Computational Economics

Estimating limited dependent variable time series models through standard extremum methods can be a daunting computational task because of the need for integration of high order multiple integrals and/or numerical optimization of difficult objective functions. This paper proposes a classical Markov Chain Monte Carlo (MCMC) estimation technique with data augmentation that overcomes both of these problems. The asymptotic properties of the proposed estimator are discussed. Furthermore, a practical and flexible algorithmic framework for this class of models is proposed and is illustrated using simulated data, thus also offering some insight into the small-sample biases of such estimators. Finally, the proposed framework is used to estimate a dynamic, discrete-choice monetary policy reaction function for the United States during the Greenspan years.

Amidst a dynamic energy market landscape, understanding evolving influencing factors is pivotal. Accurate forecasting techniques are indispensable for effective energy resource management. This study focuses on illuminating insights into economic uncertainty and commodity price forecasting. A meticulously curated dataset spanning January 2000 to December 2022 forms the foundation, incorporating diverse economic and financial uncertainty metrics. Through an innovative research framework, we discern influential factors and forecast their trajectories. Three deep learning models—Short-Term Memory, Gated Recurrent Units, and Multilayer Perception Network—are deployed. The Multilayer Perception model emerges as the standout, showcasing exceptional predictive capability rooted in its adeptness at decoding intricate market patterns. This finding holds significance for policymakers, industry experts, and energy economists. The Multilayer Perception model’s supremacy offers a robust tool for decision-making in crafting economic policies and navigating volatile markets.

PTS is a graphical interface for production and transportation system modeling. It provides a means of creating an economic model of production and transportation activities in a system of plants and markets. The model created by PTS is translated into the GAMS language and is solved using that system. The results are then returned to PTS and are displayed graphically. PTS runs under Windows on IBM PCs and compatibles. PTS is used to (1) provide a graphical interface for knowledge-based model development systems, (2) provide a test-bed for studies of parallel model representations and (3) analyze methods of production and transportation model development which are simpler than the knowledge-based methods. The system is currently implemented for linear programming production and transportation problems. Also, the current implementation provides graphical and GAMS representations of the model.

I compare the performance of solution methods in solving a standard real business cycle model with labor market search frictions. Under the conventional calibration, the model is solved by the projection method using the Chebyshev polynomials as its basis, and the perturbation methods up to third order in both levels and logs. Evaluated by two accuracy tests, the projection approximation achieves the highest degree of accuracy, closely followed by the third order perturbation in levels. Although different in accuracy, all the approximated solutions produce simulated moments similar in value.

This paper is concerned with the numerical solution of the American option valuation problem formulated as a parabolic free boundary/initial value model. We introduce and analyze a front-tracking finite difference method and compare it with other commonly used techniques. The numerical experiments performed indicate that the front-tracking method considered is an efficient alternative for approximating simultaneously the option value and free boundary functions associated with the valuation problem.

In this paper, we propose a Griddy-Gibbs sampler approach to estimate parameters and forecast volatilities for the power transformed and threshold GARCH (PTTGARCH; Pan et al. in J Econ 142:352–378, 2008) model, which includes the standard GARCH model and many other commonly used models as special cases. Simulation study indicates that the Bayesian scheme performs effectively in estimation and prediction. A real data example is presented to support our proposed Bayesian method.

One of the main problems in economic planning is to determine the amount of resources that should be allocated with the aim of reaching some specified goals. To this aim, we attack this problem by explicitly considering development programmes over a finite time horizon. We therefore, formulate a multi-regional input-output system for reaching planning goals established in terms of production in a given time interval, simultaneously minimizing transportation and investment costs. The resulting complex nonlinear programming problem is numerically solved in terms of dynamic programming techniques. The software program, the application to the Italian economy, and the numerical results are finally presented.

The variance inflation factor is one the most applied tools for diagnosing the possible existence of multicollinearity in a multiple linear regression model. However, the VIF can detect only the relationships between independent variables without considering the intercept and is not appropriate to use with binary variables. In addition, the orthogonal model from which is calculated is also controversial. All these limitations are not usually considered when the VIF is calculated which may lead to misleading conclusions. This paper parts from an alternative orthogonal model to present a redefined variance inflation factor (RVIF) which overcomes the above limitations. This method was implemented in the rvif R package (Salmerón and García in rvif: collinearity detection using redefined variance inflation factor and graphical methods [Computer software manual]. https://cran.r-project.org/package=rvif , 2022). A Monte Carlo simulation is performed to provide threshold for this new measure. The contribution of this paper is illustrated with different examples. It is also compared with the vif command from the car R package to calculate the VIF, concluding that it could be recommendable to warn non-statisticians of its controversial use.

We leverage a new algorithm for numerically solving Colonel Blotto games to gain insight into a version of the game where players have different types of resources. Specifically, the winner of a battlefield is a function of a multi-dimensional allocation vector of each player. Our main focus is on the potential benefits of fractionation, which we define as the degree to which a player can quantize its resources. When players only have one type of resource, we show that the benefits to fractionation are in general, greatest in resource poor environments and against aggregated adversaries. We then extend the model to include random dropout and show that fractionation increases robustness to failure in resource poor environments but not resource rich environments. Finally, we show that when players have different types of resources, the benefits of fractionation are no longer mitigated by an increase in the total force size. Since many real-world resource allocation problems are multi-dimensional, our results illustrate the importance of analyzing multi-resource Blotto games in tandem with the traditional specification.

We propose a general recursive algorithm for the computation of the conditional probability function of the quadratic exponential model for binary panel data given the total of the responses, which is a sufficient statistic for the individual intercept parameter. This recursion permits to implement conditional and pseudo-conditional maximum likelihood estimators of the parameters of this model, and related models such as the dynamic logit model, even when one or more statistical units are observed at many occasions. In this way we solve a typical problem in dealing with distributions with a complex normalizing constant. The advantage in terms of computational load with respect to standard techniques is assessed by simulation and illustrated by an application based on a popular dataset about brand loyalty.