Paper: arXiv 2503.20340

Authors: Nicole Bäuerle, Tamara Göll

Abstract

In this paper, we consider \(n\) agents who invest in a general financial market that is free of arbitrage and complete. The aim of each investor is to maximize her expected utility while ensuring, with a specified probability, that her terminal wealth exceeds a benchmark defined by her competitors’ performance. This setup introduces an interdependence between agents, leading to a search for Nash equilibria. In the case of two agents and CRRA utility, we are able to derive all Nash equilibria in terms of terminal wealth. For \(n>2\) agents and logarithmic utility we distinguish two cases. In the first case, the probabilities in the constraint are small and we can characterize all Nash equilibria. In the second case, the probabilities are larger and we look for Nash equilibria in a certain set. We also discuss the impact of the competition using some numerical examples. As a by-product, we solve some portfolio optimization problems with probability constraints.

Complexity vs Empirical Score

  • Math Complexity: 9.0/10
  • Empirical Rigor: 2.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper is heavily theoretical, deriving Nash equilibria using advanced stochastic control and game theory in complete markets, with extensive formal proofs and no empirical data or backtesting.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Find Nash Equilibria in<br>Relative Portfolio Optimization<br>with Value at Risk Constraints"] --> B["Methodology: Model Setup<br>Agents: n investors<br>Market: Arbitrage-free, Complete<br>Utility: CRRA or Logarithmic<br>Constraint: Terminal Wealth > Benchmark VaR"]
  B --> C{"Number of Agents n?"}
  C -->|"n = 2"| D["CRRA Utility Analysis<br>Analytical derivation of<br>all Nash equilibria in terminal wealth"]
  C -->|"n > 2"| E["Logarithmic Utility Analysis<br>Case 1: Small probabilities<br>Characterize all equilibria"]
  C -->|"n > 2"| F["Logarithmic Utility Analysis<br>Case 2: Large probabilities<br>Search equilibria in restricted set"]
  D --> G["Key Findings & Outcomes<br>- Full characterization of equilibria for 2 agents<br>- Distinct equilibria for n>2 based on risk levels<br>- Impact of competition on portfolio weights<br>- Solved new portfolio optimization problems with VaR constraints"]
  E --> G
  F --> G
  style A fill:#e1f5fe
  style B fill:#fff3e0
  style G fill:#e8f5e8

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a relative portfolio optimization framework where agents maximize expected utility subject to a stochastic value‑at‑risk (VaR) probability constraint, and studies the resulting Nash equilibria.

Source sentence
In this paper, we consider $n$ agents who invest in a general financial market that is free of arbitrage and complete. Third, this can equivalently be considered as a portfolio optimization problem with a stochastic value-at-risk constraint, i.e. a value-at-risk constraint where the risk level is stochastic.

Assumes 6

  • The financial market is arbitrage‑free and complete. market
    Source sentence
    In this paper, we consider $n$ agents who invest in a general financial market that is free of arbitrage and complete.
  • Agents have CRRA utility for the two‑agent case and logarithmic utility for the multi‑agent case. model
    Source sentence
    In the case of two agents and CRRA utility, we are able to derive all Nash equilibria in terms of terminal wealth. For $n>2$ agents and logarithmic utility we distinguish two cases.
  • Each agent must satisfy a probability constraint that her terminal wealth exceeds a benchmark defined by competitors’ performance. statistical
    Source sentence
    The aim of each investor is to maximize her expected utility while ensuring, with a specified probability, that her terminal wealth exceeds a benchmark defined by her competitors' performance. Third, this can equivalently be considered as a portfolio optimization problem with a stochastic value-at-risk constraint, i.e. a value-at-risk constraint where the risk level is stochastic.
  • Initial capitals are ordered non‑increasingly (x1⁰ ≥ … ≥ xn⁰). other
    Source sentence
    Further, we assume w.l.o.g. that the initial capitals of the agents are ordered by x1 0 ≥. . . ≥xn 0.
Show 2 more
  • The sum of the probability levels αi is ≤ 1 so that constraints can be satisfied on disjoint sets. statistical
    Source sentence
    Assume that the sum of the alphas is less or equal to 1. In this case, the probability constraints of the agents can (and will) be satisfied on disjoint sets. Theorem 4.2.
  • Self‑financing condition E[ZT X] = xi⁰ holds for each agent’s terminal wealth. model
    Source sentence
    6 N. B¨AUERLE AND T. G¨OLL Distinguishing the different cases where the maxima are attained and respecting the self-financing conditions E[X∗ 1ZT ] = x1 0, E[X∗ 2ZT ] = x2 0, we can see that we obtain a solution only in the case of x2 0 ≥β1x1 0 and the unique Nash equilibrium (here and elsewhere uniqueness is always up to sets of probability zero) is given by X∗ 1 = x1 0 ZT and X∗ 2 = x2 0 ZT . We also have to

Data

No dataset named in the text.

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • For two agents with CRRA utility, all Nash equilibria can be derived explicitly in terms of terminal wealth. theoretical
    Source sentence
    In the case of two agents and CRRA utility, we are able to derive all Nash equilibria in terms of terminal wealth.
  • For n > 2 agents with logarithmic utility and small probability levels, all Nash equilibria are characterized. theoretical
    Source sentence
    In the first case, the probabilities in the constraint are small and we can characterize all Nash equilibria.
  • When the probability levels are larger, Nash equilibria exist only in a restricted set of parameter configurations. theoretical
    Source sentence
    In the second case, the probabilities are larger and we look for Nash equilibria in a certain set.
  • If x2⁰ < α2 β1 x1⁰, no Nash equilibrium exists for the logarithmic‑utility case. theoretical
    Source sentence
    If x2 0 < α2β1x1 0, there is no Nash equilibrium.
  • Numerical examples illustrate that wealthy investors are only marginally affected by competition, while poorer agents may deviate dramatically, leading to volatile and risky investments. performance
    Source sentence
    We also discuss the impact of the competition using some numerical examples. It turns out that wealthy investors are not or only marginally affected by the competition, whereas poorer agents deviate dramatically from their benchmark optimal strategy, leading to very volatile and risky investments at least in some scenarios.

Checked by others

Qualified by

  • Predictable Relative Forward Performance Processes: Multi-Agent and Mean Field Games for Portfolio Management (2023)
    On the question: How does competition influence agents' risky investment behavior?
    Compare the findings

    This paper: Numerical examples illustrate that wealthy investors are only marginally affected by competition, while poorer agents may deviate dramatically, leading to volatile and risky investments.

    That paper: The competition‑induced component of the strategy is always positive, implying that relative performance concerns encourage higher risky investment among CARA agents.

    Both find competition raises risky investment; A says the effect is always positive for CARA agents, B shows it is marginal for wealthy investors but dramatic for poorer agents.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.