American Journal of Business and Operations Research

This article provides a data analysis framework to study asymmetric macro-financial relationships in an emerging market economy with significant energy dependence. Using annual observations for Egypt over 1990–2024, we estimate a nonlinear autoregressive distributed lag (NARDL) error-correction model in which changes in Brent crude prices are algorithmically decomposed into positive and negative cumulative partial-sum series. A composite financial development index, constructed from banking-sector depth indicators, enters the model both as a direct regressor and as an interaction term with each shock component. The results show that positive oil-price shocks carry a substantially larger long-run penalty on real GDP growth than negative shocks of equal magnitude—consistent with the cost-side exposure of a net oil-importing economy. Financial deepening conditions the transmission of these shocks but does not neutralise them; the allocation of credit toward productive private-sector activity, rather than the aggregate volume of intermediation, determines the direction of the moderating effect. Rolling-window and dynamic multiplier analyses confirm structural instability in the oil–growth relationship across sub-periods, validating the nonlinear modelling approach over standard linear alternatives. Unit root tests with structural breaks, NARDL bounds tests, and a battery of diagnostic checks support the robustness of the estimated long-run relationships. The findings carry direct implications for energy-risk management, financial-sector reform, and growth-stability policy in emerging market settings.

This paper explores why there are listed companies that are valuing significantly higher in the market based on their asset base compared to other companies. It analyses the relationship between valuation premiums and profitability, asset efficiency, the combination of the two, the size of the firm and its loss status using a cross-section of the largest publicly traded companies in the world in 2024. The empirical design integrates the predictive analytics and hypothesis testing. During the explanatory phase, a strong ordinary least squares specification is used to model the logarithm of the market value divided by the total assets. In the predictive stage, logistic regression, random forest, and gradient boosting are used to identify firms in the top quartile of the valuation-premium distribution. The findings show that profitability and asset efficiency interaction is the most positive correlate of the valuation premium, and firm scale is the most negative correlate of relative valuation after standardization by assets. The interaction-enriched specification enhances explanatory power with significant material in comparison to an interaction-free model. The discriminatory performance of the tree-based models tends to be high in the classification phase, with random forest performing out of sample with an AUC of more than 0.93. The results of these studies indicate that valuation premium should be viewed as a combined operating-quality indicator and not as a reward to margin performance in isolation and can serve as a useful guide to screen a portfolio, benchmark a company and interpret market multiples.

The paper creates the business data analytics vision of how small and medium-sized enterprises (SMEs) of GCC travel and tourism ecosystems can mitigate the commercial disruption as the perceived cost, uncertainty, or inconvenience of air travel increases due to geopolitical friction in the region. Since there is a lack of public GCC micro-level booking and itinerary data, the research paper relies on a similar public dataset: the travel mode-choice dataset published under the name of statsmodels and initially based on the intercity mode-choice literature. The benchmark is operationalized as an analogue of disruption-sensitive travel demand reallocation and poses a managerial question, not a simply transport question: in the event of a shock that increases generalized cost and waiting frictions on the most exposed mode what are the most likely demand reallocations and how should SMEs respond? Empirical design transforms the data in the long-format alternative-choice form into an analytical platform that is business-facing and integrates multinomial logit, random forest, gradient boosting, and scenario stress testing into a single analytical framework. The findings indicate that the random forest model provides the best out of sample predictive performance (accuracy 0.981; macro-F1 0.973), whereas the multinomial logit model is useful in translating scenarios that can be understood. Average predicted air share decreases by 28.0 to 16.1 percent with simulated air-travel disruption, and train-like substitutes acquire most of the share. The results suggest that GCC travel, hospitality, and mobility SMEs cannot afford to trust one open channel when a period of geopolitical escalation occurs, but rather they should develop substitution-ready packages, flexible repricing guidelines, and portfolios of partnering that encompass low-friction options. The article adds a unique business analytics template of demand reallocation sensitive to crisis through the use of repeatable public information and underlines practical resilience solutions as opposed to self-forecasting wars.

Business intelligence has emerged to be a high-level managerial competency among organizations that aim to enhance the quality of planning, responsiveness in operations and evidence-based decision making in uncertain market environments. Short-term demand forecasting is one of its most important business applications since fluctuations in demand expectations affect budgeting, inventory planning, staffing, procurement timing and revenue management. The paper formulates and tests a business intelligence system of consumer demand prediction over short-term with the help of the public macroeconomic variables. It aims to show how external economic signals may be converted into an explainable, reproducible, and useful forecasting layer to be used in dashboards and decision support systems. The research forecasts next-period real consumer spending using lagged indicators based on output, disposable income, investment, unemployment, inflation, and short-term interest rates using a publicly available U.S. macroeconomic data, which is periodically updated. Ordinary least squares, ridge regression, random forest and gradient boosting are compared by using a chronological holdout design. The empirical findings indicate that the regression-based models that are interpretable have the best out-of-sample performance, and ordinary least squares model has the lowest error and greatest explanatory power. The results suggest that effective business forecasting support can be offered using transparent analytics without the need to use complex black-box models. The study is valuable because it adds to the body of business intelligence literature a reproducible external-signal prediction pipeline, a comparison of the explainable and non-explainable models in a management context, and a translation of the forecasting results into operational and strategic planning consequences.

This paper evolves a business data analytics approach to capital allocation by exploring how the use of public panel data can aid in estimating, classifying, and profiling strategic firms. The paper examines the claim that lagged market value, capital stock, and growth signals can explain the current investment behavior and hint when the investment activity is unusually high using the public-domain Grunfeld investment data, which has annual observations of major U.S. manufacturing firms. The empirical design is deliberately non-standard as compared to typical forecasting research and it consists of three analytical layers; fixed-effects panel estimation, supervised classification of high-investment periods, and firm-level strategic segmentation. The findings indicate that the growth of lagged investment, lagged capital stock and firm value is highly correlated with the present level of investment, and that machine-learning classifiers offer helpful discrimination of high in-vestment periods. Strategic segmentation exercise also indicates the clear profiles of firms that can be used to prioritize resources and track capital. The value of the paper is two-fold. First, it illustrates how an old, conventional, public data may be re-used as a new business data analytics example to support decision-making. Second, it interprets quantitative results into a managerial advice on capital planning, growth monitoring, and portfolio-style firm evaluation. Accordingly, the paper provides a reproducible submission-ready study that has a different structure than the traditional business intelligence forecasting papers and is more in line with the requirements of strategic financial analysis and data-driven capital allocation.

The strategic importance of customer retention in small and medium-sized enterprises (SMEs) is due to the fact that the resources are limited, and the indiscriminate customer acquisition and customer retention campaigns are economically inefficient. However, the descriptive reporting used by many SMEs does not have the advantages of transactiondriven analytics that allows differentiating between high-value and low-yield customer relationships. This paper creates a repli-cable customer-analytics pipeline in SME-type retail environments, using publicly available transactional data. In con-trast to the macro-level forecasting research, the paper integrates customer value segmentation with the futureoriented repeat-purchase prediction and translates the results into retention actions explicitly. The customer-level features were based on invoices, quantities, prices, product variety, and return behavior and were derived using the public Online Retail dataset. Observation windows on a monthly were transformed into a repeat-purchase 90-day problem. Three predictive models—logistic regression, random forest, and gradient boosting—were compared after customer segmentation based on recency, frequency, and monetary behavior. The findings indicate that random forest model had the highest discrimination (ROC-AUC = 0.750; PR-AUC = 0.821), followed by logistic regression, which was only slightly less than it and more interpretable. Segment analysis also showed a very concentrated revenue base with Champions having 27.5 percent of the customers but 67.2 percent of recent revenue and 81.0 rate of repeat purchasing. The paper provides a submission-ready, transparently reproducible, and managerially understandable design that is particularly applicable in SMEs that want low-cost retention analytics, customer ranking, and allocation of marketing resources.