On the Expected Earnings Hypothesis Explanation of the Aggregate Returns–Earnings Association Puzzle

Abstract

We provide strong support for the underappreciated expected earnings hypothesis of a negative correlation between aggregate stock returns and earnings. For 1970–2000, our powerful modeling strategy incorporating macroeconomic information reveals that aggregate returns are significantly and negatively correlated with expected aggregate earnings changes but uncorrelated with unexpected aggregate earnings changes. However, this negative correlation changes after 2000, perhaps from heightened volatility or accounting changes. We also show that underlying macroeconomic information explains the power of aggregate earnings to predict future gross domestic product growth.

Aggregate accounting earnings, special items and growth in gross domestic product: evidence from Australia

“…Bailey and Lai (2020) argue that underlying macroeconomic information explains the power of aggregate earnings to predict future GDP growth. Relative to the United States, Australia is an open economy and commodity exports are an important component of GDP (Rees et al ., 2014).…”

“…Bailey and Lai (2020) argue that principal components extracted from monthly economic indicators can fully forecast future GDP growth. To examine whether our results hold after considering the effect of broader macroeconomic information, we construct two principal components from 14 macroeconomic variables.…”

“…On the other hand, it is unclear ex ante whether aggregate accounting earnings growth should necessarily be informative about future growth in GDP. An issue of particular concern is whether aggregate accounting earnings predict future GDP growth after controlling for other economic determinants that can potentially explain both corporate earnings and future growth in GDP (Bailey and Lai, 2020). Given the potential for the correlation between aggregate earnings and growth in GDP to arise from underlying economic conditions, we also consider the first hypothesis conditional on available economic information that can potentially explain both corporate earnings and future growth in GDP.…”

“…To the extent that accounting earnings are timely indicators of GDP, aggregate accounting earnings growth would be expected to be a leading indicator of growth in GDP. On the other hand, it is unclear whether aggregate accounting earnings are sufficiently timely to be informative about future GDP after controlling for other economic information associated with both aggregate earnings and growth in GDP (Bailey and Lai, 2020).…”

The Prediction of GDP by Aggregate Accounting Information. A Neural Network Model

“…Thus, researchers have been concentrating on novel approaches to enhance the precision of GDP projections. In recent times, the potential of micro-level accounting information to elucidate macroeconomic phenomena has garnered attention, initially sparked by observations that aggregate-level earnings can predict future market returns (Bailey & Lai, 2020 For instance, quarterly time series data may exhibit intercepts and trends, which can render traditional methods less suitable for forecasting. However, artificial neural network (ANN) models, which have gained prominence in recent years, can effectively avoid the limitations of linear regression (Feng & Zhang, 2014).…”