News-based forecasts of macroeconomic indicators: A semantic path model for interpretable predictions

Abstract

The macroeconomic climate influences operations with regard to, e.g., raw material prices, financing, supply chain utilization and demand quotas. In order to adapt to the economic environment, decision-makers across the public and private sectors require accurate forecasts of the economic outlook. Existing predictive frameworks base their forecasts primarily on time series analysis, as well as the judgments of experts. As a consequence, current approaches are often biased and prone to error. In order to reduce forecast errors, this paper presents an innovative methodology that extends lag variables with unstructured data in the form of financial news: (1) we apply a variety of models from machine learning to word counts as a high-dimensional input. However, this approach suffers from low interpretability and overfitting, motivating the following remedies.(2) We follow the intuition that the economic climate is driven by general sentiments and suggest a projection of words onto latent semantic structures as a means of feature engineering. (3) We propose a semantic path model, together with estimation technique based on regularization, in order to yield full interpretability of the forecasts.We demonstrate the predictive performance of our approach by utilizing 80,813 ad hoc announcements in order to make long-term forecasts of up to 24 months ahead regarding key macroeconomic indicators. Back-testing reveals a considerable reduction in forecast errors.

RETRACTED ARTICLE: COVID-19 pandemic and unemployment rate: A hybrid unemployment rate prediction approach for developed and developing countries of Asia

“…These time series incorporate some standard features of the deviation from the conventionality and nonlinearity of the arrangement of data reliance, which is obvious from past investigations on these data sets (Nagao et al 2019 ; Katris 2019 ; Feuerriegel and Gordon 2019 ). The direct ARIMA model was utilized in the first stage to get the nonlinearity.…”

“…The previous outcomes recommend that the nonstraight models are ready for long stretch assessment skylines to hold tight to the joblessness time arrangement unevenness (Wang and Zheng 2009). All things considering, an unevenness in the assurance of the joblessness rate exists, and the end will irrefutably be tested (Feuerriegel and Gordon 2019).…”

Forecasting of the business confidence index of Ukraine on the basis of neural networks technologies

“…In paper [2], an extraordinary approach to the generation of local business confidence indices and an analysis of local intersectoral relations based on textual data from newspaper articles was proposed. S. Feuerriegela та J. Gordon [3] include business confidence index in the set of macroeconomic indicators with the aim of reflecting the general German economic climate and the forecast of macroeconomic indicators is built using the semantic path model. In the collective monograph [4], various interpretations of the business activity of the enterprise are considered and the conclusion is drawn that simultaneously with the microeconomic approach, an integrated approach to the study of this category is also being developed through an attempt to combine the business activity of individual business entities with the business activity of the economy as a whole.…”

An alternative approach to modeling the country’s business climate in conditions of limited information

“…In the research S. Feuerriegela and J. Gordon [3] the business confidence index is one of the main indicator in the semantic path modeling of German economic climate. However, the weak point of indicators remains the methodology for their assessment, including the survey.…”