Pith. sign in

REVIEW 2 major objections 5 minor 36 references

Stochastic integration with respect to arbitrary collections of continuous semimartingales and applications to Mathematical Finance

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read For arbitrary collections of continuous semimartingales, extended stochastic integrals are classified by a covariation-based reproducing kernel Hilbert space, and the classifying condition is exactly market viability.

desk verdict A genuinely new bijection between extended stochastic integrals and a stochastic RKHS—important, but the necessity proof of the central theorem has a real gap that needs repair. read the letter →

arxiv 1908.03946 v2 pith:HGD5LREI submitted 2019-08-11 math.PR q-fin.MF

Constantinos Kardaras This is my paper · ORCID
classification math.PRq-fin.MF MSC 60H0591G10
keywords infinite-dimensionalstochasticintegrationcontinuoussemimartingalesaggregatereproducingkernelHilbertspacesemimartingaletopologymarketviabilitylocalmartingaledeflatorsoptionaldecompositionhedgingduality
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proves that stochastic integration against an arbitrary collection of continuous semimartingales—possibly uncountably many—can be described fully by the covariation structure of the integrators. The space of all extended stochastic integrals, defined as the closure in the semimartingale topology, is shown to be in bijection with a stochastic reproducing kernel Hilbert space built from the covariation kernel, provided the drift of the integrators belongs to that space. That drift condition is exactly what is needed for an infinite-asset financial market to be viable, in the sense that no nontrivial payment stream can be financed with arbitrarily small capital. When it holds, the abstract closure becomes an operational class of integrals written in terms of covariations, and the classical theorems of finance—optional decomposition, hedging duality, and completeness—carry over to arbitrary asset collections.

What carries the argument

The carrying object is the stochastic aggregate reproducing kernel Hilbert space $R(C)$, defined as the set of finite-variation processes $F=(F^i; i\in I)$ for which the nondecreasing process $\int_0^\cdot \|dF\|^2_{dC}$ is finite, where $C_{ij}=[P^i,P^j]$ and the norm is the essential supremum over all finite subsets of $I$. This space plays the role of the integrand space: each $F\in R(C)$ is the aggregate covariation of exactly one extended stochastic integral. The construction proceeds by finite-dimensional approximation of the kernel and uses a lemma on essential suprema of directed families of nondecreasing processes to extend from countable to arbitrary index sets; the reproducing-kernel identity $F_j=\int_0^\cdot \langle dC^{I_j},dF\rangle_{dC}$ is what makes a covariation collection determine its integral.

What would settle it

Exhibit a stochastic aggregate kernel and an essentially bounded directed family of continuous nondecreasing processes whose essential supremum is not continuous or not of finite variation; alternatively, give a collection $P$ of continuous semimartingales with $A\in R(C)$ for which two different elements of $S(P)$ have identical aggregate covariations with $P$. Either would directly contradict Theorem 2.3.

Watch

Extended reading notes

Core claim

Given $P=(P^i; i\in I)$ of continuous semimartingales with covariation kernel $C_{ij}=[P^i,P^j]$, the paper's central result (Theorem 2.3) is that the map $S(P)\ni Z\mapsto ([Z,P^i]; i\in I)$ is a bijection onto the stochastic aggregate reproducing kernel Hilbert space $R(C)$ if and only if the drift $A$ of the Doob–Meyer decomposition of $P$ lies in $R(C)$. In that case every element of $S(P)$ is of the form $\int_0^\cdot \langle dF,dA\rangle_{dC}+\int_0^\cdot \langle dF,dM\rangle_{dC}$ for a unique $F\in R(C)$, and $S(P)$ is topologically isomorphic to $R(C)$. In financial terms, a market with arbitrary assets is viable exactly when this drift condition holds, and wealth processes correspond one-to-one to covariation integrands.

Load-bearing premise

The load-bearing premise is the lemma that an essentially bounded family of continuous nondecreasing processes indexed by an arbitrary directed set has a continuous nondecreasing essential supremum reached along a monotone sequence; without it, $R(C)$ cannot be constructed for uncountable index sets and the bijection collapses.

Editorial extensions

If this is right

  • Viability of an infinite-asset market is equivalent to the drift condition $A\in R(C)$, and also to existence of a local martingale deflator (Theorem 3.3).
  • The optional decomposition theorem holds: a nonnegative process is a local supermartingale under every local martingale deflator exactly when it is a wealth process with a capital-withdrawal stream (Theorem 3.6).
  • Hedging duality holds: the minimal hedge value equals the supremum over local martingale deflators of the expected discounted payoff stream, and a minimal hedge exists whenever that value is finite (Theorem 3.9).
  • A viable market is complete exactly when there is a unique local martingale deflator (Theorem 3.11).
  • In Heath–Jarrow–Morton bond markets with possibly uncountably many maturities, viability is characterized by the integral condition (3.18), which is weaker than the classical HJM drift restriction.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same covariation-kernel construction may extend to discontinuous semimartingales once a jump-compatible analogue of the essential-supremum lemma is found; the paper itself notes that jumps make the theory harder.
  • Because the structural condition is weaker than no-free-lunch-with-vanishing-risk, the theorem predicts that infinite-asset markets can be viable while classical NFLVR fails, so the choice of viability definition is what preserves the finite-asset theorems.
  • The bijection between wealth processes and covariation integrands suggests a numerical strategy: discretize the covariation kernel rather than the trading positions, which may be more stable in infinite-dimensional models.
  • A testable extension is to check in concrete infinite-asset models whether $A\in R(C)$ reduces to a familiar integrability condition on volatility or forward-rate curves, giving a practical viability criterion.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper develops a theory of stochastic integration with respect to an arbitrary collection P=(P^i; i∈I) of continuous semimartingales, without cardinality or state-space assumptions on the index set I. It constructs a "stochastic aggregate reproducing kernel Hilbert space" R(C) from the covariation kernel C_{ij}=[P^i,P^j], and proves (Theorem 2.3) that the map Z↦([Z,P^i]; i∈I) from the S-closed space S(P) of extended stochastic integrals to R(C) is a bijection if and only if the drift vector A of P lies in R(C). Under this condition, S(P) is characterized as all processes of the form ∫⟨dF,dA⟩+∫⟨dF,dM⟩ with F∈R(C), and S(P) is topologically isomorphic to R(C). The paper then applies this structure to infinite-asset markets, proving a fundamental theorem of asset pricing (viability ⇔ existence of local martingale deflators ⇔ A∈R(C)), optional decomposition, hedging duality, market completeness, and a Heath-Jarrow-Morton bond-market example.

Significance. If the central theorem is fully proved, this is a significant contribution: it gives an operational, kernel-based description of the S-closure of finite-dimensional stochastic integrals for arbitrary index sets, and it extends the principal theorems of continuous-time mathematical finance to that setting. The construction of R(C) from finite-dimensional approximations is carefully developed, with a self-contained static RKHS appendix and no fitted parameters. The financial results identify a sharp structural condition A∈R(C) and show that this same condition yields local martingale deflators, optional decomposition, and hedging duality. The delicate Lemma 1.3, on which the definition of R(C) relies, appears to be proved correctly; the directedness assumptions are used exactly where needed. The main caveat is that the necessity direction of Theorem 2.3 currently contains two proof gaps, and these are load-bearing for the whole paper.

major comments (2)
  1. [§2.5, Theorem 2.3, Step 2] The proof asserts the existence of a nondecreasing sequence (J_n) in Fin(I) and a sequence (Π_n) of predictable disjoint sets such that, with η_n=ν^{J_n}1_{Π_n}, the processes V_n=∫⟨η_n,dA_{J_n}⟩ are finitely valued and satisfy P[V_n(T)≤exp(2n)|Γ]≤2^{-n-1}. This assertion is essential: disjointness of the Π_n is used to obtain [Y^n,Y^m]=0 for m≠n, and the lower bounds on V_n drive B^n(T)>n on the positive-probability set Λ. No construction is supplied, and the stated facts do not follow immediately from Remark 1.4, which only gives a sequence of finite sets along which the essential supremum diverges, not disjoint predictable supports for the truncated energy processes. Please provide a complete argument, or a reference containing one.
  2. [§2.5, Theorem 2.3, Step 2] The sentence "however, by its definition this sequence should converge to Z" is incorrect. The integrands 1_{∪_{k≤n}Π_k} converge pointwise to 1_{∪_{k∈N}Π_k}, not to 1, so Z_n converges to ∫ 1_{∪_{k∈N}Π_k} dZ, not necessarily to Z. This invalidates the contradiction as written. The argument can likely be repaired by observing that the finite-variation part of the limit would have to dominate B=lim_n B_n, which is infinite on Λ, but this step is not present in the manuscript. Please correct the convergence claim and complete the contradiction.
minor comments (5)
  1. [§2.4, Proposition 2.2] In the proof, "P-lim_{n→∞}[Z_n−Z,Z_n−Z](T)=0 for all T∈R" should read "for all T∈R_+".
  2. [§2.5, Theorem 2.3, Step 2] The symbol R^i appearing in the definition of F should be P^i.
  3. [§3.5, Theorem 3.9] The proof is only a sketch and sends the reader to [Kra96, Proposition 4.3]. Since hedging duality is one of the headline applications, please either expand the proof or state precisely which imported results are used.
  4. [§3.6, after Theorem 3.11] The displayed sentence beginning "Since the process YK − ..." is garbled and should be rewritten.
  5. [Abstract] "constructed though finite-dimensional approximation" should be "constructed through finite-dimensional approximation".

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central S(P)-R(C) bijection is derived in-paper from the covariation-kernel construction; self-citations in the finance section are to independent finite-dimensional prior results.

full rationale

Walking the derivation chain, Theorem 2.3 does not assume its conclusion. R(C) is built from the stochastic aggregate kernel C through Definition 1.1, Lemma 1.3, (1.10)-(1.12), and the finite-dimensional rkHs approximation of Appendix A; the bijection S(P) ∋ Z ↦ ([Z,P^i]) ∈ R(C) is then proved via Lemma 2.1, the finite-index stochastic-integration argument, and approximation along (J_n). The sufficiency of A∈R(C) is proved in Proposition 2.2, and the necessity is attempted by contradiction in Step 2 of Theorem 2.3; no displayed equation reduces the theorem to its own input. There are no fitted parameters, no empirical predictions, and no ansatz smuggled in by citation. The financial results do cite the author's own earlier work ([Kar10, Theorem 4], [KP11, §2.3], [KK15, Lemma 2.1]) for finite-asset/continuous-semimartingale base cases, but those are externally published finite-dimensional results whose assumptions do not include the present infinite-asset target, so they are supporting external lemmas rather than a self-citation chain that forces the conclusion. The reviewer-noted gaps in Step 2 of Theorem 2.3—the asserted existence of disjoint predictable sets and the misstated convergence of Z_n—are proof-completeness or correctness issues, not circularity: the assertion is a construction lemma, and the misstated limit does not substitute the desired statement A∈R(C) for an input.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

This is a pure mathematics paper: there are no fitted parameters, no data, no code. The central claim depends on standard stochastic analysis (Émery topology, Kunita-Watanabe decomposition, Dambis-Dubins-Schwarz, Yor formula) and on previously published results, several by the author, used as black boxes in the finance section. The newly introduced object R(C) is rigorously defined in Definition 1.1 and constructed from the ground up in Section 1, so it is not an unexplained postulate. The main philosophical input is the structural condition A ∈ R(C), which is not assumed ad hoc: it is proved equivalent to bijectivity of the integral map and to market viability.

assumptions (7)
  • standard math Finite-dimensional vector stochastic integration theory and the Émery S-topology: for finite I, S(P) is closed and equals the set of all stochastic integrals against P.
    Used in §2.1 and §2.3 as the starting point for finite-dimensional approximations.
  • standard math Kunita-Watanabe decomposition and the Hilbert space structure of stable subspaces of continuous local martingales.
    Used in Lemma 2.1 and in the proof of bijectivity in §2.3 to construct projections and pass to the limit.
  • domain assumption Continuous semimartingale decomposition P = A + M with A continuous finite variation and M continuous local martingale; covariations [P_i, P_j] are continuous finite variation.
    Standing assumption throughout; the paper does not treat jumps, and states in the introduction that the jump case is more delicate.
  • domain assumption For finite-asset markets, viability (no arbitrage of the first kind) is equivalent to existence of a local martingale deflator, and to the structural condition for that finite market ([Kar10], [KP11]).
    Used in Theorem 3.3 to bootstrap from finite submarkets to arbitrary I. These are the author's own published results, cited rather than reproved.
  • domain assumption Lemma 2.1 of [KK15] characterizes nondecreasing processes through local submartingale properties under deflators.
    Used in the proof of optional decomposition (Theorem 3.6) to conclude K ∈ FV^{≽}.
  • standard math Dambis-Dubins-Schwarz representation, the scaling property of Brownian motion, and Yor's formula for stochastic exponentials.
    Used in Theorem 3.3 (unbounded wealth construction) and in §3.2 (structure of deflators).
  • standard math Moore-Aronszajn theorem for reproducing kernel Hilbert spaces.
    Cited in Appendix A for the equivalence of kernel-based and completion-based RKHS definitions; the paper actually develops the concrete approximation route.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stochastic integration with respect to arbitrary collections of continuous semimartingales and applications to Mathematical Finance." pith.science (2026). https://pith.science/paper/HGD5LREI

@misc{pith2026190803946,
  author       = {Pith},
  title        = {Pith review of: Stochastic integration with respect to arbitrary collections of continuous semimartingales and applications to Mathematical Finance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HGD5LREI}},
  note         = {Machine review of arXiv:1908.03946}
}
abstract

Stochastic integrals are defined with respect to a collection $P = (P_i; \, i \in I)$ of continuous semimartingales, imposing no assumptions on the index set $I$ and the subspace of $\mathbb{R}^I$ where $P$ takes values. The integrals are constructed though finite-dimensional approximation, identifying the appropriate local geometry that allows extension to infinite dimensions. For local martingale integrators, the resulting space $\mathsf{S} (P)$ of stochastic integrals has an operational characterisation via a corresponding set of integrands $\mathsf{R} (C)$, constructed with only reference the covariation structure $C$ of $P$. This bijection between $\mathsf{R} (C)$ and the (closed in the semimartingale topology) set $\mathsf{S} (P)$ extends to families of continuous semimartingale integrators for which the drift process of $P$ belongs to $\mathsf{R} (C)$. In the context of infinite-asset models in Mathematical Finance, the latter structural condition is equivalent to a certain natural form of market viability. The enriched class of wealth processes via extended stochastic integrals leads to exact analogues of optional decomposition and hedging duality as the finite-asset case. A corresponding characterisation of market completeness in this setting is provided.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [1]

    Aliprantis and Kim C

    Charalambos D. Aliprantis and Kim C. Border, Infinite dimensional analysis, third ed., Springer, Berlin, 2006, A hitchhiker's guide

    work page 2006

  2. [2]

    Nachman Aronszajn, Theory of reproducing kernels, Trans. Amer. Math. Soc. 68 (1950), 337--404

    work page 1950

  3. [3]

    1 (1997), no

    Tomas Bj\" o rk, Giovanni Di Masi, Yuri Kabanov, and Wolfgang Runggaldier, Towards a general theory of bond markets, Finance Stoch. 1 (1997), no. 2, 141--174

    work page 1997

  4. [4]

    Alain Berlinet and Christine Thomas-Agnan, Reproducing kernel hilbert spaces in probability and statistics, Springer US, 2003

    work page 2003

  5. [5]

    Christa Cuchiero, Irene Klein, and Josef Teichmann, A new perspective on the fundamental theorem of asset pricing for large financial markets, Theory Probab. Appl. 60 (2016), no. 4, 561--579

    work page 2016

  6. [6]

    1857, Springer, Berlin, 2005, pp

    Alexander Cherny and Albert Shiryaev, On stochastic integrals up to infinity and predictable criteria for integrability, S\' e minaire de P robabilit\' e s XXXVIII , Lecture Notes in Math., vol. 1857, Springer, Berlin, 2005, pp. 165--185

    work page 2005

  7. [7]

    Carmona and Michael R

    Ren\' e A. Carmona and Michael R. Tehranchi, Interest rate models: an infinite dimensional stochastic analysis perspective, Springer Finance, Springer-Verlag, Berlin, 2006

    work page 2006

  8. [8]

    Marzia De Donno, Paolo Guasoni, and Maurizio Pratelli, Super-replication and utility maximization in large financial markets, Stochastic Process. Appl. 115 (2005), no. 12, 2006--2022

    work page 2005

Show all 36 references
  1. [9]

    152, Cambridge University Press, Cambridge, 2014

    Giuseppe Da Prato and Jerzy Zabczyk, Stochastic equations in infinite dimensions, second ed., Encyclopedia of Mathematics and its Applications, vol. 152, Cambridge University Press, Cambridge, 2014

    2014

  2. [10]

    Freddy Delbaen and Walter Schachermayer, A general version of the fundamental theorem of asset pricing, Math. Ann. 300 (1994), no. 3, 463--520

    1994

  3. [11]

    S trasbourg, S trasbourg, 1977/78), Lecture Notes in Math., vol

    Michel \' E mery, Une topologie sur l'espace des semimartingales, S\'eminaire de P robabilit\'es, XIII ( U niv. S trasbourg, S trasbourg, 1977/78), Lecture Notes in Math., vol. 721, Springer, Berlin, 1979, pp. 260--280

    1977

  4. [12]

    Control Optim

    Nicole El Karoui and Marie-Claire Quenez, Dynamic programming and pricing of contingent claims in an incomplete market, SIAM J. Control Optim. 33 (1995), no. 1, 29--66

    1995

  5. [13]

    Theory Related Fields 109 (1997), no

    Hans F \"o llmer and Dmitry Kramkov, Optional decompositions under constraints, Probab. Theory Related Fields 109 (1997), no. 1, 1--25

    1997

  6. [14]

    1, 77--105

    David Heath, Robert Jarrow, and Andrew Morton, Bond pricing and the term structure of interest rates: A new methodology for contingent claims valuation, Econometrica 60 (1992), no. 1, 77--105

    1992

  7. [15]

    Michael Harrison and Stanley R

    J. Michael Harrison and Stanley R. Pliska, Martingales and stochastic integrals in the theory of continuous trading, Stochastic Process. Appl. 11 (1981), no. 3, 215--260

    1981

  8. [16]

    Publ., River Edge, NJ, 1997, pp

    Yuri Kabanov, On the FTAP of K reps- D elbaen- S chachermayer , Statistics and control of stochastic processes ( M oscow, 1995/1996), World Sci. Publ., River Edge, NJ, 1997, pp. 191--203

    1995

  9. [17]

    Constantinos Kardaras, Finitely additive probabilities and the fundamental theorem of asset pricing, Contemporary quantitative finance, Springer, Berlin, 2010, pp. 19--34

    2010

  10. [18]

    Veroyatnost

    Yuri Kabanov and Dmitry Kramkov, Large financial markets: asymptotic arbitrage and contiguity, Teor. Veroyatnost. i Primenen. 39 (1994), no. 1, 222--229

    1994

  11. [19]

    2 (1998), no

    , Asymptotic arbitrage in large financial markets, Finance Stoch. 2 (1998), no. 2, 143--172

    1998

  12. [20]

    11 (2007), no

    Ioannis Karatzas and Constantinos Kardaras, The num\' e raire portfolio in semimartingale financial models , Finance Stoch. 11 (2007), no. 4, 447--493

    2007

  13. [21]

    , Optional decomposition for continuous semimartingales under arbitrary filtrations, Electron. Commun. Probab. 20 (2015), no. 59, 10

    2015

  14. [22]

    20 (2016), no

    Yuri Kabanov, Constantinos Kardaras, and Shiqi Song, No arbitrage of the first kind and local martingale num\' e raires , Finance Stoch. 20 (2016), no. 4, 1097--1108

    2016

  15. [23]

    Constantinos Kardaras and Eckhard Platen, On the semimartingale property of discounted asset-price processes, Stochastic Process. Appl. 121 (2011), no. 11, 2678--2691

    2011

  16. [24]

    Theory Related Fields 105 (1996), no

    Dmitry Kramkov, Optional decomposition of supermartingales and hedging contingent claims in incomplete security markets, Probab. Theory Related Fields 105 (1996), no. 4, 459--479

    1996

  17. [25]

    Shreve, Brownian M otion and S tochastic C alculus , second ed., Graduate Texts in Mathematics, vol

    Ioannis Karatzas and Steven E. Shreve, Brownian M otion and S tochastic C alculus , second ed., Graduate Texts in Mathematics, vol. 113, Springer-Verlag, New York, 1991

    1991

  18. [26]

    Dmitry Kramkov and Walter Schachermayer, The asymptotic elasticity of utility functions and optimal investment in incomplete markets, Ann. Appl. Probab. 9 (1999), no. 3, 904--950

    1999

  19. [27]

    , Necessary and sufficient conditions in the problem of optimal investment in incomplete markets, Ann. Appl. Probab. 13 (2003), no. 4, 1504--1516

    2003

  20. [28]

    2, Walter de Gruyter & Co., Berlin-New York, 1982, A course on stochastic processes

    Michel M \' e tivier, Semimartingales, de Gruyter Studies in Mathematics, vol. 2, Walter de Gruyter & Co., Berlin-New York, 1982, A course on stochastic processes

    1982

  21. [29]

    19 (2015), no

    Oleksii Mostovyi, Necessary and sufficient conditions in the problem of optimal investment with intermediate consumption, Finance Stoch. 19 (2015), no. 1, 135--159

    2015

  22. [30]

    Rozovskii, Normalized stochastic integrals in topological vector spaces, S\' e minaire de P robabilit\' e s, XXXII , Lecture Notes in Math., vol

    Remigijus Mikulevicius and Boris L. Rozovskii, Normalized stochastic integrals in topological vector spaces, S\' e minaire de P robabilit\' e s, XXXII , Lecture Notes in Math., vol. 1686, Springer, Berlin, 1998, pp. 137--165

    1998

  23. [31]

    Paulsen and Mrinal Raghupathi, An introduction to the theory of reproducing kernel H ilbert spaces , Cambridge Studies in Advanced Mathematics, vol

    Vern I. Paulsen and Mrinal Raghupathi, An introduction to the theory of reproducing kernel H ilbert spaces , Cambridge Studies in Advanced Mathematics, vol. 152, Cambridge University Press, Cambridge, 2016

    2016

  24. [32]

    293, Springer-Verlag, Berlin, 1999

    Daniel Revuz and Marc Yor, Continuous M artingales and B rownian M otion , third ed., Grundlehren der Mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences], vol. 293, Springer-Verlag, Berlin, 1999

    1999

  25. [33]

    Martin Schweizer, On the minimal martingale measure and the F \" o llmer- S chweizer decomposition , Stochastic Anal. Appl. 13 (1995), no. 5, 573--599

    1995

  26. [34]

    1686, Springer, Berlin, 1998, pp

    Christophe Stricker and Jia An Yan, Some remarks on the optional decomposition theorem, S\'eminaire de P robabilit\'es, XXXII , Lecture Notes in Math., vol. 1686, Springer, Berlin, 1998, pp. 56--66

    1998

  27. [35]

    18 (2014), no

    Koichiro Takaoka and Martin Schweizer, A note on the condition of no unbounded profit with bounded risk, Finance Stoch. 18 (2014), no. 2, 393--405

    2014

  28. [36]

    4, 543--551

    Mark Veraar, The stochastic F ubini theorem revisited , Stochastics 84 (2012), no. 4, 543--551

    2012

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.