We use a laboratory-controlled environment to provide experimental evidence on the potential intended and unintended consequences of the mandatory replacement of the Incurred Credit Loss Model (ICL) of IAS 39 by the Expected Credit Loss Model (ECL) of IFRS 9. We focus on the simplified version of the ECL model using the case of allowances for trade-receivable losses as the context. In particular, we induce incentives consistent with the existing backward-looking rule-based ICL model and the proposed forward-looking ECL model. We find that eliminating the minimum “probable” threshold condition combined with allowing managers to use forward-looking information significantly improve the adequacy of credit loss allowances and their recognition in a more timely manner. This leads to higher profits over the economic cycle compared to an environment characterized by the ICL model. However, compensation incentives lead to higher levels of earnings management and dampening profits relative to the baseline case that maximizes the profits over the economic cycle. Taken together, we interpret our findings to indicate that while the replacement of the ICL model with the ECL model has the potential to mitigate the severity of the adverse effects of the ICL model, it could entail unintended consequences for earnings management. With this new ECL model still in the pre-implementation stage, our results provide ex-ante evidence to policymakers on the advantages of the ECL model as well as about the unintended outcomes provided by its inherent incentives. credit loss models, expected credit loss, incurred credit loss, credit loss allowance, earnings management, adequacy of loss allowance; pro-cyclical bias; test-bedding rule changes