Evaluate and document usability and limitations of SEC 10K tables

[zaneselvans]

Do a review of SEC 10-K data quality / usability with someone who wasn't been working on it, and make sure that the documentation reflects the use cases that the data can address in its current state. Some issues to explore / review:

• What fraction of the companies with the electric utility sector ID in the SEC 10-K data have been successfully matched to an EIA utility?
• What fraction of the EIA utilities, electricity generation, GHG emissions, or customer base represented in the EIA data has been linked to an SEC 10-K filer? To a non-10K filing subsidiary?
• What fraction of subsidiary company records have been identified as an SEC 10-K filer?
• Subsidiary company IDs (non-10K filers) cannot be linked across time, or across different ownership slices.
• What is the impact of the missing years of data (which weren't extracted from SEC 10-K filings)?
• What is the impact of the company/block count merging issue in compiling per-company & filing information?

Next steps:

• make homework assignment for people to poke at
• recruit homework doers

[jdangerx]

@zaneselvans @katie-lamb @krivard does this feel addressed by #4562 ?

[krivard]

Mostly:

• there's one followup/enhancement I'm hoping for but it's got its own issue Update SEC 10-K source docs to clarify what low SIC coverage means #4643
• I don't think we covered the company/block count merging issue, which I'm not familiar with

[zaneselvans]

@krivard I think the company block count issue has been fixed upstream in the SEC 10-K repo. Is that right @zschira ?