Resumo

In this paper, we analyze the determinants of the recognition of impairment losses by Brazilian firms. The Brazilian context is particularly interesting given that most (56%) of the observations in our sample between 2010 and 2015 presented firms with two years of market-to-book ratio lower than one, an indicator of market expectations of impairment losses. Despite this low market-to-book state, we observe the recognition of impairment losses in only 57 out of 1,118 firm-year observations, a puzzling result.

Our research problem is summarized in our main hypothesis, which is whether Brazilian firms presenting a market-to-book ratio below one for an extended period of time are more likely to recognize an impairment loss, as suggested by the international literature. Our main objective is to contribute to the accounting literature on the recognition of impairment losses under IFRS, particularly in states of low market-to-book ratios, in which the market is supposed to expect impairment losses from firms.

Due to the low level of impairment disclosure practiced by Brazilian publicly traded companies, market participants are not able to directly observe estimates made by firms on the recoverable value of their assets. Nonetheless, the market value of firms could provide an indirect aggregate estimate of their net recoverable amount. The market-to-book ratio is taken as a potential impairment indicator, inspired by Ramanna and Watts' (2012) assertion that the market expects impairment losses on firms with book-to-market ratios higher than one for a period of at least two consecutive years.

We test our hypotheses through a dummy variables approach. In our main specification, the dummy variable of interest takes the value 1 when the company has a market-to-book ratio lower than one for the two calendar years, and zero otherwise, following Ramanna and Watts (2012). For robustness, we run our models and report results with one and three calendar years as a cutoff for a persistent impairment indicator. We first estimate a set of Random Effects Tobit regressions, and as a robustness test, we estimate Random Effects Logit Model.

Results found in a sample of 1,120 firm-years, comprised by 230 unique firms, are not favorable to our hypothesis that firms that present a market-to-book ratio below one for an extended period of time are more likely to recognize an impairment loss. Despite the fact that 56% of firms are in persistently low market-to-book states, there are only 57 observations in which a given firm has recognized an impairment loss, suggesting that the learning curve for the adoption of impairment rules may be slow.

Our results are supportive of the claim that Brazilian firms do not adopt conservative accounting behavior regarding the impairment of their assets. Some of our sample firms do recognize impairment losses when their market-to-book ratios remain persistently low. However, in 61% of our firm-year observations, firms do not recognize impairment losses in contradiction with market expectations. Why Brazilian firms collectively fail to recognize impairment losses in such scenario is an interesting question still to be addressed by future research.

Banker, R., Basu, S., & Byzalov, D. (2014). The Role of Multiple Impairment Indicators in Conditional Conservatism. SSRN Ramanna, K., & Watts, R. L. (2012). Evidence on the use of unverifiable estimates in required goodwill impairment. Review of Accounting Studies Riedl, E. J. (2004). An Examination of Long-Lived Asset Impairments. The Accounting Review, 79(3), 823–852. Watts, R. L., & Zuo, L. (2016). Understanding Practice and Institutions: A Historical Perspective. Accounting Horizons