Experimental evidence on the effect of earnings targets on managers' estimates in the financial statements

This paper reports on an experiment designed to provide evidence on whether external earnings targets, such as those imposed by analysts, influence managers' judgments about (and the related accuracy of) the value of assets/liabilities reported in the financial statements. Data from the experiment indicate that higher earnings targets result in managers reporting higher estimates of profitability/asset values, but also produce larger errors in estimating those amounts. The biased estimates and related errors are a result of managers being overly optimistic about their ability to generate outcomes that fully support their estimates. In addition, data indicate that managers, over-time, learn to make better estimates, but the relation between targets, estimates, and estimation errors persists. All of this occurred in a setting in which there were financial incentives to produce the most accurate estimates possible-nothing was to be gained by deliberately biasing estimates. This suggests that the earnings targets affected managers' judgment about amounts to be reported in the financial statements, and led to sub-conscious biases that produced results causing managers' estimates to be erroneously correlated with external earnings targets.