What Do Analysts' Provision Forecasts Tell Us about Expected Credit Loss Recognition?

• 成果类型: Article
• 署名作者: Beatty, Anne; Liao, Scott
• 署名单位: University System of Ohio; Ohio State University; University of Toronto
• 刊物名称: ACCOUNTING REVIEW
• ISSN/ISSBN: 0001-4826
• DOI: 10.2308/TAR-2018-0049
• 发表日期: 2021
• 页码: 1-21
• 关键词: ACCURACY

摘要

We document potential cross-sectional differences in how expected loss accounting will affect provision timeliness to provide important policy insights and contribute to the literature regarding the estimation of the expected loss model adoption impact and provision timeliness determinants. Our findings that analyst provision forecasts incrementally predict future nonperforming loans (NPLs) and market returns suggest that the incurred loss provision does not incorporate all available future loss information. Higher incremental coefficients on provision forecasts for banks with greater unrecognized future losses and incurred loss constraints suggest CECL could affect cross-sectional provision timeliness differences by removing these constraints. Specifically, the provision forecast and future NPL association increases with banks' unconstrained future loss estimates reflected in loan fair value disclosures and incurred loss constraints indicated by heterogeneous loans individually reviewed for impairment. This association also increases with EPS forecast errors, but decreases with target price and NPL forecast errors.