Do buy‐side analysts inform sell‐side analyst research?

Posted on by

Abstract

This paper examines whether sell-side analysts' interactions with buy-side analysts influence the quality of sell-side research output. We hypothesise that these interactions offer the sell side a view of the buy side's private information, which enhances the quality of sell-side research. Our findings show that analyst earnings forecast accuracy improves with these interactions with diminishing returns. Results are robust to alternative proxies for research quality and information flow from buy-side to sell-side analysts. Additional tests rule out endogeneity concerns, strengthening the inference that feedback from interactions with buy-side analysts improves the quality of sell-side research output.

Bagging or boosting? Empirical evidence from financial statement fraud detection

Posted on by

Abstract

Ensemble learning, specifically bagging and boosting, has been widely used in the financial field for detecting financial fraud, but their relative performance still lacks consensus. This study compares the performance of five ensemble learning models based on bagging and boosting, using data from Chinese A-share listed companies from 2012 to 2022, including the COVID-19 pandemic period. Results show that bagging outperforms boosting in various evaluation indicators, with profitability and asset quality positively affecting financial fraud. This study reveals the mechanism by which ensemble learning affects financial fraud detection and expands related research in the financial field.

Pricing cloud stocks: Evidence from China

Posted on by

Abstract

Using factor models, we examine two pricing issues of cloud stocks in China's stock market. In particular, we test whether the Fama and French factor models are useful to explain the stock prices of cloud stocks and whether there are abnormal returns unexplained by these models. Using the daily stock prices of 1670 cloud stocks from 2012 to 2022, we find that the factor models explain up to nearly 97% of the stock return variations of the cloud stocks, and mispricing. The results are robust to alternative measure of factors, outliers, sampling period and different approaches of factor modelling.