Abstract
In the context of a general semimartingale model, we aim at determining how much an investor is willing to pay to learn additional information that allows achieving arbitrage. If such a value exists, we call it the value of informational arbitrage. We are interested in the case where the information yields arbitrage opportunities but not unbounded profits with bounded risk. As in Amendinger et al. (Finance Stoch. 7:29–46, 2003), we rely on an indifference valuation approach and study optimal consumption–investment problems under initial information and arbitrage. We establish some new results on models with additional information and characterise when the value of informational arbitrage is universal.
Similar content being viewed by others
Notes
The longer preprint version is available online on arXiv at https://arxiv.org/abs/1804.00442v1.
For ease of notation, we omit to indicate explicitly the dependence on \(\omega \) in the utility stochastic field \(U\).
Similarly as above, for simplicity of notation, we omit to write explicitly the dependence on \(\omega \) in the stochastic field \(I\).
In view of Corollary 3.8, a fully explicit representation of the utility indifference value for \(k\neq 0\) can be obtained when the random variable \(\mathbb{E}[Z_{T}/q^{L}_{T}|\mathcal{G}_{0}]\) is a.s. constant or, equivalently, when \(\mathbb{Q}[q^{x}_{T}>0]\) does not depend on \(x\).
References
Acciaio, B., Fontana, C., Kardaras, C.: Arbitrage of the first kind and filtration enlargements in semimartingale financial models. Stoch. Process. Appl. 126, 1761–1784 (2016)
Aksamit, A., Choulli, T., Jeanblanc, M.: On an optional semimartingale decomposition and the existence of a deflator in an enlarged filtration. In: Donati-Martin, C., et al. (eds.) Memoriam Marc Yor—Séminaire de Probabilités XLVII. Lecture Notes in Mathematics, vol. 2137, pp. 187–218. Springer, Berlin (2015)
Amendinger, J.: Martingale representation theorems for initially enlarged filtrations. Stoch. Process. Appl. 89, 101–116 (2000)
Amendinger, J., Becherer, D., Schweizer, M.: A monetary value for initial information in portfolio optimization. Finance Stoch. 7, 29–46 (2003)
Amendinger, J., Imkeller, P., Schweizer, M.: Additional logarithmic utility of an insider. Stoch. Process. Appl. 75, 263–286 (1998)
Ankirchner, S.: Information and Semimartingales. Ph.D. Thesis, Humboldt University Berlin (2005). Available online at https://edoc.hu-berlin.de/handle/18452/15948
Ankirchner, S., Dereich, S., Imkeller, P.: The Shannon information of filtrations and the additional logarithmic utility of insiders. Ann. Probab. 34, 743–778 (2006)
Ankirchner, S., Imkeller, P.: Finite utility on financial markets with asymmetric information and structure properties of the price dynamics. Stoch. Process. Appl. 41, 479–503 (2005)
Auerhan, J., Lepingle, D., Yor, M.: Construction d’une martingale réelle continue, de filtration naturelle donnée. In: Azéma, J., Yor, M. (eds.) Séminaire de Probabilités XIV. Lecture Notes in Mathematics, vol. 784, pp. 200–204. Springer, Berlin (1980)
Beneš, V.: Existence of optimal strategies based on specified information, for a class of stochastic decision problems. SIAM J. Control Optim. 8, 179–188 (1970)
Chau, H., Cosso, A., Fontana, C., Mostovyi, O.: Optimal investment with intermediate consumption under no unbounded profits with bounded risk. J. Appl. Probab. 54, 710–719 (2017)
Chau, H., Runggaldier, W., Tankov, P.: Arbitrage and utility maximization in market models with an insider. Math. Financ. Econ. 12, 589–614 (2018)
Choulli, T., Deng, J., Ma, J.: How non-arbitrage, viability and numéraire portfolio are related. Finance Stoch. 19, 719–741 (2015)
Choulli, T., Yansori, S.: Log-optimal portfolio without NFLVR: existence, complete characterization, and duality. Preprint (2018). Available online at https://arxiv.org/abs/1807.06449
Danilova, A., Monoyios, M., Ng, A.: Optimal investment with inside information and parameter uncertainty. Math. Financ. Econ. 3, 13–38 (2010)
Delbaen, F., Schachermayer, W.: A general version of the fundamental theorem of asset pricing. Math. Ann. 300, 463–520 (1994)
Delbaen, F., Schachermayer, W.: The fundamental theorem of asset pricing for unbounded stochastic processes. Math. Ann. 312, 215–250 (1998)
Fontana, C.: Weak and strong no-arbitrage conditions for continuous financial markets. Int. J. Theor. Appl. Finance 18, 1550005 (2015)
Fontana, C.: The strong predictable representation property in initially enlarged filtrations under the density hypothesis. Stoch. Process. Appl. 128, 1007–1033 (2018)
Grorud, A., Pontier, M.: Insider trading in a continuous time market model. Int. J. Theor. Appl. Finance 1, 331–347 (1998)
Grorud, A., Pontier, M.: Asymmetrical information and incomplete markets. Int. J. Theor. Appl. Finance 4, 285–302 (2001)
He, S., Wang, J., Yan, J.: Semimartingale Theory and Stochastic Calculus. Kexue Chubanshe (Science Press), Beijing (1992)
Jacod, J.: Grossissement initial, hypothèse (H’), et théorème de Girsanov. In: Jeulin, T., Yor, M. (eds.) Grossissements de Filtrations: Exemples et Applications. Lecture Notes in Mathematics, vol. 1118, pp. 15–35. Springer, Berlin (1985)
Jacod, J., Shiryaev, A.: Limit Theorems for Stochastic Processes, 2nd edn. Springer, Berlin (2003)
Jeulin, T.: Semi-martingales et Grossissement d’une Filtration. Lecture Notes in Mathematics, vol. 833. Springer, Berlin (1980)
Karatzas, I., Kardaras, C.: The numéraire portfolio in semimartingale financial models. Finance Stoch. 11, 447–493 (2007)
Karatzas, I., Shreve, S.: Methods of Mathematical Finance. Springer, New York (1998)
Kardaras, C., Platen, E.: On the semimartingale property of discounted asset-price processes. Stoch. Process. Appl. 121, 2678–2691 (2011)
La Valle, I.: On cash equivalents and information evaluation in decisions under uncertainty: Part I: Basic theory. J. Am. Stat. Assoc. 63, 252–276 (1968)
Liu, J., Peleg, E., Subrahmanyam, A.: Information, expected utility, and portfolio choice. J. Financ. Quant. Anal. 45, 1221–1251 (2010)
Morris, J.: The logarithmic investor’s decision to acquire costly information. Manag. Sci. 21, 383–391 (1974)
Muraviev, R.: Exponential utility with non-negative consumption. Preprint (2019). Available online at http://arxiv.org/abs/1106.3006
Pikovsky, I., Karatzas, I.: Anticipative portfolio optimization. J. Appl. Probab. 28, 1095–1122 (1996)
Rheinländer, T., Schweizer, M.: On \({L}^{2}\)-projections on a space of stochastic integrals. Ann. Probab. 25, 1810–1831 (1997)
Stricker, C., Yan, J.: Some remarks on the optional decomposition theorem. In: Azéma, J., et al. (eds.) Séminaire de Probabilités XXXII. Lecture Notes in Mathematics, vol. 1686, pp. 56–66. Springer, Berlin (1998)
Stricker, C., Yor, M.: Calcul stochastique dépendant d’un paramètre. Z. Wahrscheinlichkeitstheor. Verw. Geb. 45, 109–133 (1978)
Takaoka, K., Schweizer, M.: A note on the condition of no unbounded profit with bounded risk. Finance Stoch. 18, 393–405 (2014)
Willinger, M.: Risk aversion and the value of information. J. Risk Insur. 56, 320–328 (1989)
Acknowledgements
The authors are thankful to Albina Danilova, Martin Larsson, Wolfgang Runggaldier, Martin Schweizer and seminar participants at ENSAE ParisTech, ETH Zürich and LSE for fruitful discussions and suggestions on the topic of the present paper. Valuable comments and suggestions by two anonymous referees, an Associate Editor and the Editor are gratefully acknowledged. Huy N. Chau was supported by the “Lendület” grant LP2015-6 of the Hungarian Academy of Sciences and by the NKFIH (National Research, Development and Innovation Office, Hungary) grant KH 126505.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: Proofs of the results in Sect. 3
Appendix: Proofs of the results in Sect. 3
Proof of Lemma 3.3
Let \((\vartheta ,c)\in \mathcal{A}^{\mathbf{H},k}(v)\). For ease of notation, we set \(V:=V^{v+k,\vartheta ,c}\), \(C:=\int _{0}^{\cdot }c_{u}\,\!\mathrm{d} \kappa _{u}\) and \(\widetilde{C}:=\int _{0}^{\cdot }Z^{\mathbf{H}}_{u}\, \!\mathrm{d}C_{u}\). Integration by parts gives for all \(t\in [0,T]\) that
Since \(Z^{\mathbf{H}}\in \mathcal{M}_{\text{loc}}(\mathbb{P}, \mathbf{H})\) and \(Z^{\mathbf{H}}S\in \mathcal{M}_{\text{loc}}( \mathbb{P},\mathbf{H})\), this implies that \(Z^{\mathbf{H}}V + \widetilde{C}\) is a sigma-martingale on \((\Omega ,\mathbf{H}, \mathbb{P})\) (see e.g. Fontana [18, Lemma 4.2]). Being nonnegative, it is also a supermartingale. Therefore, since \(V^{v,\vartheta ,c}_{T}\geq 0\) a.s., we get
so that \(\mathbb{E}[\widetilde{C}_{T}|\mathcal{H}_{0}]\leq v+k(1- \mathbb{E}[Z^{\mathbf{H}}_{T}|\mathcal{H}_{0}])\) a.s. Conversely, let \(C:=\int _{0}^{\cdot }c_{u}\,\!\mathrm{d}\kappa _{u}\) and suppose that
Consider the process \(\widehat{V}=(\widehat{V}_{t})_{t\in [0,T]}\) defined for \(t\in [0,T]\) by
The process \(\widehat{V}\) is well defined as an element of \(\mathcal{M}_{\text{loc}}(\mathbb{P},\mathbf{H})\). Due to Assumption 2.1 (and Proposition 2.5, in the case \(\mathbf{H}= \mathbf{G}\)), there exists \(\psi \in L(S,\mathbf{H})\) such that
The process \(V^{v+k,\psi ,c}=(V^{v+k,\psi ,c}_{t})_{t\in [0,T]}\) associated to the pair \((\psi ,c)\) satisfies
a.s. for all \(t\in [0,T]\). By construction, we have \(Z^{\mathbf{H}} _{t}V_{t}^{v+k,\psi ,c}\geq 0\) a.s. for all \(t\in [0,T]\) and \(Z^{\mathbf{H}}_{T}V^{v,\psi ,c}_{T}\geq 0\) a.s. Therefore \((\psi ,c)\in \mathcal{A}^{\mathbf{H},k}(v)\), proving that \(c\in \mathcal{C}^{\mathbf{H},k}(v)\). □
Proof of Proposition 3.5
Under the given assumptions, the process \(c^{\mathbf{H}}=(c^{\mathbf{H}}_{t})_{t\in [0,T]}\) satisfies
so that \(c^{\mathbf{H}}\in \mathcal{C}^{\mathbf{H},k}(v)\) by Lemma 3.3. Consider an arbitrary consumption process \(c\in \mathcal{C}^{\mathbf{H},k}(v)\). By the Fenchel–Legendre duality (see e.g. Karatzas and Shreve [27, Lemma 3.4.3]), the definitions of the stochastic field \(I\) and of the process \(c^{ \mathbf{H}}\) imply that
Therefore, it holds that
where the last inequality follows from the fact that in view of Lemma 3.3,
The result then follows by the arbitrariness of \(c\in \mathcal{C}^{ \mathbf{H},k}(v)\). □
Proof of Corollary 3.8
In view of Proposition 3.5, to compute \(u^{\mathbf{H},k}(v)\), it suffices to find the \(\mathcal{H}_{0}\)-measurable random variable \(\Lambda ^{\mathbf{H},k}(v)\) satisfying (3.3).
(i) If \(U(\omega ,t,x)=\log x \), then \(I(\omega ,t,y) = 1/y\) for all \(y \in (0,+\infty )\). Therefore (3.3) can be explicitly solved and we find
By Proposition 3.5, the solution \(c^{\mathbf{H}}=(c^{ \mathbf{H}}_{t})_{t\in [0,T]}\) is therefore given by
Under the integrability assumption in the corollary, the optimal expected utility \(u^{\mathbf{H},k}(v)\) in (3.4) can be obtained by means of a straightforward computation.
(ii) If \(U(\omega ,t,x)=x^{p}/p\), then \(I(\omega ,t,y) = y^{1/(p-1)}\) for all \(y \in (0,+\infty )\). By Proposition 3.5, the \(\mathcal{H}_{0}\)-measurable random variable \(\Lambda ^{\mathbf{H},k}(v)\) must solve
Therefore if \(\mathbb{E}[\int _{0}^{T}(Z^{\mathbf{H}}_{u})^{p/(p-1)}\, \!\mathrm{d}\kappa _{u}|\mathcal{H}_{0}]<+\infty \) a.s., it holds that
By Proposition 3.5, the optimal consumption process \(c ^{\mathbf{H}}=(c^{\mathbf{H}}_{t})_{t\in [0,T]}\) is given by \(c_{t} ^{\mathbf{H}} = (\Lambda ^{\mathbf{H},k}(v) Z^{\mathbf{H}}_{t})^{1/(p-1)} \) for all \(t\in [0,T]\). If
the optimal expected utility \(u^{\mathbf{H},k}(v)\) is finite and can be explicitly computed as in (3.5).
(iii) We first show that (3.7) admits an a.s. unique solution for every \(v>v^{\mathbf{H}}_{k}\). Arguing similarly as in Muraviev [32, Theorem 3.2], define the \(\mathcal{H}_{0}\)-measurable function \(g:\Omega \times (0,+\infty ) \rightarrow \mathbb{R}_{+}\) by
Note that \(g\) is well defined since
Clearly, \(g\) is a decreasing function. Furthermore, dominated convergence implies that \(g\) is continuous. Again by dominated convergence, we get \(\lim _{\lambda \rightarrow +\infty }g( \lambda )=0\) a.s., and a straightforward application of Fatou’s lemma yields that \(\lim _{\lambda \downarrow 0}g(\lambda )=+\infty \) a.s. Moreover, for all \(0<\lambda '<\lambda <+\infty \), it holds that \(g(\lambda ')>g( \lambda )\) a.s. on \(\{g(\lambda )>0\}\). Indeed, if the \(\mathcal{H}_{0}\)-measurable set \(G_{\lambda ,\lambda '}:=\{g( \lambda )=g(\lambda '),g(\lambda )>0\}\) has strictly positive probability, then it holds on that set that
However, since \(\log (\alpha /(\lambda 'Z^{\mathbf{H}}_{u}))>\log ( \alpha /(\lambda Z^{\mathbf{H}}_{u}))\) for all \(u\in [0,T]\), this contradicts the assumption that \(g(\lambda )>0\). In view of these observations,
for a.a. \(\omega \in \Omega \). Therefore, by Benes̆ [10, Lemma 1], (3.7) admits a unique strictly positive \(\mathcal{H}_{0}\)-measurable solution \(\Lambda ^{\mathbf{H},k}(v)\) for every \(v>v^{\mathbf{H}}_{k}\). If \(U( \omega ,t,x)=-e^{-\alpha x}\), then \(I(\omega ,t,y)=(1/\alpha )( \log (\alpha /y))^{+}\) for all \(y\in (0,+\infty )\). By Proposition 3.5, the optimal consumption process \(c^{\mathbf{H}}=(c ^{\mathbf{H}}_{t})_{t\in [0,T]}\) is given by
for all \(t\in [0,T]\), thus proving (3.6). □
Rights and permissions
About this article
Cite this article
Chau, H.N., Cosso, A. & Fontana, C. The value of informational arbitrage. Finance Stoch 24, 277–307 (2020). https://doi.org/10.1007/s00780-020-00418-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00780-020-00418-3