Brain Language Metrics on Company Filings (History Trial)

The Brain Language Metrics on Company Filings (BLMCF) dataset has the objective of monitoring several language metrics on 10-Ks and 10-Qs company reports for approximately 6000+ US stocks.

Recent literature works claim inefficiencies in the market response to company filings information due to the increased complexity and length of such reports, see for example "Lazy Prices" Cohen et al. 2018  or "The Positive Similarity of Company Filings and the Cross-Section of Stock Returns", M. Padysak 2020 .

Some literature works claim inefficiencies in the market response to company filings information due to the increased complexity and length of such reports; over the last 20 years, the length of the average 10-K has in fact increased dramatically.

Our dataset is made of two parts; the first one includes the language metrics of the most recent 10-K or 10-Q report for each firm, namely:

• Financial sentiment

• Percentage of words belonging to financial domain classified by language types: constraining, litigious, uncertainty and interesting language.

• Readability score

• Lexical metrics such as lexical density and richness

• Text statistics such as the report length and the average sentence length

The second part includes the differences between the two most recent 10-Ks or 10-Qs reports of the same period for each company, namely:

• Difference of the various language metrics (e.g. delta sentiment, delta readability score delta, delta percentage of a specific language type etc.)

• Similarity metrics between documents, also with respect to a specific language type (for example similarity with respect to “litigious” language or “uncertainty” language)

Our dataset includes the metrics and related differences both for the whole report and for specific sections (Risk Factors and Management Discussion and Analysis)

The dataset is updated with a daily frequency since new 10-Ks and 10-Qs reports are released every day for some of the universe companies. Clearly the largest update will be around February, April, August and November when the largest number of reports is released. The historical dataset is available from year 2010.

Data Dictionary 

Factsheet 

The dataset contains historical data from January 2010 that can be freely accessed for 2 months. For a live feed please contact us at support@braincompany.co  and we will make accessible a customized version of the product on AWS Data Exchange according to Client requirements.

The content of this dataset is not to be intended as investment advice. The material is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provvaluee investment advisory or other services by Brain. Brain makes no guarantees regarding the accuracy and completeness of the information expressed in the dataset.