arxiv: 2604.13334 · v1 · submitted 2026-04-14 · 💱 q-fin.TR

Recognition: unknown

Against a Universal Trading Strategy: No-Arbitrage, No-Free-Lunch, and Adversarial Cantor Diagonalization

Karl Svozil

Pith reviewed 2026-05-10 13:09 UTC · model grok-4.3

classification 💱 q-fin.TR

keywords no-arbitrageuniversal trading strategiesno-free-lunch theoremCantor diagonalizationequivalent martingale measureadversarial computationfinancial markets

0 comments

The pith

No trading strategy can guarantee strict profits across all possible market trajectories.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that a universal trading strategy—one generating strict profit across every possible market trajectory—cannot exist under standard market assumptions. Such a strategy would amount to strong arbitrage, which is ruled out when competitive markets admit an equivalent martingale measure. Combinatorially, the no-free-lunch result establishes that universal outperformance is impossible under uniform averaging over trajectories. Computationally, an adversary can diagonalize against any given computable algorithm to produce defeating price paths. A time-reversal heuristic links these financial constraints to thermodynamic detailed balance, while practical examples like the wheel options strategy illustrate dependence on transient regime assumptions that increase tail risks.

Core claim

Under standard admissibility constraints, the existence of such a strategy is a strict subset of strong arbitrage, which is mathematically precluded in competitive markets admitting an equivalent martingale measure. Beyond this, the No-Free-Lunch theorem demonstrates that outperformance requires exploitation of non-uniform market structure. A Turing diagonalization argument constructs an adversarial environment that defeats any computable trading algorithm. These limits are framed by a time-reversal heuristic that establishes a formal analogy between financial martingale measures and thermodynamic detailed balance, and are illustrated by the wheel options strategy succeeding only for all-pr.

What carries the argument

The reduction of universal strategies to strong arbitrage under admissibility constraints, together with adversarial Cantor diagonalization that defeats any computable trading algorithm.

If this is right

Any purported universal strategy must either be inadmissible or create arbitrage opportunities forbidden by equivalent martingale measures.
Strategies that succeed in practice depend on non-uniform or transient market structures rather than working universally.
Automated execution of regime-dependent strategies systematically amplifies tail risks.
The time-reversal heuristic implies that market dynamics maintain a detailed balance analogous to thermodynamic equilibrium.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Any general-purpose trading algorithm will remain vulnerable to adaptive market responses constructed against its own rules.
The results connect computational limits in finance to equilibrium principles that prevent extraction of perpetual advantage.
Relaxing the assumption of computable strategies might allow non-standard approaches, though this lies outside the paper's scope.

Load-bearing premise

Competitive markets admit an equivalent martingale measure that precludes strong arbitrage, or that trading algorithms are computable and subject to adaptive adversaries.

What would settle it

Constructing a computable trading algorithm that produces strict positive profit against every possible adaptive adversarial price path in a market model.

read the original abstract

We investigate the impossibility of universally winning trading strategies -- those generating strict profit across all market trajectories -- through three distinct mathematical paradigms. Fundamentally, under standard admissibility constraints, the existence of such a strategy is a strict subset of strong arbitrage, which is mathematically precluded in competitive markets admitting an equivalent martingale measure. Beyond this rigorous measure-theoretic foundation, we explore analogous limitations in two alternative modeling regimes. Combinatorially, the No-Free-Lunch theorem demonstrates that outperformance requires exploitation of non-uniform market structure, as uniform averaging precludes universal dominance. Computationally, a Turing diagonalization argument constructs an adversarial environment that defeats any computable trading algorithm, shifting the impossibility from exogenous price paths to adaptive adversaries. These mathematical limits are framed by a time-reversal heuristic that establishes a formal analogy between financial martingale measures and thermodynamic detailed balance, resolving the Maxwell's Demon analogy for markets without relying on physically irrelevant Landauer erasure costs. Using the Wheel Options Strategy as a case study, we demonstrate that strategies succeeding ``for all practical purposes'' (FAPP) inherently depend on transient regime assumptions, meaning their automated execution systematically amplifies tail risks.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper stitches together no-arbitrage, no-free-lunch, and a computability diagonal argument to show universal trading strategies cannot exist, with a time-reversal link to thermodynamics as the freshest piece.

read the letter

The core takeaway is straightforward: under standard admissibility, a strategy that wins on every path is just a strong arbitrage, which the fundamental theorem rules out once an equivalent martingale measure exists. The paper then runs the same conclusion through the no-free-lunch theorem (uniform markets cannot be dominated) and a Turing-style diagonal construction (any computable rule can be beaten by an adaptive price path). It closes with a heuristic that maps martingale measures to thermodynamic detailed balance, sidestepping Maxwell-demon worries without invoking erasure costs, and uses the Wheel Options Strategy to show how “for all practical purposes” rules still blow up tails when the regime shifts. That synthesis and the thermodynamic framing are the parts that feel new. The rest is competent application of known results rather than fresh derivations. The main soft spot is that the central claims rest on the existing literature without new proofs or exhaustive edge-case checks; the diagonal argument, for instance, is a direct transplant and the case study stays illustrative. Nothing circular or invented, but the computational section would benefit from more explicit construction of the adversary. Readers already comfortable with FTAP and basic recursion theory will get the most out of it; it is not a methods paper for practitioners. It is coherent and engages the literature honestly, so it deserves a serious referee who can ask for tighter derivations in the diagonalization part. I would send it to review.

Referee Report

2 major / 2 minor

Summary. The manuscript claims that no universal trading strategy—one generating strict profit across all market trajectories—can exist. This is established in three regimes: (1) measure-theoretically, such strategies form a strict subset of strong arbitrage and are precluded in markets admitting an equivalent martingale measure under standard admissibility; (2) combinatorially, the No-Free-Lunch theorem shows that uniform market structure prevents universal dominance; (3) computationally, a Turing diagonalization constructs an adaptive adversary defeating any computable trading algorithm. A time-reversal heuristic analogizes martingale measures to thermodynamic detailed balance (resolving Maxwell's Demon without Landauer costs), and the Wheel Options Strategy illustrates that apparent practical success relies on transient regime assumptions that amplify tail risks.

Significance. If the central derivations hold, the work unifies standard no-arbitrage and NFL results with computability arguments to bound universal strategies, offering a clear framework for why automated trading cannot guarantee outperformance. The thermodynamic analogy provides a novel interpretive lens, and the case study highlights practical risk implications. Strengths include reliance on established external theorems (FTAP, NFL) without ad-hoc parameters or invented entities.

major comments (2)

[Measure-theoretic foundation] The claim that a universal strategy is a strict subset of strong arbitrage (abstract and measure-theoretic section) is load-bearing for applying the EMM preclusion; the manuscript must supply an explicit definition of 'universal' versus 'strong arbitrage' and a short proof of the subset relation rather than asserting it, as edge cases (e.g., strategies that are admissible but not strictly positive) could affect the argument.
[Computational diagonalization] In the computational regime, the diagonalization argument defeats computable strategies via adaptive adversaries, but the manuscript does not address whether the impossibility extends beyond computable algorithms or how non-computable strategies (if admissible) would be treated; this boundary condition is central to the scope of the 'no universal strategy' conclusion.

minor comments (2)

[Case study] The Wheel Options Strategy case study would benefit from a concrete numerical example or simulation showing how regime assumptions lead to tail-risk amplification, rather than a qualitative description.
[Time-reversal heuristic] The time-reversal heuristic linking martingales to detailed balance is conceptually appealing but requires a more formal statement (e.g., an explicit mapping or theorem) to support the Maxwell's Demon resolution.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the manuscript's unification of no-arbitrage, NFL, and computability results. We address each major comment below and will incorporate the suggested clarifications into the revised manuscript.

read point-by-point responses

Referee: [Measure-theoretic foundation] The claim that a universal strategy is a strict subset of strong arbitrage (abstract and measure-theoretic section) is load-bearing for applying the EMM preclusion; the manuscript must supply an explicit definition of 'universal' versus 'strong arbitrage' and a short proof of the subset relation rather than asserting it, as edge cases (e.g., strategies that are admissible but not strictly positive) could affect the argument.

Authors: We agree that an explicit definition and short proof are needed to make the subset relation rigorous. In the revision we will define a universal trading strategy as one that produces strictly positive terminal wealth almost surely under every admissible price process. We will then supply a concise proof that any such strategy is necessarily a strong arbitrage (non-negative terminal value a.s. with positive probability of strict positivity), which is ruled out by the existence of an equivalent martingale measure under standard admissibility. Edge cases such as admissible but non-strictly-positive strategies will be explicitly excluded from the universal class. revision: yes
Referee: [Computational diagonalization] In the computational regime, the diagonalization argument defeats computable strategies via adaptive adversaries, but the manuscript does not address whether the impossibility extends beyond computable algorithms or how non-computable strategies (if admissible) would be treated; this boundary condition is central to the scope of the 'no universal strategy' conclusion.

Authors: The computational section is intentionally limited to computable trading algorithms, which are the only ones that can be realized by any physical or algorithmic trading system. We will add a clarifying paragraph stating that the Turing diagonalization applies exclusively to computable functions and that non-computable strategies, while mathematically conceivable, fall outside the scope of implementable trading and are therefore not addressed by the paper's conclusions on practical universal strategies. revision: yes

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's derivation chain relies on standard external results: the Fundamental Theorem of Asset Pricing precluding strong arbitrage (hence its strict subset, universal strategies) when an equivalent martingale measure exists; the No-Free-Lunch theorem requiring non-uniform structure for outperformance; and Turing diagonalization constructing an adaptive adversary defeating any computable strategy. These follow directly from the stated definitions of admissibility constraints and computability without reduction to self-defined quantities, fitted inputs renamed as predictions, or load-bearing self-citation chains. The time-reversal heuristic is presented as an analogy to thermodynamic balance, not a circular derivation step. The Wheel Options case study illustrates practical dependence on transient assumptions without internal circularity. The overall structure is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claims rest on standard domain assumptions from mathematical finance and computability theory rather than new free parameters or invented entities.

axioms (2)

domain assumption Competitive markets admit an equivalent martingale measure under standard admissibility constraints
Invoked in the first paradigm to preclude strong arbitrage.
domain assumption Trading strategies are computable algorithms
Required for the Turing diagonalization argument to construct an adversarial environment.

pith-pipeline@v0.9.0 · 5501 in / 1291 out tokens · 39191 ms · 2026-05-10T13:09:03.141480+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

17 extracted references · 1 canonical work pages

[1]

Bachelier, Th´ eorie de la sp´ eculation, Annales Scien- tifiques de l’ ´Ecole Normale Sup´ erieure17, 21 (1900)

L. Bachelier, Th´ eorie de la sp´ eculation, Annales Scien- tifiques de l’ ´Ecole Normale Sup´ erieure17, 21 (1900)

1900
[2]

P. A. Samuelson, Proof that properly anticipated prices fluctuate randomly, Industrial Management Review6, 41 (1965)

1965
[3]

E. F. Fama, Efficient capital markets: A review of the- ory and empirical work, The Journal of Finance25, 383 (1970)

1970
[4]

J. M. Harrison and S. R. Pliska, Martingales and stochas- tic integrals in the theory of continuous trading, Stochas- tic Processes and their Applications11, 215 (1981)

1981
[5]

Delbaen and W

F. Delbaen and W. Schachermayer, A general version of the fundamental theorem of asset pricing, Mathematische Annalen300, 463 (1994)

1994
[6]

D. H. Wolpert and W. G. Macready, No free lunch theo- rems for optimization, IEEE Transactions on Evolution- ary Computation1, 67 (1997)

1997
[7]

A. M. Turing, On computable numbers, with an appli- cation to the Entscheidungsproblem, Proceedings of the London Mathematical Society, Series 242, 43, 230 (1936- 7 and 1937)

1936
[8]

Landauer, Irreversibility and heat generation in the computing process, IBM Journal of Research and Devel- opment5, 183 (1961)

R. Landauer, Irreversibility and heat generation in the computing process, IBM Journal of Research and Devel- opment5, 183 (1961)

1961
[9]

C. H. Bennett, The thermodynamics of computation—a review, International Journal of Theoretical Physics21, 905 (1982)

1982
[10]

C. I. Isasa Mart´ ın,Computability-Based Analysis of Mar- ket Predictability, Trabajo de fin de m´ aster, Universidad Complutense de Madrid, Madrid, Spain (2022), m´ aster en M´ etodos Formales en Ingenier´ ıa Inform´ atica. Tutor: Ismael Rodr´ ıguez Laguna. Score: 9.8/10

2022
[11]

N. S. Yanofsky, A universal approach to self-referential paradoxes, incompleteness and fixed points, Bulletin of Symbolic Logic9, 362 (2003), arXiv:math/0305282

work page arXiv 2003
[12]

H. G. Rice, Classes of recursively enumerable sets and their decision problems, Transactions of the American Mathematical Society74, 358 (1953)

1953
[13]

K. J. Arrow,Social Choice and Individual Values, 3rd ed. (Yale University Press, New Haven, CT, 2012) foreword to the Third Edition by Eric S. Maskin

2012
[14]

J. S. Bell, Against ‘measurement’, Physics World3, 33 (1990)

1990
[15]

A. W. Lo, The adaptive markets hypothesis: Market effi- ciency from an evolutionary perspective, The Journal of Portfolio Management30, 15 (2004)

2004
[16]

N. F. Brady, J. C. Cotting, R. G. Kirby, J. R. Opel, and H. M. Stein,Report of the Presidential Task Force on 6 Market Mechanisms(U.S. Government Printing Office, Washington, D.C., 1988)

1988
[17]

Kirilenko, A

A. Kirilenko, A. S. Kyle, M. Samadi, and T. Tuzun, The flash crash: High-frequency trading in an electronic mar- ket, The Journal of Finance72, 967 (2017)

2017