Paper: arXiv 2311.15635

Authors: Kerstin Lamert, Benjamin R. Auer, Ralf Wunderlich

Abstract

This study evaluates the practical usefulness of continuous-time arbitrage strategies designed to exploit serial correlation in fractional financial markets. Specifically, we revisit the strategies of Shiryaev (1998) and Salopek (1998) and transfer them to a real-world setting by distretizing their dynamics and introducing transaction costs. In Monte Carlo simulations with various market and trading parameter settings as well as a formal analysis of discretization error, we show that both are promising with respect to terminal portfolio values and loss probabilities. These features and complementary sparsity make them worth serious consideration in the toolkit of quantitative investors.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper is mathematically dense, featuring advanced stochastic calculus (fBm SDEs, Riemann-Stieltjes integrals) and rigorous discretization error analysis, warranting a high math score. However, it lacks real-world data application, relying instead on Monte Carlo simulations with parameter sweeps and sensitivity checks, which is moderately rigorous but falls short of live backtesting, justifying a mid-range empirical score.

Research Flowchart

  flowchart TD
  A["Research Goal: Assess practicality of continuous-time arbitrage strategies for fractional markets with transaction costs"] --> B["Key Methodology: Discretization & Simulation"]
  B --> C["Computational Process: Monte Carlo simulations with various market/trading parameters"]
  C --> D["Computational Process: Formal analysis of discretization error"]
  D --> E["Key Findings: Both strategies are promising regarding terminal values & loss probabilities"]
  E --> F["Outcome: Suggested as viable tools for quantitative investors"]

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 to discretize the continuous-time arbitrage strategies of Shiryaev (1998) and Salopek (1998) and to incorporate transaction costs for real‑world application.

Source sentence
Specifically, we revisit the strategies of Shiryaev (1998) and Salopek (1998) and transfer them to a real-world setting by distretizing their dynamics and introducing transaction costs.

Assumes 2

  • All risky asset price processes start at time zero with identical initial prices. market
    Source sentence
    Assumption 2.1 All price processes (Si t)t∈[0,T] of the risky assets i = 1,...,d start at time t = 0 with identical initial prices Si 0 = es > 0.
  • Asset price processes are driven by independent fractional Brownian motions, implying uncorrelated prices at the terminal time. model
    Source sentence
    This implies the strict inequality VΨ T = Mβ(ST)−Mα(ST) > 0 with probability one such that arbitrage condition (iii) is also satisfied.

Data

  • Simulated Monte‑Carlo scenarios of discretized strategies synthetic unknown
    Source sentence
    144.7 226.2 0.0 0.5 59.4 581.3 4,477.2 0.00 (144.2) (222.9) (0.0) (0.5) (59.6) (575.7) (∞) (0.00) Φ (0,0) 109.4 222.8 −53.5 −32.9 25.7 537.9 4,392.8 0.39 (0.1,0) 91.9 219.7 −73.7 −48.1 9.2 514.2 4,339.9 0.46 (0.1,0.5) −17.3 221.0 −178.5 −157.9 −100.4 407.8 4,245.5 0.73 This table reports some descriptive statistics for the simulated terminal portfolio value distributions in Panel (a) of Figure 3.2. Besides the mean
  • Refinitiv Eikon daily price data for 1,689 emerging‑market stocks · Refinitiv Eikon market-prices commercial
    2013‑2023 · 1,689 full‑history stocks in 41 country sub‑indices
    Source sentence
    Specifically, we obtain an illustrative data set from Refinitiv Eikon which spans from January 2013 to December 2023 and covers daily price information of the 1,689 full-history stocks listed in the 41 country subindices of the Datastream Emerging Arbitrage strategies in markets with fractional Brownian motion 31 E[V Φ T ] (E[VΨ T −V Φ T ])/∆t2H−1 Frequency N simulated approximated simulated Monthly 12 323.2 304.6

Costs

Modeled Costs are part of the reported headline results.

Various transaction‑cost variants are considered, including proportional costs and minimum‑fee settings.

Source sentence
11 In Proposition 3.1, we show that, for example, discretizing the Shiryaev strategy (2.8) almost surely leads to a strictly positive capital requirement. Table 3.2: Shiryaev portfolio value statistics for different transaction costs 91.9 but remains positive.

Finds 5

  • Monte‑Carlo simulations show that the discretized strategies yield promising terminal portfolio values and low loss probabilities. performance
    Source sentence
    In Monte Carlo simulations with various market and trading parameter settings as well as a formal analysis of discretization error, we show that both are promising with respect to terminal portfolio values and loss probabilities.
  • For the Shiryaev strategy, loss probability can rise to 46% under certain cost settings but remains considered reasonably low. performance
    Source sentence
    Simultaneously, the loss probability rises to 46% but can still Arbitrage strategies in markets with fractional Brownian motion 17 be considered reasonably low (see Hogan et al., 2004).
  • As the Hurst exponent approaches 1, the discretized Salopek strategy converges to an almost perfect arbitrage strategy. theoretical
    Source sentence
    Put differently, even though discretiza-tion causes VΨ T −V Φ T̸ = 0, the discretized Salopek strategy converges to an almost perfect arbitrage strategy.
  • Empirically, the Salopek strategy achieves similar mean terminal wealth to Shiryaev but with significantly lower loss probabilities. performance
    Source sentence
    In comparison, the Salopek strategy shows the same mean terminal wealth signs but at significantly lower loss probabilities.
  • Higher trading frequencies improve mean terminal values and reduce loss probabilities even in the presence of transaction costs. performance
    Source sentence
    The resulting mean terminal portfolio values in Panel (a) imply that, even in the presence of transaction costs, higher trading frequencies are beneficial for investors. The loss probabilities in Panel (b) also recommend more frequent trading.

Checked by others

No other paper in the archive has been found to support or contradict these findings. 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.