Skip to content
All library documents

Why SEC Company Facts Can Show Multiple Values for One Filing

Article Quant Q&A · Author: Skyliquid
Read the original

Summary

This exchange explains why a single company metric in the SEC’s structured company facts data may appear in several rows for the same fiscal year or filing accession. The examples involve diluted earnings per share filtered to quarterly and annual reports, where the expected set of one annual and three quarterly observations is not the only possible result.

The answer recommends checking the underlying filing manually. Multiple entries can arise when a company amends a report or when a filing reports comparative figures for an earlier fiscal year as well as the current one. The response offers plausible explanations rather than diagnosing the specific rows, so analysts still need to inspect the filing context and the dataset’s fields before choosing a value.

Key ideas

  • SEC company facts can contain multiple observations for one metric and filing accession.
  • An amended annual report may create an additional entry.
  • A filing may include comparative data for an earlier fiscal year.
  • Inspect the underlying filing to determine what each reported value represents.

Tags

EquitiesUS marketsStatistics
Full text
# Help with pulling company fundamentals from EDGAR


# Help with pulling company fundamentals from EDGAR












I need help understanding why there are many different values for a specific metric reported by a company's 10-K or 10-Q in the EDGAR database.

I've downloaded the entire US equity universe in JSON from the SEC website here (Bulk Data).

Here is my code for getting the Diluted Earnings Per Share of a company in 2016

```
import json
import pandas as pd

f = open('CIK0001511198.json')
data = json.load(f)

df = pd.DataFrame(data['facts']['us-gaap']['EarningsPerShareDiluted']['units']['USD'])

df = df[(df['form'] == '10-K') | (df['form'] == '10-Q')]

df[df['fy'] == 2016]
```

I was expecting 4 rows with three 10-Q and one 10-K... some values are different even within the same form id (`accn`)

Can someone explain why there are many different values and what the differences are?

Cheers!

## Answer by Jase (score 1)

https://quant.stackexchange.com/a/66245

You should check Bamsec for one of the stocks manually to see what's going on, and what the "val" column is actually referring to. Manually compare those results w/ what you see here.

It could be something like the company amending the 10-K which leads to two entries in this dataset.

It could be that the company reports data for previous FY as well as current FY inside the same 10-K, which leads to two rows being entered into this dataset for the one report.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.