This paper examines whether the differences in accounting information between stocks affect cross-asset return predictability. We use a comprehensive set of accounting variables and find that abnormal accruals, earnings smoothness, book-to-market, firm age, leverage, abnormal capital investment and investment growth, among others, explain the variation in return predictability across pairing stocks. Moreover, our results show that cross-asset predictability varies over time and is associated with funding liquidity and market sentiment. A simple trading strategy based on our findings yields a higher mean return, lower standard deviation and higher Sharpe ratio compared to a buy-and-hold strategy.

中文翻译：

---

会计信息和市场环境对于跨资产可预测性是否重要？

本文考察了股票之间会计信息的差异是否会影响跨资产收益的可预测性。我们使用了一套全面的会计变量，发现异常应计、收益平滑度、账面市值比、公司年龄、杠杆、异常资本投资和投资增长等，解释了配对股票之间回报可预测性的变化。此外，我们的结果表明，跨资产可预测性随时间变化，并与资金流动性和市场情绪有关。与买入并持有策略相比，基于我们发现的简单交易策略可产生更高的平均回报、更低的标准差和更高的夏普比率。