C53 - Forecasting Models; Simulation MethodsReturn

Logistics concepts are becoming less functional nowadays. The successful concept of just-in-time manufacturing seems untenable for 21st-century Europe. New ideas and new practices are emerging. One of the major trends in contemporary logistics is the demand-driven enterprise. Managing supply, fields and other processes in a company based on demand requires accurate demand forecasting at the relevant time. Relevant data are necessary for this forecasting. In the time of big data, the problem is not collecting data but evaluating them correctly. Data analytics is of great interest in business. The aim of the paper is to verify the possibilities of demand forecasting using traditional statistical methods (exponential smoothing, ARIMA), more advanced statistical methods (TBATS, etc.), methods based on artificial neural networks and a hybrid method. This is in conjunction with thorough data preparation, especially data normality testing. My research question is to provide, on data that contain a number of outliers, the results and the accuracy of the models when the data are or are not normalized. Demand forecasting for poetry books was chosen for the case study. The data were obtained from Google Trends data, i.e., searches for the topic of poetry for the period from 1 September 2013 to 31 September 2023. The results showed that the selected data contain a number of outliers that recur at regular intervals and are the result of a logical order of demand. The expected result was that data normalization increases the accuracy of the model. A method based on artificial neural networks provided significantly more accurate results. However, the resulting estimated underlying trend remained very similar. The article thus opens a discussion about the necessity of excluding outlying observations in time series where outliers exist at regular intervals.
Implications for Central European audience: The article poses fundamental scientific questions for Central Europe. Logistics concepts are developing rapidly in this area and Europe is the creator of significant innovations. Europe is currently facing great competition from foreign companies and new logistics concepts are becoming a necessity. Therefore, many European companies now feel the need to change their logistics processes. One of them may be to switch to an adaptive enterprise based on demand. The practical implications are also based on data from Europe.

The appearance of coronavirus in the spring of 2020 has significantly revalued risk exposures in the business environment, which required new approaches in the measurement of financial risks. This paper defines and explains a new approach to the measurement of liquidity risk in companies in a time of an economic crisis. This approach is more responsive to stressful circumstances in the business environment and measures the time in an economic crisis when a company can still pay maturing liabilities out of its own inventory of liquid assets, where sales of a company on the market are limited or completely prohibited, as this was the case for some industries during the first wave of Covid -19 in spring 2020. The objective of this paper was to define a new metric for the measurement of corporate liquidity, which is sensitive to this environment and shows the propensity of a company to the risk of illiquidity if an economic crisis appears. This paper, in the next step, leverages the newly defined metric of corporate liquidity and calculates the average liquidity of all companies, as well as the average liquidity by business sectors in the Serbian economy, to discuss corporate liquidity in a crisis. The results show that the average liquidity of the Serbian economy in crisis at the end of 3Q 2020 was 78,02 days, i.e., the average company in the Serbian economy was able to survive 78,02 days in stress by the end of 3Q 2020 if its sales on the market were prohibited.
Implications for Central European audience: The model in this paper is in the interest of every company and bank to measure the liquidity risk of business partners in times of an economic crisis and to predict which business partners may have liquidity problems in crisis and may therefore not be able to pay their open liabilities. The contribution of this paper to liquidity risk measurement in companies in times of stress is high as the currently available literature does not offer an alternative approach.

There are many methods of forecasting, often based on the specific conditions of the given time series which are frequently the result of research in scientific centres and universities. Nevertheless, there are also models that were created by scientists in a particular company, examples may be Google or Facebook. The latter one has developed one of the latest Prophet forecasting models published in 2017 by Taylor & Letham. This model is completely new and so it is appropriate to subject it to further research, which is the topic of this article. To accomplish this research objective, the aim of this work is to identify seasonal trends in revenue development in a selected e-commerce segment based on the assessment of the applicability of the Facebook Prophet forecasting tool. To accomplish this goal, the Python Prophet is decomposed with a subsequent two-year forecast. Accuracy of this model is measured by RMSA and coverage. The e-commerce subject selected is active primarily in the field of sales of professional outdoor supplies and organizing outdoor educational courses, seminars and competitions. It is clear from the prediction that the e-commerce entity shows a strong sales period with the beginning of the spring season and then, due to the summer, decline, until the pre-Christmas period. The subject has little growth potential and a new impetus is needed to increase sales and thus restore the growth trend. It has been confirmed that Prophet is a suitable tool for debugging seasonal tendencies.