Paper: arXiv 2610.08631
Authors: Balazs Hoffmann, Miklos Rasonyi
Abstract
We investigate a continuous-time financial market where the asset price exhibits weak (sublinear) mean reversion and has a nonzero drift. Complementing earlier work on strong (superlinear) mean reversion, we show that, for an investor maximizing expected exponential utility, the certainty equivalent grows as \(O(T^{2β+1})\) where \(0<β<1\) is the strength of mean reversion. An explicit asymptotically optimal strategy is also given.
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: This paper is highly theoretical, focusing on advanced stochastic calculus and utility maximization. While it extends previous work with a novel approach to weak mean reversion, it lacks any empirical validation or data analysis. The mathematical derivations are rigorous and detailed.
Research Flowchart
flowchart TD
A["Research Goal: Optimal Investment with Weak Mean Reversion"] --> B("Methodology: Continuous-Time Financial Market Model");
B --> C{"Inputs: Asset Price Dynamics, Exponential Utility"};
C --> D["Computational Process: Solve Optimization Problem"];
D --> E("Key Finding 1: Certainty Equivalent Growth O(T²β+1)");
E --> F["Key Finding 2: Explicit Asymptotically Optimal Strategy"];
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 an explicit asymptotically optimal trading strategy for an exponential utility investor in a continuous-time market with weak (sublinear) mean-reverting asset prices, and shows that the certainty equivalent grows on the order of T^{2β+1}.
Source sentence
Complementing earlier work on strong (superlinear) mean reversion, we show that, for an investor maximizing expected exponential utility, the certainty equivalent grows as $O(T^{2β+1})$ where $0<β<1$ is the strength of mean reversion. An explicit asymptotically optimal strategy is also given.
Assumes 6
- The asset price follows a continuous-time dynamics with weak (sublinear) mean reversion and a non‑zero drift. market
Source sentence
We investigate a continuous-time financial market where the asset price exhibits weak (sublinear) mean reversion and has a nonzero drift.
- The mean‑reversion strength parameter satisfies 0 < β < 1. model
Source sentence
Assume 0 < β < 1, α > 0, µ̸ = 0.
- The volatility parameter α is positive (α > 0). model
Source sentence
Assume 0 < β < 1, α > 0, µ̸ = 0.
- The drift µ is non‑zero (µ ≠ 0). model
Source sentence
Assume 0 < β < 1, α > 0, µ̸ = 0.
Show 2 more
- The investor maximizes expected exponential utility. other
Source sentence
Complementing earlier work on strong (superlinear) mean reversion, we show that, for an investor maximizing expected exponential utility, the certainty equivalent grows as $O(T^{2β+1})$ where $0<β<1$ is the strength of mean reversion.
- Admissible trading strategies belong to the class 𝒜_T defined in the paper. model
Source sentence
(3.3) Then Hβ,ε,T ∈AT for every T ≥1.
Data
No dataset named in the text.
Costs
Not applicable
Finds 5
- The certainty equivalent of the optimal exponential‑utility investor grows as O(T^{2β+1}) for 0<β<1. theoretical
Source sentence
Complementing earlier work on strong (superlinear) mean reversion, we show that, for an investor maximizing expected exponential utility, the certainty equivalent grows as $O(T^{2β+1})$ where $0<β<1$ is the strength of mean reversion.
- An upper bound c_T ≤ C_{α,β,µ} T^{2β+1} holds for all admissible families of strategies. theoretical
Source sentence
(4.6) Combining (4.5) and (4.6), and using T = O(T 2β+1) and T β+1 = O(T 2β+1) for β > 0, yields cT ≤Ent(QT | P) ≤Cα,β,µT 2β+1, T ≥1.
- The constructed family H_{β,ε} achieves C(H_{β,ε}) = 2β + 1, matching the upper bound. theoretical
Source sentence
The upper bound from section 4 gives C(H) ≤2β + 1 for every admissible family H. Since the family Hβ,ε is admissible by Section 6, we conclude that C(Hβ,ε) = 2β + 1.
- The strategy H_{β,ε,T}(t) = (T+1−t)^{β−1} φ_ε(S_t) is admissible for every T ≥ 1. theoretical
Source sentence
(3.3) Then Hβ,ε,T ∈AT for every T ≥1.
- The lower bound proof shows that E[exp(−V_{β,ε,T})] ≤ C exp(−b T^{2β+1}) for large T, establishing the optimal growth rate. theoretical
Source sentence
9 10 Completion of the proof Equations (8.6) and (9.4) show that, for every 0 < β < 1, every fixed ε > 0, and all sufficiently large T, there are constants C, b > 0 such that E exp{−V β,ε T } ≤Ce−bT 2β+1.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 19 related papers). Coverage is partial, so this is neither replication nor refutation.
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.