Impact of COVID-19 on SMEs in Brazil and managerial perception drivers: a novel neural model based on entropy-weighted utility functions

Abstract

Abstract Departing from the inconclusive results of the scant literature on the COVID-19 impact on Small and Medium Enterprises (SMEs), this paper proposes a novel evaluation model for addressing this issue through managerial perceptions. Over 6000 SMEs responded to twelve rounds of surveys from 2020 to 2021 during the pandemic, allowing to track the evolution over time of the perceived impact of the pandemic on small businesses. A novel entropy-weighted utility function approach is proposed here, followed by artificial neural network regression to map the variables related to the SME’s businesses that most foster the perceived utility of each business criterion during the pandemic. First, weights of business-related criteria were computed using Stepwise Weight Assessment Ratio Analysis (SWARA), sorting their relative importance - or perceptions - based on information entropy ranks derived from questionnaires collected. Transfer entropy measurements also helped in unveiling the hidden cause-effect relationships among criteria. Second, business utility functions for each criterion were computed using Complex Proportional Assessment based on SWARA weights. Third, neural network regressions were used to explain the managerial perceptions on each business criterion during the pandemic, considering each business variable. Our expected and unexpected results suggest that more resilient SMEs in Brazil are 5-10 years old and operating in the services and construction sectors. Moreover, loan success is the second most impactful criterion, deeply impacting the continuity of economic activity levels, and it is not impacted by any other business criteria. Implications for policymakers and governmental actions are highlighted.