Forecasting methods currently available assume that established patterns or relationships will not change during the post-sample forecasting phase. This, however, is not a realistic assumption for business and economic series. This paper describes a new approach to forecasting which takes into account possible pattern changes beyond the historical data. This approach is based on the development of two models: one short, the other long term. These models are then reconciled to produce the final forecasts by setting certain parameters as a function of the number, extent, and duration of pattern changes that have occurred in the past. The proposed method has been applied to the 111 series used in the M-Competition. Post-sample forecasting accuracy comparisons show the superiority of the proposed approach over the most accurate methods in the M-Competition.
This paper investigates fundamental investment strategies to detect and exploit the public's systematic errors in horse race wager markets. A handicapping model is developed and applied to win-betting in the pari-mutuel system. A multinomial logit model of the horse racing process is posited and estimated on a data base of 200 races. A recently developed procedure for exploiting the information content of rank ordered choice sets is employed to obtain more efficient parameter estimates. The variables in this discrete choice probability model include horse and jockey characteristics, plus several race-specific features. Hold-out sampling procedures are employed to evaluate wagering strategies. A wagering strategy that involves unobtrusive bets, with a side constraint eliminating long-shot betting, appears to offer the promise of positive expected returns, even in the presence of the typically large track take encountered at Thoroughbred racing events.
A decision maker needs to give his subjective probability for a single event of interest, A. Being aware that he has little substantive knowledge of the factors affecting A, the decision maker asks an expert for advice. The expert gives his subjective probability for A. In the light of this information, how should the decision maker form his subjective probability? In particular, does it matter to the decision maker whether the expert is a well-calibrated probability assessor? In this paper these questions are explored and it is suggested that concepts such as calibration and refinement cannot usefully be defined independently of the decision maker.
When designing large-scale systems, managers and engineers must often balance the desire for optimality with the need for analytic tractability. When new technologies are involved the problem may be further complicated by the need to conduct local tradeoffs among risk, cost, and time factors. In order to formally deal with these issues, this paper presents a decomposition scheme in which individual subsystems may be evaluated separately and a representative set of alternatives obtained for each. In the development, a set of multiple objectives is introduced to account for the range of organizational priorities underlying the decision making process. Pareto-optimal solutions are then found with a general purpose parametric programming algorithm and ranked with the help of the Analytic Hierarchy Process. The methodology is demonstrated with an example centering on the selection of automation options for the upcoming Space Station, but is general enough to be applicable to the design of any complex system.
Much of the current thinking about competitive strategy focuses on ways that firms can create imperfectly competitive product markets in order to obtain greater than normal economic performance. However, the economic performance of firms does not depend simply on whether or not its strategies create such markets, but also on the cost of implementing those strategies. Clearly, if the cost of strategy implementation is greater than returns obtained from creating an imperfectly competitive product market, then firms will not obtain above normal economic performance from their strategizing efforts. To help analyze the cost of implementing strategies, we introduce the concept of a strategic factor market, i.e., a market where the resources necessary to implement a strategy are acquired. If strategic factor markets are perfect, then the cost of acquiring strategic resources will approximately equal the economic value of those resources once they are used to implement product market strategies. Even if such strategies create imperfectly competitive product markets, they will not generate above normal economic performance for a firm, for their full value would have been anticipated when the resources necessary for implementation were acquired. However, strategic factor markets will be imperfectly competitive when different firms have different expectations about the future value of a strategic resource. In these settings, firms may obtain above normal economic performance from acquiring strategic resources and implementing strategies. We show that other apparent strategic factor market imperfections, including when a firm already controls all the resources needed to implement a strategy, when a firm controls unique resources, when only a small number of firms attempt to implement a strategy, and when some firms have access to lower cost capital than others, and so on, are all special cases of differences in expectations held by firms about the future value of a strategic resource. Firms can attempt to develop better expectations about the future value of strategic resources by analyzing their competitive environments or by analyzing skills and capabilities they already control. Environmental analysis cannot be expected to improve the expectations of some firms better than others, and thus cannot be a source of more accurate expectations about the future value of a strategic resource. However, analyzing a firm’s skills and capabilities can be a source of more accurate expectations. Thus, from the point of view of firms seeking greater than normal economic performance, our analysis suggests that strategic choices should flow mainly from the analysis of its unique skills and capabilities, rather than from the analysis of its competitive environment.
We present a model which enables efficient analysis of certain types of closed queueing networks with blocking due to limited buffer spaces. The networks analyzed are those in which the limited buffers occur in tandem subnetworks. A new model, with variable buffer-size, is introduced as a conceptual tool to model part of a tandem network with blocking, using only product-form submodels. Using this model we iteratively solve for the whole network. The technique is illustrated first for a simple system with tandem queues, and then for more complex systems. The method is compared with exact solutions or simulations, and found to be reasonably accurate. The method is easily implemented using standard software for closed queueing networks. Given the complexity of the blocking problem, our approach offers a simple and efficient alternative to exact analysis.
This paper considers a one-machine scheduling problem where the objective is to minimize the sum of weighted completion times subject to release dates. A polynomial time algorithm is developed for the case when the jobs are clustered. The jobs in each cluster must be processed sequentially and the clusters are ordered. The insights developed are used in an efficient heuristic for the weighted completion time problem without clusters. Also, a class of release date problems is described for which the heuristic finds an optimal solution.
A number of authors have used the portfolio standard deviation to model the risk reduction advantages of naive diversification. Other authors have pointed out that when risk is modelled by the portfolio's variance the modelling process becomes much simpler and is computationally more efficient. In this note we derive an exact parametric relationship between portfolio standard deviation and size and thus highlight the dangers of using the standard deviation in conjunction with O.L.S. regression techniques to model the risk reduction advantages of naive diversification. It is then shown that past empirical studies which have used this methodology are deficient.
The impact of forecasted demand and forecast error, introduced in the Master Production Schedule, upon Material Requirements Planning (MRP) Systems is investigated. A computerized simulation was built to examine several questions. Results indicate that forecasting error, especially the mean error, does impact MRP system inventory costs and shortages; the greater the forecast error the greater the shortages. An exception to this general relationship was that a slight forecast BIAS may improve MRP system performance, which was the case for systems studied herein. Lot-sizing rules and product structure (bill of material structure) were also found to impact total MRP system inventory costs and shortages. The more complicated the MRP structure, the greater the differentiation among lot-sizing rules and the greater the cost impact of forecast errors. A good lot-sizing rule appears to be the period order quantity rule. However, as the forecast error level gets higher, it becomes difficult to select the better lot-sizing rule. Based on this study, suggestions are presented for the production manager's consideration, especially the inventory-production control manager.
Innovation is defined as the development and implementation of new ideas by people who over time engage in transactions with others within an institutional order. This definition focuses on four basic factors (new ideas, people, transactions, and institutional context). An understanding of how these factors are related leads to four basic problems confronting most general managers: (1) a human problem of managing attention, (2) a process problem in managing new ideas into good currency, (3) a structural problem of managing part-whole relationships, and (4) a strategic problem of institutional leadership. This paper discusses these four basic problems and concludes by suggesting how they fit together into an overall framework to guide longitudinal study of the management of innovation.