Portfolio management is one of the fundamental problems in financial decision making. In a typica... more Portfolio management is one of the fundamental problems in financial decision making. In a typical portfolio management problem, an investor is concerned with an optimal allocation of the capital among a number of available financial assets to maximize the return on the investment while minimizing the risk. This problem was formulated in the mean-variance portfolio management framework proposed by Markowitz in 1952. Since then, it has been widely studied by researchers and the practitioners. However, the solution is sensitive to model parameters due to data uncertainty. In this chapter, we review robust approaches to deal with data uncertainty for a single-period portfolio allocation problem. We first introduce the main ideas of robust optimization using symmetric and asymmetric uncertainty sets where the uncertain asset returns belong to. We then focus on data driven and distributionally robust optimization approaches.
International Series in Operations Research & Management Science, 2016
In this chapter, we are concerned with performance evaluation of olive oil production using Data ... more In this chapter, we are concerned with performance evaluation of olive oil production using Data Envelopment Analysis (DEA) under uncertainty. In order to measure production efficiency of olive-growing farms, we first apply an imprecise DEA approach by taking into account optimistic and pessimistic perspectives on uncertainty realized in olive oil production yield. We then consider robust optimization based DEA under an uncertainty set where the random data belong. The robust DEA model enables to adjust level of conservatism that is defined by the price of robustness of the uncertainty set. The performance of imprecise and robust DEA models is illustrated via a case study of olive-growing farms located in the Aegean Region of Turkey. The numerical experiments reveal that the efficiency scores and efficiency discriminations dramatically depend on how the uncertainty is treated both in imprecise and robust DEA modeling. There exists a trade-off between the protection level and conservatism of the efficiency scores.
This entry considers the problem of a typical pension fund that collects premiums from sponsors o... more This entry considers the problem of a typical pension fund that collects premiums from sponsors or employees and is liable for fixed payments to its customers after retirement. The fund manager’s goal is to determine an investment strategy so that the fund can cover its liabilities while minimizing contributions from its sponsors and maximizing the value of its assets. We develop robust optimization and scenario-based stochastic programming approaches for optimal asset-liability management, taking into consideration the uncertainty in asset returns and future liabilities. Our focus is on computational tractability and ease of implementation under conditions typically encountered in practice, such as asymmetries in the distributions of asset returns. Computational results from tests with real and generated data are presented to illustrate the performance of these models.
Financial decision making involves uncertainty and consequently risk. It is well known that asset... more Financial decision making involves uncertainty and consequently risk. It is well known that asset return forecasts and risk estimates are inherently inaccurate. The inaccuracy in forecasting and estimation can be addressed through the specification of rival scenarios. In this paper, we extend the multi-period mean-variance portfolio optimization and asset liability management problems to the robust worst-case design with multiple rival return and risk scenarios. A worst-case optimal strategy would yield the best decision determined simultaneously with the worst-case scenario. In risk management terms, such robust strategy would ensure that the min-max optimal performance will improve if the worst-case scenarios do not materialize.
In this paper, we are concerned with multi-agent team modelling and decision-making problems aris... more In this paper, we are concerned with multi-agent team modelling and decision-making problems arising in mission planning. Team coordination is essential for a good team management to achieve specific objectives. Team structure has an influence on team processes such as the ability to communicate or to coordinate team members ' actions properly, and consequently team performance. We consider a cooperative team decision making problem and develop a multi-agent mine search model using a centralised team structure. We present the results of our computational experiments and analyse the performance of the model in terms of an optimal or a feasible plan in a multi-agent system.
ABSTRACT In this paper, we consider interaction between spot and forward trading under demand and... more ABSTRACT In this paper, we consider interaction between spot and forward trading under demand and cost uncertainties, deriving the equilibrium of the multi-player dynamic games. The stochastic programming and worst-case analysis models based on discrete scenarios are developed to analyze the impact of demand uncertainty and risk aversion on oligopoly (forward and spot) markets’ structure in terms of the forwards and spot pricing, traded quantities and production. A real case of the Iberian electricity market is studied to illustrate performance of the models. The numerical experiments show that cost uncertainty impacts on the strategic decisions more than demand uncertainty.
Multistage stochastic programming is used to model the problem of financial portfolio management,... more Multistage stochastic programming is used to model the problem of financial portfolio management, given stochastic data provided in the form of a scenario tree. The mean or variance of total wealth at the end of the planning horizon can be optimised by solving either a linear stochastic program or a quadratic stochastic program, respectively; solution of many almost identical quadratic stochastic programs yields points describing the Markowitz efficient frontier. Computational results and backtesting are presented on a number of models, simulated and real.
Portfolio management is one of the fundamental problems in financial decision making. In a typica... more Portfolio management is one of the fundamental problems in financial decision making. In a typical portfolio management problem, an investor is concerned with an optimal allocation of the capital among a number of available financial assets to maximize the return on the investment while minimizing the risk. This problem was formulated in the mean-variance portfolio management framework proposed by Markowitz in 1952. Since then, it has been widely studied by researchers and the practitioners. However, the solution is sensitive to model parameters due to data uncertainty. In this chapter, we review robust approaches to deal with data uncertainty for a single-period portfolio allocation problem. We first introduce the main ideas of robust optimization using symmetric and asymmetric uncertainty sets where the uncertain asset returns belong to. We then focus on data driven and distributionally robust optimization approaches.
International Series in Operations Research & Management Science, 2016
In this chapter, we are concerned with performance evaluation of olive oil production using Data ... more In this chapter, we are concerned with performance evaluation of olive oil production using Data Envelopment Analysis (DEA) under uncertainty. In order to measure production efficiency of olive-growing farms, we first apply an imprecise DEA approach by taking into account optimistic and pessimistic perspectives on uncertainty realized in olive oil production yield. We then consider robust optimization based DEA under an uncertainty set where the random data belong. The robust DEA model enables to adjust level of conservatism that is defined by the price of robustness of the uncertainty set. The performance of imprecise and robust DEA models is illustrated via a case study of olive-growing farms located in the Aegean Region of Turkey. The numerical experiments reveal that the efficiency scores and efficiency discriminations dramatically depend on how the uncertainty is treated both in imprecise and robust DEA modeling. There exists a trade-off between the protection level and conservatism of the efficiency scores.
This entry considers the problem of a typical pension fund that collects premiums from sponsors o... more This entry considers the problem of a typical pension fund that collects premiums from sponsors or employees and is liable for fixed payments to its customers after retirement. The fund manager’s goal is to determine an investment strategy so that the fund can cover its liabilities while minimizing contributions from its sponsors and maximizing the value of its assets. We develop robust optimization and scenario-based stochastic programming approaches for optimal asset-liability management, taking into consideration the uncertainty in asset returns and future liabilities. Our focus is on computational tractability and ease of implementation under conditions typically encountered in practice, such as asymmetries in the distributions of asset returns. Computational results from tests with real and generated data are presented to illustrate the performance of these models.
Financial decision making involves uncertainty and consequently risk. It is well known that asset... more Financial decision making involves uncertainty and consequently risk. It is well known that asset return forecasts and risk estimates are inherently inaccurate. The inaccuracy in forecasting and estimation can be addressed through the specification of rival scenarios. In this paper, we extend the multi-period mean-variance portfolio optimization and asset liability management problems to the robust worst-case design with multiple rival return and risk scenarios. A worst-case optimal strategy would yield the best decision determined simultaneously with the worst-case scenario. In risk management terms, such robust strategy would ensure that the min-max optimal performance will improve if the worst-case scenarios do not materialize.
In this paper, we are concerned with multi-agent team modelling and decision-making problems aris... more In this paper, we are concerned with multi-agent team modelling and decision-making problems arising in mission planning. Team coordination is essential for a good team management to achieve specific objectives. Team structure has an influence on team processes such as the ability to communicate or to coordinate team members ' actions properly, and consequently team performance. We consider a cooperative team decision making problem and develop a multi-agent mine search model using a centralised team structure. We present the results of our computational experiments and analyse the performance of the model in terms of an optimal or a feasible plan in a multi-agent system.
ABSTRACT In this paper, we consider interaction between spot and forward trading under demand and... more ABSTRACT In this paper, we consider interaction between spot and forward trading under demand and cost uncertainties, deriving the equilibrium of the multi-player dynamic games. The stochastic programming and worst-case analysis models based on discrete scenarios are developed to analyze the impact of demand uncertainty and risk aversion on oligopoly (forward and spot) markets’ structure in terms of the forwards and spot pricing, traded quantities and production. A real case of the Iberian electricity market is studied to illustrate performance of the models. The numerical experiments show that cost uncertainty impacts on the strategic decisions more than demand uncertainty.
Multistage stochastic programming is used to model the problem of financial portfolio management,... more Multistage stochastic programming is used to model the problem of financial portfolio management, given stochastic data provided in the form of a scenario tree. The mean or variance of total wealth at the end of the planning horizon can be optimised by solving either a linear stochastic program or a quadratic stochastic program, respectively; solution of many almost identical quadratic stochastic programs yields points describing the Markowitz efficient frontier. Computational results and backtesting are presented on a number of models, simulated and real.
Uploads
Papers by Nalan Gulpinar