INSOLVENCY OF BRAZILIAN ELECTRICITY DISTRIBUTORS: A DEA BOOTSTRAP APPROACH

Abstract

This study investigates the financial and operational indicators that explain the insolvency of Brazilian electricity distributors, using a data envelopment analysis (DEA) bootstrap approach. The Wagner and Shimshak (2007) stepwise procedure was used to select the variables that had the greatest impact on average efficiency estimated by DEA in the construction of an inefficient frontier. Through a second stage analysis, the Simar and Wilson (2007) bootstrapped truncated regression analyzed contextual variables associated with inefficiency, and consequently with firm insolvency. The sample was composed of electricity distributors, whose financial information for the 2000–2015 period was available on the Brazilian Securities Exchange (CVM) website. The results indicated that the Actual Equivalent Frequency of Power Interruptions/Regulatory Equivalent Frequency of Power Interruptions and Overall Indebtedness were the most important indicators in explaining insolvency. The second-stage analysis showed that the inefficiencies calculated using the selected indicators are positively related to insolvency criteria used by the literature, state control, dollar and geographical location, and negatively related to the domestic inflation index. The results provide valuable information for the Brazilian electricity sector’s regulatory body, which recently began to hold public hearings prior to setting up procedures for monitoring financial sustainability using financial and operational indicators.