The impact of forecasting on companies’ performance: Analysis in a multivariate setting

Over the years, practitioners and researchers have devoted their attention to forecasting techniques and methods that can be adopted to improve companies’ performance. However, forecasting techniques alone are not enough since companies should also consider several other issues associated with forecasting process management, e.g. how companies collect and use information on the market, or how the forecast is used in different decision-making processes. It is also important to understand the existence of interaction effects between these different forecasting variables, as they could determine a positive additional synergistic effect on companies’ performance. This paper aims to investigate what relevant forecasting variables should be considered to improve companies’ performance, and whether some forecasting variables can interact and influence performance with a synergistic effect. Analyses are conducted by means of data collected by the Global Manufacturing Research Group (GMRG). Data from a sample of 343 manufacturing companies in 6 different countries demonstrate that when companies intend to improve cost and delivery performances, they should devote their attention to all the different forecasting variables. In addition, the results found reveal the existence of positive interaction effects between the collection and use of information on the market and the other forecasting variables, as well as the existence of a negative interaction effect between the adoption of forecasting techniques and the use of forecasts in several decision-making processes. These results have important implications for managers as they provide guidance on how to lever on the different forecasting variables to maximize companies’ performance.     

数据挖掘方法在传统预测模型中的应用

数学建模领域中,对于解决不同类型的预测问题有一些经典的预测模型。最近日益流行的数据挖掘技术也有一些针对解决预测问题的独特的数学模型和建模方法。在此对使用传统解析式模型的预测方法与应用数据挖掘技术的数据库模型的预测方法进行了对比,重点分析了2种方法在影响因素的确定、数据规律的寻找方法以及模型侧重点选择上的不同之处,并简单讨论了如何结合2种模型的优点,将数据挖掘技术应用于传统解析式模型的建模过程之中,对已有的解析式预测模型进行改进,进而得出更加可靠准确的预测结果。     

On the Use of Spline Smoothing in Estimating Hedonic Housing Price Models: Empirical Evidence Using Hong Kong Data

Traditionally in estimating hedonic housing price functions, investigators use parametric models involving specific functional forms and a finite number of unknown parameters. Some investigators have questioned whether the underlying theory is capable of conveying sufficient information to enable a correct and successful specification of parametric models and have instead proposed the less restrictive semiparametric approach to the problem. In this paper, we illustrate how the technique of smoothing splines can be used to estimate hedonic housing price models. Smoothing splines are a powerful approach to the analysis of housing data as they are exceptionally flexible in their functional forms and provide a computationally tractable method even with a large number of explanatory variables. Our illustration takes the form of a rather limited, but very promising, application with Hong Kong data. In the forecasting comparison, the spline smoothing procedure outperforms the traditional parametric Box–Cox model in mean square error terms for out-of-sample predictions. Our results also suggest that the distortion caused by underfitting the model is smaller for spline smoothing than for the kernel and k-nearest-neighbor semiparametric procedures.     

The experience curve from the economist's perspective

This paper undertakes a critique of experience curves from several angles. It considers the extent to which they can be regarded as an extension of learning curves, and concludes that the benefits from learning-by-doing at plant level are exhausted relatively early. It goes on to consider the evidence that there is a common slope to experience curves, their usefulness for forecasting prices, and possible reasons for a spurious correlation between accumulated output and average cost. It concludes by demonstrating the differences in strategic implications between the various possible economic factors which may underlie the experience curve. The conclusion is that experience curves are partly spurious, and of little practical value in forecasting or decision making.     

Forecasting Market Size and Market Growth Rates for New Products

How many times is a forecast of a technological development correct? According to many experienced managers, it almost never is. Then what good is a forecast? A forecast helps make important innovation decisions, according to Brian C. Twiss. He argues that precision in forecasting a technological development is seldom needed for purposes of long-term planning and that any innovation so marginal that small errors in forecasting will make a big difference should not be considered anyway. Twiss suggests that technology forecasting can be of real value once it is accepted that it is essentially concerned with modeling human behavior. This is the unexpected viewpoint that Twiss presents in this article. He explains how to develop and how to use a technology forecast in long-range planning.     

A comparison of export sales forecasting practices among UK firms

Although the practice of sales forecasting is a widely researched area, only recently have empirical studies differentiated between export and domestic sales forecasting. To identify whether firms adapt their export sales forecasting activities according to the organizational and environmental context in which they are operating, this investigation develops a typology of exporters which is then used to contrast export sales forecasting practices and performance. The findings show how export forecasting activities are adapted according to organizational and environmental influences facing the firm; the impact of such influences on both perceived and actual forecast performance is also noted.     

Fashion retail forecasting by evolutionary neural networks

Recent literature on nonlinear models has shown that neural networks are versatile tools for forecasting. However, the search for an ideal network structure is a complex task. Evolutionary computation is a promising global search approach for feature and model selection. In this paper, an evolutionary computation approach is proposed in searching for the ideal network structure for a forecasting system. Two years’ apparel sales data are used in the analysis. The optimized neural networks structure for the forecasting of apparel sales is developed. The performances of the models are compared with the basic fully connected neural networks and the traditional forecasting models. We find that the proposed algorithms are useful for fashion retail forecasting, and the performance of it is better than the traditional SARIMA model for products with features of low demand uncertainty and weak seasonal trends. It is applicable for fashion retailers to produce short-term retail forecasting for apparels, which share these features.     

Forecasting organizational adoption of high-technology product innovations separated by impact: Are traditional macro-level diffusion models appropriate?

This study examines forecasting accuracy when applying macro-level diffusion models to high-tech product innovations among organizational adopters. In addition, it explores whether the accuracy of macro-level diffusion models differs according to the impact of the new product. As a benchmark for comparison, three types of basic diffusion models are compared to three simple trend extrapolation models. The role of innovation impact in explaining forecasting accuracy is also considered. These issues are addressed by empirically testing organizational adoption data for 39 new high-tech products. Results indicate that for radical innovations the Bass model is best while for incremental innovations an external influence model is best. However, simple trend extrapolation models produced the most accurate overall forecasts. The purpose of the study is to reintroduce an important topic and give practitioners better insight into forecasting the organizational adoption of high-tech products once initial sales data becomes available.     

A comparison of opinion and regression forecasting for an industrial product

Opinion-based forecasting techniques are widely used by industrial marketers. Rarely, however, are the results of these forecasts compared with alternative forecasting techniques and/or evaluated against actual operating results. In this study, opinion-based forecasting results for an industrial equipment manufacturer are evaluated against actual sales data. Further, the opinion-based predictions are compared with the predictions of a naive regression model. The results of these analyses suggest that the current opinion-based forecasting system is deficient.     

Estimating Market Growth for New Products: An Analogical Diffusion Model Approach

There has been a lot of interest in diffusion models as a basis for prelaunch estimates of the sales of new products, and indeed there have been several models developed that have achieved fairly good acceptance by new product managers. One of the limitations of such models, however, has been the requirement that a sales history for the new product, even a short one from a test market, for example, be available to derive the parameters of the model. For some types of products—consumer durables, services, industrial products, for example—a sales history isn't available. In this article, Professor Robert Thomas suggests some steps toward the development of models that incorporate the attractive features of diffusion models. His approach is to use, in a systematic way, the sales histories of products that can be considered to have analogous features from a buyer's point of view. He illustrates the approach by forecasting the sales of a new service.     

The role of the forecasting process in improving forecast accuracy and operational performance

Several operations decisions are based on proper forecast of future demand. For this reason, manufacturing companies consider forecasting a crucial process for effectively guiding several activities and research has devoted particular attention to this issue. This paper investigates the impact of how forecasting is conducted on forecast accuracy and operational performances (i.e. cost and delivery performances). Attention is here paid on three factors that characterize the forecasting process: whether structured techniques are adopted, whether information from different sources is collected to elaborate forecasts, and the extent to which forecasting is used to support decision-making processes. Analyses are conducted by means of data provided by the fourth edition of the Global Manufacturing Research Group survey. Data was collected from 343 companies belonging to several manufacturing industries from six different countries. Results show that companies adopting a structured forecasting process can improve their operational performances not simply because forecast accuracy increases. This paper highlights the importance of a proper forecasting-process design, that should be coherent with how users intend to exploit forecast results and with the aim that should be achieved, that is not necessarily improving forecast accuracy.     

The use for the competition theory of the industrial investment decisions—a case study of the Taiwan IC assembly industry

This study empirically analyzes model accuracy, and applies grey forecasting to handle non-linear problems, insufficient data resources and forecasting involving small samples, and to construct the co-opetition diffusion model for the Lotka–Volterra (L.V.) system. Furthermore, this study examines historical data comprising revenue trends in the Taiwanese IC assembly industry during the past ten years and selects from a range of forecasting models.Empirical study uses MAPE to precisely analyze revenue trends in the L.V. dynamic co-opetition diffusion model relation to the IC assembly industry. The nine companies will be selected from 4 to 11 of the modeling, the results of the LV model 64 accuracy test, its accuracy is higher than 95% accounted for 59 times, five times better than the grey prediction, showing LV competing diffusion model not only with grey prediction, and better than the traditional grey forecasting model to make a higher accuracy of the predicted value. Like grey forecasting, MAPE can promptly respond even given insufficient data. Additionally, MAPE is able to provide more accurate forecasting values than the traditional Grey forecasting model. This study demonstrates the applicability of the dynamic co-opetition theory forecasting model to the Taiwanese IC assembly industry and provides management with a reference for use in decisions aimed to increase managerial competitiveness.     

A big data driven framework for demand-driven forecasting with effects of marketing-mix variables

This study aims to investigate the contributions of promotional marketing activities, historical demand and other factors to predict, and develop a big data-driven fuzzy classifier-based framework, also called “demand-driven forecasting,” that can shape, sense and respond to real customer demands. The availability of timely information about future customer needs is a key success factor for any business. For profit maximization, manufacturers want to sense demand signals and shape future demands using price, sales, promotion and others economic factors so that they can fulfil customer's orders immediately. However, most demand forecasting systems offer limited insight to manufacturers as they fail to capture contemporary market trends, product seasonality and the impact of forecasting on the magnitude of the bullwhip effect. This paper aims to improve the accuracy of demand forecasts. In order to achieve this, a back-propagation neural network-based model is trained by fuzzy inputs and compared with benchmark forecasting methods on a time series data, by using historical demand and sales data in combination with advertising effectiveness, expenditure, promotions, and marketing events data. A statistical analysis is conducted, and the experiments show that the method used in the proposed framework outperforms in optimality, efficiency and other statistical metrics. Finally, some invaluable insights for managers are presented to improve the forecast accuracy of fuzzy neural networks, develop marketing plans for products and discuss their implications in several fields.     

Multipoint contact without forbearance? How coverage synergies shape equity analysts' forecasting performance

Research Summary: Scholars regularly use multipoint contact (MPC) to explain how encountering rivals in different domains shapes performance. While most explanations rely on mutual forbearance theory, I propose that competitive deterrence does not adequately explain how MPC shapes performance in knowledge intensive work and argue instead that cross-domain synergies may play a central role. I examine how security analysts' MPC with publicly traded firms captures synergies in their coverage portfolio, which improves forecasting accuracy and information leadership. The advantages of greater MPC for a focal analyst are counterbalanced by rivals' observational learning, which reduces the focal analyst's forecasting differentiation. A natural experiment helps corroborate my argument: rival analysts' forecasting accuracy dropped for firms in which high MPC analysts perished in the terrorist attack on September 11, 2001. Managerial Summary: Competition in the knowledge economy often unfolds across multiple domains including product markets, geographic locations, and customer segments. In these settings, an actor's level of multipoint contact (MPC) in a domain captures the knowledge and other synergies available to the focal actor, which can improve performance in the domain. In the equity research setting, an analyst's MPC on a focal firm captures the likelihood that the analyst also covers that firm's suppliers, customers and important competitors. Using data on analysts' forecasting performance between 2001 and 2013, I find that greater levels of MPC on a focal firm predicts greater forecasting accuracy and information leadership but also lowers forecasting differentiation by attracting rivals who observe and benefit from the focal analyst's knowledge.     

Optimal safety stock levels of subassemblies and manufacturing components

In order to control the time to market and manufacturing costs, companies produce and purchase many parts and components before receiving customer orders. Consequently, demand forecasting is a critical decision process. Using modular product design and super bills of materials are two effective strategies for developing a reliable demand forecasting process. They reduce the probability of stockouts in diversified production contexts. Furthermore, managing and controlling safety stocks for pre-assembled modules provide an effective solution to the problem of minimizing the effects of forecast errors. This paper develops, evaluates, and applies innovative cost-based analytical models so that the optimal safety stock of modular subassemblies and components in assembly to order and manufacturing to order systems, respectively, can be rapidly quantified. The implementation of the proposed models in two industrial case applications demonstrates that they significantly reduce the safety stock inventory levels and the global logistical cost.     

The Relationship Between Union Membership and Net Fiscal Impact

This paper develops the first evidence on how individuals’ union membership status affects their net fiscal impact, the difference between taxes they pay and cost of public benefits they receive, enriching our understanding of how labor relations interacts with public economics. Current Population Survey data between 1994 and 2015 in pooled cross‐sections and individual first‐difference models yield evidence that union membership has a positive net fiscal impact through the worker‐level channels studied.     

A Regime‐Switching Approach to Modeling Rental Prices of U.K. Real Estate Sectors

This article uses regime‐switching models of the threshold type to analyze the adjustment process of rental prices for three U.K. commercial real estate sectors over the period 1974–2008. The nonlinear models outperform their linear counterparts in in‐sample fit. Their out‐of‐sample forecasting ability is better whenever the corresponding linear models contain a significant amount of neglected nonlinearity. Regime switches are triggered when the growth rates of rental price exceed certain threshold levels. For the industrial and retail sectors such regime switches occur in situations of strong excess demand, for the office sector they occur when there is strong excess supply.     

Kurzfrist-Lastprognose für die Day-Ahead- und Intraday-Stromnachfrage bei sich systematisch ändernden Lasten

Load forecasts are used in various fields of the German energy economic to plan and to optimize the schedule of the power generation or the purchase of power from the markets based on the results of the forecasts. Therefor accurate load forecasts are necessary. But many load forecasting models reach their limits when dealing with systematic changes in the profile of the energy demand, since the model is usually calibrated by historic data so the relation between the load and the input parameters are estimated. Due to changes in the load profile the load level is moving to another level compared to the historic one. While the forecasting model is still calibrated on the old level, this can lead to higher forecasting errors and these can in turn have negative consequences on the following optimization steps. That is why a methodological approach is presented so that the forecasting model is able to adapt a systematic change in the load profile. Therefor the presented approach is at first applied to a case of application, before it is applied to two more extreme variations of the load profile to identify possible limits of the presented approach.     

Assessing the Forecasting Performance of Regime-Switching, ARIMA and GARCH Models of House Prices

While price changes on any particular home are difficult to predict, aggregate home price changes are forecastable. In this context, this paper compares the forecasting performance of three types of univariate time series models: ARIMA, GARCH and regime-switching. The underlying intuition behind regime-switching models is that the series of interest behaves differently depending on the realization of an unobservable regime variable. Regime-switching models are a compelling choice for real estate markets that have historically displayed boom and bust cycles. However, we find that, while regime-switching models can perform better in-sample, simple ARIMA models generally perform better in out-of-sample forecasting.     

Extended revenue forecasting within a service industry

Revenue forecasting is an important topic for management to track business performance and support related decision making processes (e.g. headcount or capital expenditure). It focuses on how a business recognises operating revenue, which can differ from the point at which a sales order is won. Whilst there are many publications detailing forecasting theory, in a business context these largely focus on sales order recognition alone.This paper describes the development of a revenue forecasting tool appropriate for service provision. The organisation involved in the development of the revenue forecasting tool will remain anonymous for commercial reasons but will be referred to as “Organisation A”. The targeted outcome was to extend the forecast window from one month to three months with an error rate of no more than ±10%. The tool was required to consolidate supporting data, adopt appropriate analysis/projection techniques and extend the forecast window in a specific and complex business environment.The resulting tool returned high level results that were aligned to the original targets, and was developed with three components using a combination of projection approaches appropriate to the operating environment. Whilst limited to a specific service industry as a trial, the paper provides a useful reference point for revenue forecasting in complex service businesses and provides a basis for further research opportunities for extended revenue forecasting and business analysis approaches within other service industries.