Paper: arXiv 2308.15135

Authors: Owen Futter, Blanka Horvath, Magnus Wiese

Abstract

In this article we introduce a portfolio optimisation framework, in which the use of rough path signatures (Lyons, 1998) provides a novel method of incorporating path-dependencies in the joint signal-asset dynamics, naturally extending traditional factor models, while keeping the resulting formulas lightweight and easily interpretable. We achieve this by representing a trading strategy as a linear functional applied to the signature of a path (which we refer to as “Signature Trading” or “Sig-Trading”). This allows the modeller to efficiently encode the evolution of past time-series observations into the optimisation problem. In particular, we derive a concise formulation of the dynamic mean-variance criterion alongside an explicit solution in our setting, which naturally incorporates a drawdown control in the optimal strategy over a finite time horizon. Secondly, we draw parallels between classical portfolio stategies and Sig-Trading strategies and explain how the latter leads to a pathwise extension of the classical setting via the “Signature Efficient Frontier”. Finally, we give examples when trading under an exogenous signal as well as examples for momentum and pair-trading strategies, demonstrated both on synthetic and market data. Our framework combines the best of both worlds between classical theory (whose appeal lies in clear and concise formulae) and between modern, flexible data-driven methods that can handle more realistic datasets. The advantage of the added flexibility of the latter is that one can bypass common issues such as the accumulation of heteroskedastic and asymmetric residuals during the optimisation phase. Overall, Sig-Trading combines the flexibility of data-driven methods without compromising on the clarity of the classical theory and our presented results provide a compelling toolbox that yields superior results for a large class of trading strategies.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 6.0/10 (capped from 7.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 7.0)

Why this score: The paper employs advanced rough path theory (signatures, tensor algebra) and extends the mean-variance framework with path-dependent dynamics, indicating high mathematical complexity. It includes concrete implementation sections, numerical results on synthetic and market data for pairs trading, momentum, and exogenous signals, demonstrating strong empirical backing for a theoretical paper.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Integrate Path-Dependencies<br>into Mean-Variance Framework"] --> B{"Methodology: Signature Trading"}
  B --> C["Represent Strategy as Linear Functional<br>of Rough Path Signature"]
  C --> D["Derive Dynamic Mean-Variance<br>Optimization Formulation"]
  D --> E["Compute Solution with<br>Integrated Drawdown Control"]
  E --> F["Construct Signature<br>Efficient Frontier"]
  F --> G["Outcomes: Unified Framework<br>Interpretable Formulas & Superior Performance"]

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 Signature Trading (Sig‑Trading) framework that represents a trading strategy as a linear functional of a path signature and provides a closed‑form solution to the dynamic mean‑variance optimisation problem.

Source sentence
... Specifically, we achieve this by representing a trading strategy as a linear functional applied to the signature of a path (which we refer to as “Signature Trading” or “Sig‑Trading”).

Assumes 5

  • The underlying asset or signal need not be Markovian. model
    Source sentence
    the Sig‑Trading framework does not impose the restriction that the underlying asset or signal be Markovian
  • A filtered probability space is assumed for the stochastic processes. model
    Source sentence
    Let (Ω, F, (Ft)t∈[0,T], P) be a filtered probability space.
  • Asset dynamics are modelled as a diffusion process. model
    Source sentence
    In such stochastic control problems, the asset dynamics are explained by a diffusion process
  • No market impact (trading costs) is considered in the optimisation. execution
    Source sentence
    In this work, we do not consider market impact
Show 1 more
  • The signal and asset returns are assumed to have a well‑posed covariance structure. statistical
    Source sentence
    The performance of such a trading strategy is then mostly reliant on the predictive power of the signal that goes into the model, as well as the well‑posedness of the covariance structure

Data

No dataset named in the text.

Costs

Not modeled A trading or execution result is reported without costs.

Finds 4

  • The Sig‑Trader achieves higher Sharpe ratios than a comparable factor model, especially at higher truncation orders of the signature. performance
    Source sentence
    for higher orders of truncation, the Sharpe ratio improves as the Sig‑Trader can better approximate the non‑linear relationship between the signal and the underlying asset process
  • The closed‑form solution is lightweight, tractable and easily interpretable, enabling dynamic optimisation with drawdown control. theoretical
    Source sentence
    a closed form solution to the optimal mean‑variance Sig‑Trading strategy and derive a clear intuition for portfolio managers to navigate and use this formula in a path‑dependent context
  • Sig‑Trading combines the flexibility of data‑driven methods with the clarity of classical theory, yielding superior results for a large class of trading strategies. performance
    Source sentence
    our presented results provide a compelling toolbox that yields superior results for a large class of trading strategies
  • The framework does not require explicit prediction of returns, bypassing the accumulation of asymmetric residuals. theoretical
    Source sentence
    by lifting the mean‑variance optimisation into the lead‑lag signature space

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 55 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 3 papers in the archive (2 by independent authors): Randomized Signature Methods in Optimal Portfolio Selection (used as a baseline)

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.