RCMARS: Robustification of CMARS with different scenarios under polyhedral uncertainty set

Our recently developed CMARS is powerful in handling complex and heterogeneous data. We include into CMARS the existence of uncertainty about the scenarios. Indeed, data include noise in both output and input variables. Therefore, solutions of the optimization problem may reveal a remarkable sensitivity to perturbations in the parameters of the problem. The data uncertainty results in uncertain constraints and objective function. To overcome this difficulty, we refine our CMARS algorithm by a robust optimization technique proposed to cope with data uncertainty. In our previous study, we present the new robust CMARS (RCMARS) in theory and method and illustrate it with a numerical example. In this study, we present RCMARS results with different uncertainty scenarios for our numerical example.     

The new robust conic GPLM method with an application to finance: prediction of credit default

This paper contributes to classification and identification in modern finance through advanced optimization. In the last few decades, financial misalignments and, thereby, financial crises have been increasing in numbers due to the rearrangement of the financial world. In this study, as one of the most remarkable of these, countries’ debt crises, which result from illiquidity, are tried to predict with some macroeconomic variables. The methodology consists of a combination of two predictive regression models, logistic regression and robust conic multivariate adaptive regression splines (RCMARS), as linear and nonlinear parts of a generalized partial linear model. RCMARS has an advantage of coping with the noise in both input and output data and of obtaining more consistent optimization results than CMARS. An advanced version of conic generalized partial linear model which includes robustification of the data set is introduced: robust conic generalized partial linear model (RCGPLM). This new model is applied on a data set that belongs to 45 emerging markets with 1,019 observations between the years 1980 and 2005.     

A new approach to multivariate adaptive regression splines by using Tikhonov regularization and continuous optimization

This paper introduces a model-based approach to the important data mining tool Multivariate adaptive regression splines (MARS), which has originally been organized in a more model-free way. Indeed, MARS denotes a modern methodology from statistical learning which is important in both classification and regression, with an increasing number of applications in many areas of science, economy and technology. It is very useful for high-dimensional problems and shows a great promise for fitting nonlinear multivariate functions. The MARS algorithm for estimating the model function consists of two algorithms, these are the forward and the backward stepwise algorithm. In our paper, we propose not to use the backward stepwise algorithm. Instead, we construct a penalized residual sum of squares for MARS as a Tikhonov regularization problem which is also known as ridge regression. We treat this problem using continuous optimization techniques which we consider to become an important complementary technology and model-based alternative to the concept of the backward stepwise algorithm. In particular, we apply the elegant framework of conic quadratic programming. This is an area of convex optimization which is very well-structured, herewith, resembling linear programming and, hence, permitting the use of powerful interior point methods. Based on these theoretical and algorithmical studies, this paper also contains an application to diabetes data. We evaluate and compare the performance of the established MARS and our new CMARS in classifying diabetic persons, where CMARS turns out to be very competitive and promising.     

Minmax robustness for multi-objective optimization problems

In real-world applications of optimization, optimal solutions are often of limited value, because disturbances of or changes to input data may diminish the quality of an optimal solution or even render it infeasible. One way to deal with uncertain input data is robust optimization, the aim of which is to find solutions which remain feasible and of good quality for all possible scenarios, i.e., realizations of the uncertain data. For single objective optimization, several definitions of robustness have been thoroughly analyzed and robust optimization methods have been developed. In this paper, we extend the concept of minmax robustness (Ben-Tal, Ghaoui, & Nemirovski, 2009) to multi-objective optimization and call this extension robust efficiency for uncertain multi-objective optimization problems. We use ingredients from robust (single objective) and (deterministic) multi-objective optimization to gain insight into the new area of robust multi-objective optimization. We analyze the new concept and discuss how robust solutions of multi-objective optimization problems may be computed. To this end, we use techniques from both robust (single objective) and (deterministic) multi-objective optimization. The new concepts are illustrated with some linear and quadratic programming instances.     

An integrated performance evaluation of financial holding companies in Taiwan

There has been a worldwide trend for financial institutions to become larger in scale and more diversified in scope, with Taiwan being no exception. Fourteen financial holding companies (FHCs) have each begun to function as a management umbrella in Taiwan by investing in different types of financial services such as banking, insurance, and securities. This paper focuses on this local financing issue from an integrated methodological perspective by model innovations proposed in several earlier studies. For example, the efficiency of profitability and marketability are combined to evaluate the FHCs’ performance. To conduct a valid and reliable evaluation process while applying the FHC’s case in Taiwan, we integrate the slacks-based measure (SBM) and slacks-based measure of super efficiency (super-SBM) models in order to directly handle the slacks and identify the best performers. A new scheme that deals with the negative output data in the SBM/super-SBM is also introduced. Inter-temporal efficiency change, which is decomposed into ‘catch-up’ and ‘frontier-shift’ effects, is analyzed by means of the SBM-based Malmquist index. A decision-making matrix is also presented to help the FHCs’ managerial authorities position themselves in the industry. The above techniques show with a high degree of consistency that large-sized FHCs perform better than small-sized ones.     

Evaluation of credit risk based on firm performance

This paper investigates whether productive inefficiency measured as the distance from the industry’s ‘best practice’ frontier is an important ex-ante predictor of business failure. We use samples of French textiles, wood and paper products, computers and R&D companies to obtain efficiency estimates for individual firms in each industry. These efficiency measures are derived from a directional technology distance function constructed empirically using non-parametric data envelopment analysis (DEA) methods. Estimating binary and ordered logit regression models we find that productive efficiency has significant explanatory power in predicting the likelihood of default over and above the effect of standard financial indicators.     

Branch & sample: A simple strategy for constraint satisfaction

Many constraint satisfaction problems have too many solutions for exhaustive generation. Optimization techniques may help in selecting a small number of solutions for consideration, but a reasonable measure of optimality is not always at hand. A simple algorithm called Branch & Sample is suggested as an alternative to optimization. Combining breadth-first and depth-first search Branch & Sample finds solutions distributed over the search tree. The aim is to obtain a limited set of solutions that corresponds well to the intuitive notion of a representative, uniformly scattered sample. A precise definition of this notion is discussed in relation to the algorithm whose effect is illustrated by two geometric design problems. The performance of the algorithm is evaluated and it is concluded that Branch & Sample is applicable to certain types of problems, and refinements can extend the scope of application.     

Pricing and hedging of financial derivatives using a posteriori error estimates and adaptive methods for stochastic differential equations

The efficient and accurate calculation of sensitivities of the price of financial derivatives with respect to perturbations of the parameters in the underlying model, the so-called ‘Greeks’, remains a great practical challenge in the derivative industry. This is true regardless of whether methods for partial differential equations or stochastic differential equations (Monte Carlo techniques) are being used. The computation of the ‘Greeks’ is essential to risk management and to the hedging of financial derivatives and typically requires substantially more computing time as compared to simply pricing the derivatives. Any numerical algorithm (Monte Carlo algorithm) for stochastic differential equations produces a time-discretization error and a statistical error in the process of pricing financial derivatives and calculating the associated ‘Greeks’. In this article we show how a posteriori error estimates and adaptive methods for stochastic differential equations can be used to control both these errors in the context of pricing and hedging of financial derivatives. In particular, we derive expansions, with leading order terms which are computable in a posteriori form, of the time-discretization errors for the price and the associated ‘Greeks’. These expansions allow the user to simultaneously first control the time-discretization errors in an adaptive fashion, when calculating the price, sensitivities and hedging parameters with respect to a large number of parameters, and then subsequently to ensure that the total errors are, with prescribed probability, within tolerance.     

60 Years of portfolio optimization: Practical challenges and current trends

The concepts of portfolio optimization and diversification have been instrumental in the development and understanding of financial markets and financial decision making. In light of the 60 year anniversary of Harry Markowitz’s paper “Portfolio Selection,” we review some of the approaches developed to address the challenges encountered when using portfolio optimization in practice, including the inclusion of transaction costs, portfolio management constraints, and the sensitivity to the estimates of expected returns and covariances. In addition, we selectively highlight some of the new trends and developments in the area such as diversification methods, risk-parity portfolios, the mixing of different sources of alpha, and practical multi-period portfolio optimization.     

A note on supply chain coordination for joint determination of order quantity and reorder point using a credit option

Credit options and side payments are two methods suggested for achieving coordination in a two-echelon supply chain. We examine the credit option coordination mechanism introduced by Chaharsooghi and Heydari [Chaharsooghi, S., & Heydari, J. (2010). Supply chain coordination for the joint determination of order quantity and reorder point using credit option. European Journal of Operational Research, 204(1), 86–95]. This method assumes that the supplier’s opportunity costs are equal to the reduction in the buyer’s financial holding costs during the credit period. In this note, we show that Chaharsooghi and Heydari’s method is not applicable when buyer and supplier opportunity costs are not equal. We introduce an alternate per order rebate method that reduces supply chain costs to centralized management levels.