working papers
projects I'm currently working on
-
Projecting Financial Statements with Artificial Intelligence
We introduce a novel artificial intelligence framework for projecting the entire set of financial statements. Our approach integrates multi-target learning and chained learning to predict interdependent financial statement items, capturing the intricate relationships across income statement and balance sheet components. Leveraging gradient boosting machines (GBMs) as the base learner, the framework employs a three-step process to optimise chaining sequences and expand feature sets, in order to effectively model inter-item correlations. Empirical validation using out-of-sample predictions for a large sample of U.S. public firms demonstrates the model's ability to produce accurate and internally consistent financial statement projections. Furthermore, we establish the utility of these projections for applications such as detecting financial irregularities and forecasting earnings. Our methodology can be adapted to other tasks involving the prediction of complex and interdependent outputs.
-
Leveraging AI to Decipher How Supply Contracts Map to Revenues
This paper leverages Generative Artificial Intelligence (GAI) tools to examine the role of supply contracts in revenue recognition, an important yet unexplored area in the literature. Using GAI to analyze material supply contracts disclosed in SEC filings, we first provide an overview of their common structure and specific contents. We then investigate how individual contract provisions relate to revenue recognition along three dimensions (1) the revenues expected to be recognized from the contract, (2) ex-ante uncertainty regarding the revenue to be recognized, and (3) managerial discretion in revenue reporting. Finally, we apply machine learning techniques to assess the predictive power of GAI-extracted contract information for reported revenues and revenue-related accounting issues. The results show that contract information outperforms traditional financial variables and firm characteristics in both predictive settings. Overall, our paper highlights the value of detailed information embedded in supply contracts and the advantages of using GAI in contract analysis.
-
Coordinated Innovation--The Role of Product Development Disclosures
This study explores the role of disclosures in facilitating coordinated innovation between supply chain partners. We argue that disclosures related to product development, referred to as "product disclosures", serve as a commitment device that mitigates the first-mover risk in coordination efforts, thereby facilitating coordinated innovation. We capture product disclosures from product-related press releases and measure coordinated innovation based on congruence in patent class vectors. Consistent with our hypothesis, we find a significantly positive association between product disclosures and coordinated innovation along the supply chain. This positive relation is more pronounced when coordination uncertainty is higher and information asymmetry is greater between supply chain partners. Our results are robust to two different instrumental variables approaches. We also find that coordinated innovation along the supply chain increases following a plausibly exogenous shock that enhances transparency in product development.
-
Corporate Responses to Generative AI--Early Evidence from Conference Calls
We provide early evidence of generative artificial intelligence (GAI)’s potential impact on corporations through managerial discussions of GAI in conference calls after the release of ChatGPT in November 2022. Following the release, managerial discussions of GAI in conference calls increase substantially, and the increase is more pronounced for firms with greater innovation intensity, cybersecurity threats, product differentiation, labor exposure to AI, and customer operations, suggesting that these firms are more likely to be affected by GAI. Managers tend to believe that GAI is more beneficial to firms with greater innovation intensity and cybersecurity threats but is more detrimental to firms with greater product differentiation. While they hold mixed views on GAI’s impact on firms with greater labor exposure to AI and customer operations, their views are more likely to be positive than negative. Managers of firms with greater innovation intensity, cybersecurity threats and customer operations (product differentiation) increase initiative-related (non-initiative-related) discussions more. While managers of firms with greater labor exposure to AI increase discussions of both types more, they are more likely to increase initiative-related discussions. Overall, our study sheds light on the heterogeneous corporate perceptions of GAI’s impacts and responses.