Manag Sci Research Articles

Prospect and Markowitz stochastic dominance

Levy and Wiener (J Risk Uncertain 16(2), 147–163, 1998), Levy and Levy (Manage Sci 48(10), 1334–1349, 2002; Rev Fin Stud 17(4), 1015–1041, 2004) develop the prospect and Markowitz stochastic dominance theory with S-shaped and reverse S-shaped utility functions for investors. In this paper, we extend their work on prospect stochastic dominance theory (PSD) and Markowitz stochastic dominance theory (MSD) to the first three orders and link the corresponding S-shaped and reverse S-shaped utility functions to the first three orders. We also provide experiments to illustrate each case of the MSD and PSD to the first three orders and demonstrate that the higher order MSD and PSD cannot be replaced by the lower order MSD and PSD. Furthermore, we formulate the following PSD and MSD properties: hierarchy exists in both PSD and MSD relationships; arbitrage opportunities exist in the first orders of both PSD and MSD; and for any two prospects under certain conditions, their third order MSD preference will be ‘the opposite of’ or ‘the same as’ their counterpart third order PSD preference. By extending the work of Levy and Wiener and Levy and Levy, we provide investors with more tools to identify the first and third order PSD and MSD prospects and thus they could make wiser choices on their investment decision.

Eliciting conditional and unconditional rank correlations from conditional probabilities

Causes of uncertainties may be interrelated and may introduce dependencies. Ignoring these dependencies may lead to large errors. A number of graphical models in probability theory such as dependence trees, vines and (continuous) Bayesian belief nets [Cooke RM. Markov and entropy properties of tree and vine-dependent variables. In: Proceedings of the ASA section on Bayesian statistical science, 1997; Kurowicka D, Cooke RM. Distribution-free continuous Bayesian belief nets. In: Proceedings of mathematical methods in reliability conference, 2004; Bedford TJ, Cooke RM. Vines—a new graphical model for dependent random variables. Ann Stat 2002; 30(4):1031–68; Kurowicka D, Cooke RM. Uncertainty analysis with high dimensional dependence modelling. New York: Wiley; 2006; Hanea AM, et al. Hybrid methods for quantifying and analyzing Bayesian belief nets. In: Proceedings of the 2005 ENBIS5 conference, 2005; Shachter RD, Kenley CR. Gaussian influence diagrams. Manage Sci 1998; 35(5) [15].] have been developed to capture dependencies between random variables. The input for these models are various marginal distributions and dependence information, usually in the form of conditional rank correlations. Often expert elicitation is required. This paper focuses on dependence representation, and dependence elicitation. The techniques presented are illustrated with an application from aviation safety.

Maximum likelihood estimation of the double exponential jump-diffusion process

The double exponential jump-diffusion (DEJD) model, recently proposed by Kou (Manage Sci 48(8), 1086–1101, 2002) and Ramezani and Zeng (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=606361, 1998), generates a highly skewed and leptokurtic distribution and is capable of matching key features of stock and index returns. Moreover, DEJD leads to tractable pricing formulas for exotic and path dependent options (Kou and Wang Manage Sci 50(9), 1178–1192, 2004). Accordingly, the double exponential representation has gained wide acceptance. However, estimation and empirical assessment of this model has received little attention to date. The primary objective of this paper is to fill this gap. We use daily returns for the S&P-500 and the NASDAQ indexes and individual stocks, in conjunction with maximum likelihood estimation (MLE) to fit the DEJD model. We utilize the BIC criterion to assess the performance of DEJD relative to log-normally distributed jump-diffusion (LJD) and the geometric brownian motion (GBM). We find that DEJD performs better than these alternatives for both indexes and individual stocks.

Determinants of state diesel fuel excise tax rates: the political economy of fuel taxation in the United States

Ever-escalating gasoline and diesel fuel prices in the United States (US) are prompting some states to temporarily suspend the collections of gasoline, and in fewer instances, diesel fuel, taxes. The focus of this paper is on the determinants of US state excise taxes levied specifically on diesel fuel. Among other things, we find the freight trucking industry’s contribution to total state employment is a highly significant determinant of a state’s diesel tax rate, consistently suggesting that the greater this contribution, the lower the tax rate, ceteris paribus. We find little evidence that the degree of freight transportation usage on state highways impacts diesel tax rates. These two findings taken together suggest that state legislators, when determining the diesel tax rate, exhibit behavior consistent with Stigler’s (Bell J Econ Manag Sci 2: 3–21, 1971) economic or positive theory of regulation. Moreover, state law makers appear less concerned over the impact that heavy freight transportation has on highway infrastructure, which appears to be at odds with the public interest or normative theory of regulation as articulated by Posner (Bell J Econ Manag Sci 4: 335–358, 1974).

An effective modeling and solution approach for the generalized independent set problem

The generalized independent set (GIS) problem was first introduced by Hochbaum and Pathria (Forest Sci 43(4), 544–554, 1997) and independently explored in greater detail by Hochbaum (Manage Sci 50(6), 709–123, 2004). This problem, with applications in forest management and a variety of related areas, is a generalization of the classical maximum independent set problem. In this paper we highlight a natural, nonlinear formulation for the problem that is an attractive alternative to the linear model found in the literature. The effectiveness of this alternative formulation is demonstrated by computational experience on test problems of varying size and density, disclosing a dramatic reduction in the time to obtain optimal and near optimal solutions and an ability to solve much larger problems.

Preferences, the Agenda Setter, and the Distribution of Power in the EU

In this paper, we present a generalization of power indices which includes the preferences of the voters. Using a Multilinear Extension perspective (Owen in Manage Sci 18:p64–p72, 1972a) we measure the probability of the players’ voting “yes” for a particular political issue. Further, we randomize the issues and show the influence that the Agenda Setter can have on a player’s power. We demonstrate these results using data from the European Union to show how the power distribution may shift after enlargement and under the new Constitutional Treaty.

A scatter search approach for general flowshop scheduling problem

This paper proposes a new evolutionary technique called scatter search for scheduling problems of a general flow-shop. Scatter search (SS) is applied to this problem as it is able to provide a wide exploration of the search space through intensification and diversification. In addition it has a unifying principle for joining solutions which exploit the adaptive memory principle to avoid generating or incorporating duplicate solutions at various stages of the problem. This methodology provides substantially better results than the Tabu search approach of Nowicki and Smutnicki (Manage Sci 42(6):797–813, 1996) and Jain and Meeran (Comput Oper Res 29:1873–1901, 2002). The proposed framework achieves an average deviation of 14.25% from the lower bound solution of benchmark problems of Demirkol et al. (Eur J Oper Res 109(1):137–141, 1998), while the scatter search technique gives the best solutions for 32 of 40 of their benchmark problems.

Covalent binding of chloroacetamide herbicides to the active site cysteine of plant type III polyketide synthases

Chloroacetamide herbicides inhibit very-long-chain fatty acid elongase, and it has been suggested that covalent binding to the active site cysteine of the condensing enzyme is responsible [Pest Manage Sci 56 (2000), 497], but direct evidence was not available. The proposal implied that other condensing enzymes might also be targets, and therefore we have investigated four purified recombinant type III plant polyketide synthases. Chalcone synthase (CHS) revealed a high sensitivity to the chloroacetamide metazachlor, with 50% inhibition after a 10 min pre-incubation with 1–2 molecules per enzyme subunit, and the inactivation was irreversible. Stilbene synthase (STS) inactivation required 20-fold higher amounts, and 4-coumaroyltriacetic acid synthase and pyrone synthase revealed no response at the highest metazachlor concentrations tested. A similar spectrum of differential responses was detected with other herbicides that also inhibit fatty acid elongase (metolachlor and cafenstrole). The data indicate that type III polyketide synthases are potential targets of these herbicides, but each combination has to be investigated individually. The interaction of metazachlor with CHS was investigated by mass spectrometric peptide mapping, after incubation of the enzymes with the herbicides followed by tryptic digestion. A characteristic mass shift and MS/MS sequencing of the respective peptide showed that metazachlor was covalently bound to the cysteine of the active site, and the same was found with STS. This is the first direct evidence that the active site cysteine in condensing enzymes is the primary common target of these herbicides.

Cumulative sum techniques for assessing surgical results

The report by Novick and colleagues [1], in this issue of The Annals of Thoracic Surgery uses a cumulative sum (CUSUM) technique to assess the learning curve in telerobotic surgery. Novick and associates [2] have previously used CUSUMs to describe the learning curve of an academic surgeon, the change from on-pump to off-pump coronary bypass surgery [3], and the learning curve for off-pump surgery [4]. From these studies, they found that CUSUM provided “ . . . a more sensitive indicator of a cluster of surgical failures than standard statistical techniques” [1].

A batch arrival queue with a vacation time under single vacation policy

We consider an M x / G/1 queueing system with a vacation time under single vacation policy, where the server takes exactly one vacation between two successive busy periods. We derive the steady state queue size distribution at different points in times, as well as the steady state distributions of busy period and unfinished work (backlog) of this model. Scope and purpose This paper addresses issues of model building of manufacturing systems of job-shop type, where the server takes exactly one vacation after the end of each busy period. This vacation can be utilized as a post processing time after clearing the jobs in the system. To be more realistic, we further assume that the arrivals occur in batches of random size instead of single units and it covers many practical situations. For example in manufacturing systems of job-shop type, each job requires to manufacture more than one unit; in digital communication systems, messages which are transmitted could consist of a random number of packets. These manufacturing systems can be modeled by M x / G/1 queue with a single vacation policy and this extends the results of Levy and Yechiali, Manage Sci 22 (1975) 202, and Doshi, Queueing Syst 1 (1986) 29.

Predicting mutual fund performance using artificial neural networks

This study utilizes an artificial neural network (ANN) approach to predict the performance of equity mutual funds that follow value, blend and growth investment styles. Using a multi-layer perceptron model and GRG2 nonlinear optimizer, fund-specific historical operating characteristics were used to forecast mutual funds' risk-adjusted return. Results show that ANN generates better forecasting results than linear models for funds of all styles. In addition, our model outperforms that of Chiang et al. [Chiang WC, Urban TL, Baldridge GW. A neural network approach to mutual fund net asset value forecasting. Omega Int J Manage Sci 1996:24;205–215.] in predicting the performance of growth funds. We also employed a heuristic approach of variable selection via neural networks and compared it with the stepwise selection method of linear regression. Results are encouraging in that the reduced ANN models still outperform the linear models for growth and blend funds and yield similar results for value funds.