A Text-Based Analysis of Corporate Innovation

Bellstam, Gustaf; Bhagat, Sanjai; Cookson, J. Anthony

doi:10.2139/ssrn.2803232

Cited by 27 publications

(63 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…10 Recently, academic papers have started using alternative innovation proxies. For instance, Bellstam et al (2017) compute a text-based innovation measure by analysing analyst reports and capturing the intensity with which analysts write about innovation. The advantage of this measure is that it is not influenced by firms' choices to disclose R&D expenditures or to file for patents.…”

Section: Measurement Of Innovationmentioning

confidence: 99%

Real effects of financial reporting and disclosure on innovation

Simpson

Tamayo

2020

Accounting and Business Research

View full text Add to dashboard Cite

This paper reviews the literature on the real effects of financial reporting and disclosure on corporate innovation, highlighting both the possible channels of influence and the potential challenges that researchers face when attributing causal effects. We discuss the concept of innovation, emphasising the specific characteristics that make investments in innovation difficult to report. We then provide a review of the nascent work relating disclosure to innovation, which we organise around three channels: financing, compensation and learning. Finally, we discuss recent efforts aimed at increasing the quality of corporate disclosures, including disclosures of firms' innovative activities. Throughout the paper, we highlight the trade-offs of disclosure (reduced information asymmetry and increased proprietary costs), which are particularly exacerbated in the context of corporate innovation.

show abstract

Section: Measurement Of Innovationmentioning

confidence: 99%

Real effects of financial reporting and disclosure on innovation

Simpson

Tamayo

2020

Accounting and Business Research

View full text Add to dashboard Cite

show abstract

“…For textual analysis of documents that contain a variety of topics or significant variations in information content across topics or factors, generative models such as the Latent Dirichelet Allocation (LDA) are used in recent studies. For example, Bellstam, Bhagat, and Cookson (2016) measure corporate innovation using Topic modeling; similarly, Jegadeesh and Wu (2017) quantify the economic and policy content of the Federal Reserve communications and their impact on financial markets. Still, this approach usually does not scale well when facing textual datasets with millions of words, and cannot capture complex linguistic structure within a computationally reasonable budget.…”

Section: Introductionmentioning

confidence: 99%

Textual Factors: A Scalable, Interpretable, and Data-driven Approach to Analyzing Unstructured Information

2018

View full text Add to dashboard Cite

“…Qualitatively, Bellstam et al. [] find that LDA generates topics from analyst reports that conform well to concepts used to describe innovation activities. Quantitative analyses further confirm that these topics are strongly correlated with innovation measures.…”

Section: Data and Empirical Measuresmentioning

confidence: 99%

What Are You Saying? Using topic to Detect Financial Misreporting

Brown

Crowley

Elliott

2020

J of Accounting Research

166

View full text Add to dashboard Cite

We use a machine learning technique to assess whether the thematic content of financial statement disclosures (labeled topic) is incrementally informative in predicting intentional misreporting. Using a Bayesian topic modeling algorithm, we determine and empirically quantify the topic content of a large collection of 10-K narratives spanning 1994 to 2012. We find that the algorithm produces a valid set of semantically meaningful topics that predict financial misreporting, based on samples of Securities and Exchange Commission (SEC) enforcement actions (Accounting and Auditing Enforcement Releases [AAERs]) and irregularities identified from financial restatements and 10-K filing amendments. Our out-of-sample tests indicate that topic significantly improves the detection of financial misreporting by as much as 59% when added to models based on commonly used financial and textual style variables. Furthermore, models that incorporate topic significantly outperform traditional models when detecting serious revenue recognition and core expense errors. Taken together, our results suggest that the topics discussed in annual report filings and the attention devoted to each topic are useful signals in detecting financial misreporting. JEL codes: C80; K22; K42; M40; M41; M48

show abstract

A Text-Based Analysis of Corporate Innovation

Cited by 27 publications

References 58 publications

Real effects of financial reporting and disclosure on innovation

Real effects of financial reporting and disclosure on innovation

Textual Factors: A Scalable, Interpretable, and Data-driven Approach to Analyzing Unstructured Information

What Are You Saying? Using topic to Detect Financial Misreporting

Contact Info

Product

Resources

About