Mikhail Zhitlukhin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mikhail Zhitlukhin has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Igor V. Evstigneev
Department of Economics, University of Manchester, Manchester, UK
5 shared publications - Konstantin Borovkov
School of Mathematics and Statistics The University of Melbourne 3010 Parkville Australia
3 shared publications - Albert N. Shiryaev
Russian Academy of Sciences
2 shared publications - Alexander Novikov
University of Technology Sydney
2 shared publications Alexey Muravlev
Department of Probability Theory and Mathematical Statistics, Steklov Mathematical Institute of the Russian Academy of Sciences, Moscow, Russia
2 shared publications- Alexey Muravlev
Steklov Mathematical Institute of Russian Academy of Sciences
2 shared publications
Latest Publications
- 25 September 2025
- Published by Taylor & Francis in Sequential Analysis
- Vol. 45 (1) , 140-152
Disorder detection with reversible decisions
- 25 September 2025
- Published by Taylor & Francis in Sequential Analysis
- Vol. 45 (1) , 140-152
Disorder detection with reversible decisions
- Open Access
Team problems and stochastic programming
The paper bridges two vast areas of research: stochastic team decision problems and convex stochastic programming. New methods developed in the latter are applied to the study of fundamental problems in the former. The main results are concerned with the Lagrangian relaxation of informational and material constraints in convex stochastic team problems.
Open AccessTeam problems and stochastic programming
The paper bridges two vast areas of research: stochastic team decision problems and convex stochastic programming. New methods developed in the latter are applied to the study of fundamental problems in the former. The main results are concerned with the Lagrangian relaxation of informational and material constraints in convex stochastic team problems.
- Open Access
Contest vs. Competition in Cournot Duopoly: Schaffer's Paradox
The paper compares two types of industrial organization in the Cournot duopoly: (a) the classical one, where the market players maximize profits and the outcome of the game is a Cournot-Nash equilibrium; (b) a contest in which players strive to win a fixed prize/bonus employing unbeatable strategies. Passing from (a) to (b) leads to a perfect competition with zero profits of the players (Schaffer's paradox). Transition from (b) to (a) results in a substantial decline in the production output, which also seems paradoxical, as it is commonly accepted that competition increases efficiency. We examine these phenomena in two versions of the Cournot model: with a homogeneous good and with differentiated goods.
Open AccessContest vs. Competition in Cournot Duopoly: Schaffer's Paradox
The paper compares two types of industrial organization in the Cournot duopoly: (a) the classical one, where the market players maximize profits and the outcome of the game is a Cournot-Nash equilibrium; (b) a contest in which players strive to win a fixed prize/bonus employing unbeatable strategies. Passing from (a) to (b) leads to a perfect competition with zero profits of the players (Schaffer's paradox). Transition from (b) to (a) results in a substantial decline in the production output, which also seems paradoxical, as it is commonly accepted that competition increases efficiency. We examine these phenomena in two versions of the Cournot model: with a homogeneous good and with differentiated goods.
- Open Access
On convergence of forecasts in prediction markets
We propose a dynamic model of a prediction market in which agents predict the values of a sequence of random vectors. The main result shows that if there are agents who make correct (or asymptotically correct) next-period forecasts, then the aggregated market forecasts converge to the next-period conditional expectations of the random vectors.
Open AccessOn convergence of forecasts in prediction markets
We propose a dynamic model of a prediction market in which agents predict the values of a sequence of random vectors. The main result shows that if there are agents who make correct (or asymptotically correct) next-period forecasts, then the aggregated market forecasts converge to the next-period conditional expectations of the random vectors.
- Open Access
- 12 February 2024
- Published by Springer Nature in Research Square
Behavioral Equilibrium and Evolutionary Dynamics in Asset Markets: Survival of the Fittest for Portfolio Rules
This paper analyzes a dynamic stochastic equilibrium model of an asset market based on behavioral and evolutionary principles. The core of the model is a non-traditional game-theoretic framework combining elements of stochastic dynamic games and evolutionary game theory. Its key characteristic feature is that it relies only on objectively observable market data and does not use hidden individual agents’ characteristics (such as their utilities and beliefs). A central goal of the study is to identify an investment strategy that allows an investor to survive in the market selection process, i.e., to keep with probability one a strictly positive, bounded away from zero share of market wealth over an infinite time horizon, irrespective of the strategies used by the other players. The main results show that under very general assumptions, such a strategy exists, is asymptotically unique and easily computable. Most of the related models currently considered in this field assume that asset payo¤s are exogenous and depend only on the underlying stochastic process of states of the world. The present work develops a modeling framework where the payo¤s are endogenous: they depend on the total market wealth invested in the asset. JEL Classification: C73, D53, D58, G11, G02
- 12 February 2024
- Published by Springer Nature in Research Square
Open AccessBehavioral Equilibrium and Evolutionary Dynamics in Asset Markets: Survival of the Fittest for Portfolio Rules
This paper analyzes a dynamic stochastic equilibrium model of an asset market based on behavioral and evolutionary principles. The core of the model is a non-traditional game-theoretic framework combining elements of stochastic dynamic games and evolutionary game theory. Its key characteristic feature is that it relies only on objectively observable market data and does not use hidden individual agents’ characteristics (such as their utilities and beliefs). A central goal of the study is to identify an investment strategy that allows an investor to survive in the market selection process, i.e., to keep with probability one a strictly positive, bounded away from zero share of market wealth over an infinite time horizon, irrespective of the strategies used by the other players. The main results show that under very general assumptions, such a strategy exists, is asymptotically unique and easily computable. Most of the related models currently considered in this field assume that asset payo¤s are exogenous and depend only on the underlying stochastic process of states of the world. The present work develops a modeling framework where the payo¤s are endogenous: they depend on the total market wealth invested in the asset. JEL Classification: C73, D53, D58, G11, G02
- 27 November 2023
- Published by Springer Nature in Annals of Operations Research
- p. 1-21
Survival strategies in an evolutionary finance model with endogenous asset payoffs
- 27 November 2023
- Published by Springer Nature in Annals of Operations Research
- p. 1-21
Survival strategies in an evolutionary finance model with endogenous asset payoffs
- 1 August 2023
- Published by Society for Industrial & Applied Mathematics (SIAM) in SIAM Journal on Financial Mathematics
- Vol. 14 (3) , 812-837
Capital Growth and Survival Strategies in a Market with Endogenous Prices
We consider a multiagent asset market model which includes assets of two types: long-lived assets with exogenous prices and short-lived assets with endogenous prices. The first main result consists of constructing a strategy which allows an agent to maintain a nonvanishing share of market wealth over the infinite time horizon (a “survival” strategy). The second result shows that an agent who uses this strategy becomes a single surviving agent in the market with relative wealth converging to 1 if the representative strategy of the other agents is asymptotically different from it.
- 1 August 2023
- Published by Society for Industrial & Applied Mathematics (SIAM) in SIAM Journal on Financial Mathematics
- Vol. 14 (3) , 812-837
Capital Growth and Survival Strategies in a Market with Endogenous Prices
We consider a multiagent asset market model which includes assets of two types: long-lived assets with exogenous prices and short-lived assets with endogenous prices. The first main result consists of constructing a strategy which allows an agent to maintain a nonvanishing share of market wealth over the infinite time horizon (a “survival” strategy). The second result shows that an agent who uses this strategy becomes a single surviving agent in the market with relative wealth converging to 1 if the representative strategy of the other agents is asymptotically different from it.
- Open Access
Optimal growth strategies for a representative agent in a continuous-time asset market
We propose a multi-agent model of an asset market and study conditions that guarantee that the strategy of an individual agent cannot outperform the market. The model assumes a mean-field approximation of the market by considering an infinite number of infinitesimal agents who use the same strategy and another infinitesimal agent with a different strategy who tries to outperform the market. We show that the optimal strategy for the market agents is to split their investment budgets among the assets proportionally to their discounted expected relative dividend intensities.
Open AccessOptimal growth strategies for a representative agent in a continuous-time asset market
We propose a multi-agent model of an asset market and study conditions that guarantee that the strategy of an individual agent cannot outperform the market. The model assumes a mean-field approximation of the market by considering an infinite number of infinitesimal agents who use the same strategy and another infinitesimal agent with a different strategy who tries to outperform the market. We show that the optimal strategy for the market agents is to split their investment budgets among the assets proportionally to their discounted expected relative dividend intensities.
- 27 September 2022
- Published by With Intelligence LLC in The Journal of Portfolio Management
- Vol. 49 (1) , 172-197
Using a Mean-Changing Stochastic Processes Exit–Entry Model for Stock Market Long–Short Prediction
This article has a unique application to the study of mean-changing models in stock markets. The idea is to enter and exit stock markets like Apple Computer stock (AAPL) and the broad S&P 500 Index (S&P 500) at good times and prices (long and short). Mean estimation is far more important to portfolio success than the estimation of variance or covariance. The idea in the stochastic-process model is to determine when the mean changes and then reverse the position direction. This was applied to AAPL in 2012 when its stock price rallied dramatically and then had a large fall, and to AAPL and the S&P 500 in the extremely difficult and volatile January to June 2020 period of COVID-19 and through 2021.
- 27 September 2022
- Published by With Intelligence LLC in The Journal of Portfolio Management
- Vol. 49 (1) , 172-197
Using a Mean-Changing Stochastic Processes Exit–Entry Model for Stock Market Long–Short Prediction
This article has a unique application to the study of mean-changing models in stock markets. The idea is to enter and exit stock markets like Apple Computer stock (AAPL) and the broad S&P 500 Index (S&P 500) at good times and prices (long and short). Mean estimation is far more important to portfolio success than the estimation of variance or covariance. The idea in the stochastic-process model is to determine when the mean changes and then reverse the position direction. This was applied to AAPL in 2012 when its stock price rallied dramatically and then had a large fall, and to AAPL and the S&P 500 in the extremely difficult and volatile January to June 2020 period of COVID-19 and through 2021.
- 8 June 2022
- Published by Springer Nature in Finance and Stochastics
- Vol. 26 (3) , 587-630
A continuous-time asset market game with short-lived assets
- 8 June 2022
- Published by Springer Nature in Finance and Stochastics
- Vol. 26 (3) , 587-630
A continuous-time asset market game with short-lived assets