm cooperative scenario. The path integral method derives a distinct optimal strategy compared to the Pontryagin benchmark, driven by the specific choice of the integrating factor $g(s, X)$ .
Example 4 & 5 (Resource Extraction): Applied to resource extraction games. The path integral method successfully handles non-linear drift terms where the Pontryagin principle struggles with the value function's characteristic equation.
Numerical Simulations:
- Monte Carlo simulations (2000 paths) demonstrate the stability of the algorithm.
- The distribution of terminal market shares and the "effective sample size" (ESS) of the importance weights confirm that the method remains well-conditioned and does not suffer from severe weight degeneracy under the tested parameters.

5. Significance and Implications

Theoretical Unification: The paper bridges Quantum Mechanics (Feynman path integrals, Schrödinger equations) and Economic Theory (Dynamic programming, Game Theory). It suggests that economic agents in stochastic environments can be modeled similarly to quantum particles, where the "optimal path" is a probabilistic aggregation of all feasible histories.
Computational Advantage: By replacing the "curse of dimensionality" inherent in grid-based PDE solvers with Monte Carlo sampling, the method offers a practical tool for solving high-dimensional stochastic differential games in macroeconomics and finance.
Robustness to Non-Linearity: The approach is particularly significant for markets with strong non-linearities (e.g., stochastic volatility, saturation effects) where standard dynamic programming fails to provide unique or smooth solutions.
Policy Implications: The framework allows policymakers to model forward-looking commitments and aggregate uncertainty without needing to explicitly solve for a global value function, offering a new lens for analyzing equilibrium selection in large economies.

In summary, Pramanik presents a novel "Quantum" approach to economic control that leverages path integrals to solve complex, non-linear, and high-dimensional stochastic games, offering a computationally efficient and theoretically robust alternative to the traditional HJB/Pontryagin paradigm.