SEC EDGAR Filing Trends & Patterns from 1994 to 2022

   

On this page:

• Dataset Overview
  • Dataset Structure
  • Data Sources
  • Important Considerations
  • Analysis Workflow
• Getting Started
• Plotting Functions Used in This Tutorial
• Annual Filing Volume by Form Type
  • Amended Filings
• Monthly Filing Volume Trends
• Filing Volume per EDGAR Filer
• Filing Volume per Sector
• Filing Volume per SEC Office
• Filing Volume per SIC Codes
• IPO & Registration Statement Volumes
• Late-Filing Notification Trends
• Delisting Notification Trends - 25-NSE Filings
• Insider Trading Filing Trends
• 8-K Filing Trends

In this tutorial, we'll analyze all EDGAR filings spanning from 1994 to 2022 to uncover patterns, trends, and insights across sectors, IPOs, form types, SEC offices and more.

Key areas we'll delve into include:

• Filing Volume Dynamics: Explore how filing volumes have shifted across sectors and years, painting a picture of the economic landscape.
• IPO Trends: Using S-1 filings, we'll uncover insights into the initial public offering (IPO) patterns across sectors, revealing sectors that have seen surges in new public companies.
• M&A Activity: Mergers and acquisitions are pivotal moments in a company's journey. We'll analyze the ebbs and flows in M&A activities over the years.
• Delistings Analysis: Through 25-NSE filings, we'll examine the trends in company delistings, providing insights into sectors that might be experiencing turbulence or consolidations.
• Material Events with 8-Ks: Dive deep into 8-K filings to understand significant corporate events across sectors, from mergers and leadership changes to financial disclosures.

To bring these insights to life, we'll leverage visualization techniques such as bar charts, box plots, and heatmaps. Here's a snapshot of the key insights we're about to uncover.

Key Insights:

1. EDGAR Filing Volume: 2021 experienced the highest EDGAR filing volume ever recorded since the introduction of EDGAR in 1994, with over 800,000 filings. 2022 followed closely, registering the second-highest annual volume.

2. Dominance of Specific Form Types: Just 2% of EDGAR form types, equivalent to 20 distinct forms, accounted for 80% of all 2022 filings. Thus, understanding these 20 form types effectively unlocks 80% of information for that year.

3. Popular Form Types of 2022: The leading ten EDGAR form types in 2022 were: Form 4 (representing 29% of all filings), 8-K (11%), 424B2 (9%), followed by NPORT-P, D, 13F-HR, 6-K, FWP and 10-Q.

4. Trends in Form Types:

   • A noticeable surge was observed in 424B2 filings from 2017 onward.
   • The number of amended filings has consistently risen since 1994, hitting a record of 60,000 amended filings in 2022.
   • Form D/A, constituting 34% of amended filings, has seen a steady increase since its 2009 introduction, peaking at over 20,000 in 2022. Together, Form D/A and Form SC 13G/A comprised 64% of the top 20 amended filings in 2022.

5. Monthly Filing Patterns: February recorded the highest filing volume consistently between 2004 and 2022. In contrast, September had the lowest, followed by October and July.

6. Filing Frequency: In 2022, 63% (100,474) of all EDGAR filers (158,517) submitted between one to two filings.

7. Prominent Filers: The top five entities with the most EDGAR filings in 2022 were all banks, led by JPMORGAN CHASE, filing more than 18,000 filings that year.

8. Sectoral Analysis: The Financial Services and Healthcare sectors generated 47% of 2022's filings. The former sector's filing volume doubled since 2009, whereas the tech sector's volume declined since its 2007 peak.

9. SEC Oversight: Among the 12 SEC Offices, the Office of Finance supervised 31% of 2022 filings. The Office of Manufacturing, ranking second, managed only 10%.

10. Asset-backed Securities: ABS filings peaked in 2006, plummeted by 2009, and have been on an upward trajectory since, surpassing their pre-financial crisis peak in 2022.

11. SIC Classification Analysis: NATIONAL COMMERCIAL BANKS (SIC 6021) filings have consistently risen since 1994, with a rapid growth from 2016 and a peak in 2022, representing 20% of all filings that year. SECURITY BROKERS, DEALERS & FLOTATION COMPANIES (SIC 6211) was the second most voluminous filer in 2022, with 55,000 filings, accounting for 15% of all filings.

12. M&A and IPO Trends: M&A activity, as indicated by S-4 and S-4/A filings, peaked between 2010-2013 and has been declining since. IPOs, represented by S-1 filings, reached an all-time high in 2021, with the Industrials and Healthcare sectors leading the charge.

13. Delistings: 2021 marked the record year for delistings with 2,026 25-NSE filings, closely followed by 2022 with 1,970 cases and 2007 with 1,944.

14. Late Filing Notices: NT 10-X filings seem to act as early indicators of financial crises, as evidenced by spikes before major market downturns.

15. Sector-specific Form Trends: Only the Healthcare sector saw a consistent rise in Form 4 filings since 2012. Healthcare's annual 8-K filings have increased since 2012, while other sectors have seen a decline since their 2005 peak.

We generated two primary datasets to analyze filing volumes across dimensions like filings per form types annually from 1994 to 2022. To enhance this data, we integrated these primary datasets with two supplementary datasets, enabling us to associate a filer's CIK with additional dimensions, including sectors and SEC offices.

The two primary and two secondary datasets are structured as follows:

• filing_count_stats: contains four columns cik, formType, count, year, month. It tracks the number of filings published on EDGAR per CIK per form type in a month and year. In other words, it indicates the type and quantity of filings an entity submitted each month/year. A single filing can be tallied more than once; for instance, if a filing includes five CIKs, it gets counted five times, once for every CIK (i.e., each individual filer).
• volume_by_accNo: holds four columns formType, count, year, month. It tracks all form types, and the number of filings published per form type per month and year. It indicates the quantity of filings for each form type submitted in a given month/year. Filings are counted based on their accession numbers rather than CIKs. Thus, a filing with multiple CIKs is counted only once, rather than repeatedly for each CIK.
• filers: contains seven columns year, fiscalYearEnd, stateOfIncorporation, cik, companyName, sic, irsNo, sector. It lists all EDGAR filers that submitted at least one filing in the specified year, alongside other associated details.
• sic_codes includes three columns sic_code, office, industry_title and maps SIC codes to their respective industry titles and SEC Offices overseeing them.

The datasets, namely filing_count_stats and volume_by_accNo, were constructed using the Query API. We employed a combination of the Mapping API and Query API to create the filers dataset. The sic_codes dataset was derived from the SEC's mapping between SIC codes and their respective SEC Offices.

• EDGAR filers are identified by their CIK, while EDGAR filings are recognized by their accession number.
• All CIKs are normalized by removing leading zeros; for instance, "0000872243" is transformed to "872243".
• The CIK aids in appending SIC code, SEC Office, and sector data to the filing_count_stats dataset.
• The filedAt column signifies the Accepted timestamp of the filing as reported on the EDGAR system, and it assists in determining the associated year and month.

The methodology for preparing data for our analysis generally adheres to the steps outlined below:

1. Choose the dimension(s) of interest, such as "across sectors".
2. Group the dataset temporally (by years or months) and by the chosen dimension (like sectors).
3. Quantify a target variable, for instance, the annual number of filings per sector.
4. Construct a pivot table, setting the year as the index and the selected dimension (e.g., sector) as columns.
5. Refine the pivot table through cleaning, filtering, and sorting actions, possibly omitting certain years like 1994-1996 to eliminate outliers.
6. Visualize the data using line charts, stacked bar charts, and box plots.

The dimensions of interest include:

• Specific form types such as 10-K, 10-Q, etc.
• Form type categories, for instance, all registration statements like S-1, S-4, F-1, N-1, etc.
• Sectors
• SEC Offices
• SIC Codes
• Timeframes: years and months
• Combinations of the aforementioned, like form types analyzed by year and sector.

Let's import the relevant packages and helper functions, and load the two primary datasets: filing_count_stats and volume_by_accNo.

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.ticker as mtick
import seaborn as sns
import modules.helpers as helpers
# force reloading of modules.helpers
import importlib
importlib.reload(helpers)

plt.rcParams.update({'font.size': 8})

filing_count_stats = helpers.get_volume_by_cik()

filing_count_stats.head()

	cik	formType	count	year	month
0	100240	424B2	1	1994	01
1	100240	424B3	1	1994	01
2	100240	8-K	1	1994	01
3	100712	SC 13G	1	1994	01
4	100826	424B5	1	1994	01

volume_by_accNo = helpers.get_volume_by_accessionNo()

volume_by_accNo.sort_values(by=['year', 'month']).head()

	formType	year	month	count
1611	10-12B/A	1994	1	1
2566	10-C	1994	1	11
3011	10-K	1994	1	29
3359	10-K/A	1994	1	18
4191	10-Q	1994	1	106

---

Throughout this tutorial, we leverage two primary plotting functions to simplify our data visualization process.

plot_2_figs(pivot_table): This function simultaneously generates a line graph and a stacked bar chart within a single figure. The line graph is displayed on the left, while the stacked bar chart appears on the right.

• X-axis: Represents the time values, typically spanning from the years 1994 to 2022.
• Y-axis: Denotes the count values.
• Legend: A single, unified legend is utilized for both plots.

plot_box_plot(pivot_table): This function produces a box plot based on the provided pivot table.

Input Requirements for Both Functions: Both plotting functions expect a pivot table in the form of a dataframe. The requirements for the dataframe are:

• Index: Represents the time scale, which can be years, months, or any other time measurement.
• Columns: Correspond to any particular dimension, such as sectors.
• Values: Should be numerical counts.

def plot_2_figs(data, title, figsize=(10, 4)):
  fig, (ax1, ax2) = plt.subplots(1, 2, figsize=figsize)

  # create color map for factors
  factors = data.columns
  # viridis, plasma, magma, inferno, cividis, tab20
  colors = plt.cm.tab20(np.linspace(0, 1, len(factors)))
  colormap = dict(zip(factors, colors))

  sorted_data_1 = data.T.sort_values(2022, ascending=False).T
  sorted_data_2 = data.T.sort_values(2022, ascending=True).T

  sorted_data_1.plot(ax=ax1, color=[colormap[sector] for sector in sorted_data_1.columns])
  ax1.set_title(""), ax1.set_ylabel("Filing Volume"), ax1.set_xlabel("Year"),
  ax1.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
  ax1.legend(bbox_to_anchor=(1.05, 1), loc='upper left')
  ax1.grid(), ax1.set_axisbelow(True)
  ax1.set_xticks(range(1994, 2023)), ax1.set_xticklabels(range(1994, 2023), rotation=90, ha='center')
  ax1.legend().set_visible(False)

  sorted_data_2.plot(kind='bar', stacked=True, ax=ax2, color=[colormap[sector] for sector in sorted_data_2.columns])
  ax2.set_title(""), ax2.set_ylabel(""), ax2.set_xlabel("Year"), ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, ha='center')
  ax2.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
  handles, labels = ax2.get_legend_handles_labels()
  ax2.legend(reversed(handles), reversed(labels), bbox_to_anchor=(1.05, 1), loc='upper left')
  ax2.grid(), ax2.set_axisbelow(True)

  fig.suptitle(title, fontsize=12)
  plt.tight_layout()
  plt.show()

def plot_box_plot(data: pd.DataFrame, title, x_label="", y_label="", figsize=(5, 4)):
  data.boxplot(figsize=figsize)
  plt.title(title), plt.ylabel(y_label), plt.xlabel(x_label), plt.tight_layout()
  ax = plt.gca()
  ax.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
  max_xticklabels = max([len(label.get_text()) for label in ax.get_xticklabels()])
  rotation = 90 if max_xticklabels > 3 else 0
  ax.set_xticklabels(ax.get_xticklabels(), rotation=rotation, ha='center')
  plt.show()

Let's demonstrate the functions with some sample data.

np.random.seed(42)
years = list(range(1994, 2023))
sectors = ["Technology", "Finance", "Health", "Energy"]

sample_1 = pd.DataFrame({
    "Year": np.tile(years, len(sectors)),
    "Sector": np.repeat(sectors, len(years)),
    "Filing Volume": np.random.randint(500, 5000, len(years) * len(sectors))
})

pivot_data_sample_1 = sample_1.pivot(index='Year', columns='Sector', values='Filing Volume')

plot_2_figs(pivot_data_sample_1, "Sample Plot: Filing Volume by Sector (1994 - 2022)", figsize=(7,3))
plot_box_plot(pivot_data_sample_1, "Sample Plot", figsize=(2,2))

---

To understand the yearly distribution of filings, we'll consolidate our data to display the aggregate count of filings for each form type per year. We begin by grouping the volume_by_accNo dataframe based on two columns: formType and year. For each combination of formType and year, we sum up the values in the count column across all months of the specific year. This results in a DataFrame that shows the total number of filings for each form type in each year. The result is then stored in the type_year_stats DataFrame.

type_year_stats = pd.DataFrame(volume_by_accNo.groupby(['formType', 'year'])['count'].sum())

Next, the type_year_stats DataFrame is pivoted using the pivot_table method. The formType is set as the index, and the year values are set as the columns. This effectively reshapes the data such that each row corresponds to a specific formType and each column corresponds to a specific year. The values in the pivoted table represent the total number of filings for a given formType in a specific year.

type_year_stats_piv = type_year_stats.pivot_table(index='formType', columns='year', values='count')

Some combinations of formType and year might not have any filings, resulting in NaN (Not a Number) values in the pivoted table. This line fills those NaN values with zeros using the fillna method. Additionally, the astype(int) method is used to ensure all the numbers in the DataFrame are integers. Then, we re-arranges the columns of the type_year_stats_piv DataFrame in reverse order, starting with the year 2022 instead of 1994. The iloc method is used to select all rows and columns in reverse order (from the latest year to the earliest year). Finally, the rows of the type_year_stats_piv DataFrame are sorted based on the values in the 2022 column. This results in a DataFrame where the form types with the highest filing volumes in 2022 appear at the top, allowing for better visualization and reading the figure legend later on.

type_year_stats_piv = type_year_stats_piv.fillna(0).astype(int)
type_year_stats_piv = type_year_stats_piv.iloc[:, ::-1]
type_year_stats_piv = type_year_stats_piv.sort_values(2022, ascending=False)

print('Number of EDGAR Filings per Form Type per Year (1994 - 2022)')
type_year_stats_piv

Number of EDGAR Filings per Form Type per Year (1994 - 2022)

year	2022	2021	2020	2019	2018	2017	2016	2015	2014	2013	...	2003	2002	2001	2000	1999	1998	1997	1996	1995	1994
formType
4	182342	194456	180689	175434	180457	180800	182191	191904	193757	193172	...	160028	29298	11304	11756	10310	9467	6595	2253	0	0
8-K	68214	72895	70558	64664	66434	68772	70483	74391	74899	74153	...	65427	43121	33543	28440	26836	27072	23565	15381	6245	3484
424B2	59582	60345	44968	34295	31987	27144	18451	15494	13829	12713	...	1568	1557	1517	1153	1050	1655	1835	1630	1518	1133
NPORT-P	49843	48590	47500	8998	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
D	41672	42506	28153	27502	27272	24589	23097	23030	22194	19841	...	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
U-33-S/A	0	0	0	0	0	0	0	0	0	0	...	0	0	1	0	0	0	0	0	0	0
U-3A-2	0	0	0	0	0	0	0	0	0	0	...	86	82	76	90	99	101	86	75	72	53
U-3A-2/A	0	0	0	0	0	0	0	0	0	0	...	23	23	10	6	11	11	12	6	7	21
U-3A3-1	0	0	0	0	0	0	0	0	0	0	...	2	2	2	2	3	3	2	1	1	1
U-57	0	0	0	0	0	0	0	0	0	0	...	23	61	74	96	92	74	92	42	12	2

893 rows × 29 columns

We are now ready to plot the number of EDGAR filings by form type per year, from 1994 to 2022. We use pandas' builtin .plot function to visualize the filing volumes as stacked bar chart. Executing the plot snippet takes around 1 to 2 minutes.

type_year_stats_piv.iloc[:, ::-1].T.plot(kind='bar', stacked=True, figsize=(6, 5))
plt.title("Filing Volume by Form Type per Year (1994 - 2022)"), plt.ylabel("Filing Volume"), plt.xlabel("Year")
ax = plt.gca()
ax.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
ax.get_legend().remove(), ax.grid(), ax.set_axisbelow(True)
plt.show()

The next line sums up all EDGAR filings per year.

pd.DataFrame(type_year_stats_piv.sum(axis=0), columns=['volume']).head(10).T

year	2022	2021	2020	2019	2018	2017	2016	2015	2014	2013
volume	786768	817572	733003	660181	652410	655758	645884	666349	671145	657310

Let's identify the top 20 EDGAR form types from 2022 and then explore their filing trends since 1994.

top_20_types = list(type_year_stats_piv.sort_values(2022, ascending=False).head(20).index)
top_20_types_piv = type_year_stats_piv.loc[top_20_types]

print('Top 20 most commonly filed from types in 2022')
print(top_20_types_piv[2022][:20].index.to_list())

Top 20 most commonly filed from types in 2022
['4', '8-K', '424B2', 'NPORT-P', 'D', '13F-HR', '6-K', 'FWP', 'D/A', '10-Q', '497K', 'SC 13G/A', '3', '497', '10-D', 'CORRESP', '424B3', 'SC 13G', '10-K', 'LETTER']

plot_2_figs(top_20_types_piv.iloc[:, ::-1].T, "Filing Volume of Top 20 Most Common Form Types per Year (1994 - 2022)")

print('2022 Filing Volume of Top 20 Most Common Form Types')
print("-"*60)
size_top_20_volume = top_20_types_piv[2022].sum() / type_year_stats_piv[2022].sum()
size_top_20_form_types = top_20_types_piv[2022].count() / type_year_stats_piv[2022].count()
print("Total filing volume: {:,}".format(type_year_stats_piv[2022].sum()))
print("Top 20 form types volume: {:,}".format(top_20_types_piv[2022].sum()))
print("Volume of top 20 to total volume: {:.1%}".format(size_top_20_volume))
print("-"*60)
print("All form types in 2022: {}".format(type_year_stats_piv[2022].count()))
print("Top 20 form types to all form types: {:.1%}".format(size_top_20_form_types))

2022 Filing Volume of Top 20 Most Common Form Types
------------------------------------------------------------
Total filing volume: 786,768
Top 20 form types volume: 633,272
Volume of top 20 to total volume: 80.5%
------------------------------------------------------------
All form types in 2022: 893
Top 20 form types to all form types: 2.2%

plot_box_plot(top_20_types_piv[range(2003, 2023)].T,
              title="Top 20 Most Common Filing Types (2003 - 2022)",
              x_label="Form Type",
              y_label="Annual Filing Volume")

top_20_types_piv.T.describe().astype(int)

formType	4	8-K	424B2	NPORT-P	D	13F-HR	6-K	FWP	D/A	10-Q	497K	SC 13G/A	3	497	10-D	CORRESP	424B3	SC 13G	10-K	LETTER
count	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29	29
mean	138033	61307	13305	5342	11998	11467	16174	6879	7529	21788	8783	13568	12262	13469	3067	6504	6103	7486	7033	6140
std	89920	29024	17304	15065	13856	7423	9929	7032	8278	4730	10102	3713	7732	2263	3459	5343	1083	2311	1754	5437
min	0	3484	1050	0	0	0	0	0	0	6402	0	2499	0	5782	0	0	3321	1757	1853	0
25%	11756	33543	1557	0	0	8114	2805	0	0	19252	0	12466	3598	12133	0	0	5614	6456	6514	0
50%	182191	70483	5914	0	379	11674	20923	7333	61	21636	0	14564	14034	14080	1166	9016	6123	7469	7259	7027
75%	193757	76922	15494	0	23030	16561	22776	10589	15949	25210	19446	16130	16026	14983	6707	10823	6603	9093	8178	11273
max	234772	112225	60345	49843	42506	27500	26805	24327	20609	28178	24252	18092	24846	16402	9035	14251	8498	11329	9496	14912

top_20_types_2022 = type_year_stats_piv[2022].sort_values(ascending=False).to_frame().head(20)
top_20_types_2022['pct'] = (top_20_types_2022[2022] / top_20_types_2022[2022].sum()) * 100
top_20_types_2022['pct'] = top_20_types_2022['pct'].round(1)
top_20_types_2022

	2022	pct
formType
4	182342	28.8
8-K	68214	10.8
424B2	59582	9.4
NPORT-P	49843	7.9
D	41672	6.6
13F-HR	27500	4.3
6-K	26805	4.2
FWP	24327	3.8
D/A	20609	3.3
10-Q	20287	3.2
497K	19446	3.1
SC 13G/A	18092	2.9
3	14023	2.2
497	10731	1.7
10-D	9035	1.4
CORRESP	9019	1.4
424B3	8498	1.3
SC 13G	8439	1.3
10-K	7540	1.2
LETTER	7268	1.1

Observations:

• 2021 experienced the highest EDGAR filing volume ever recorded, with over 800,000 filings. 2022 followed closely, registering the second-highest annual volume.
• In 2022, the top 20 most common EDGAR form types accounted for 80% of all filings, even though they represented just 2% of the total 893 EDGAR form types available.
• Conversely, the remaining 98% of form types, which translates to 873 distinct forms, contributed to only 20% of the total filings in 2022.
• This stark contrast demonstrates a significant information asymmetry: a small fraction of form types dominate the vast majority of the filing volume.
• The leading ten EDGAR form types in 2022 were: 4 (29%), 8-K (11%), 424B2 (9%), followed by 'NPORT-P', 'D', '13F-HR', '6-K', 'FWP', 'D/A', and '10-Q'.
• A narrow candle in the box plot for filings like 6-K indicates a stable filing volume over the past 20 years, with minimal annual fluctuations.

Occasionally companies need to make corrections or provide updated information to previously submitted filings. These corrected or updated submissions are known as amended filings.

For instance, the Form S-1 is a document that companies must submit when they intend to go public through an Initial Public Offering (IPO). It serves as a preliminary prospectus, commonly referred to as a "red herring" due to the bold red disclaimer on its cover. This disclaimer indicates that the information within the prospectus is not final and may undergo changes. The S-1 offers potential investors a detailed view of the company's business operations, financial status, management, and both the risks and benefits of the offered securities. However, it often lacks finalized details such as the exact number of shares to be issued or the price range for the offering. During the IPO process, situations may arise where the company needs to update material information, rectify inaccuracies, or provide additional context based on feedback from investors and regulators. To address these changes, the company submits an amended registration statement, the Form S-1/A. Like its predecessor, the S-1/A might also be presented as a "red herring" with the characteristic disclaimer, signifying that the information is closer to being finalized but is still subject to change.

Another example represents 10-K filings (annual reports) that publicly traded companies must submit to the SEC. It provides an in-depth look at a company's operational and financial performance for the past fiscal year. Sometimes, during the SEC's review of these filings, the staff might identify areas of concern, seek clarifications, or request additional disclosures. This process often results in a series of correspondence between the SEC staff and the company. The content of these correspondences can often serve as triggers for companies to amend their initial filings. If the SEC staff's review highlights material discrepancies, omissions, or misrepresentations in the original 10-K, the company may be prompted (or even required) to submit an amended 10-K, known as Form 10-K/A.

Amended filings are identified by the suffix '/A' added to the original form type. So, an amended S-1 filing would appear as "S-1/A".

amended_types_piv = type_year_stats_piv[type_year_stats_piv.index.str.contains('/A')]
top_20_types_amended = list(amended_types_piv.sort_values(2022, ascending=False).head(20).index)
top_20_types_amended_piv = amended_types_piv.loc[top_20_types_amended]
top_20_types_amended_piv

year	2022	2021	2020	2019	2018	2017	2016	2015	2014	2013	...	2003	2002	2001	2000	1999	1998	1997	1996	1995	1994
formType
D/A	20609	18931	17964	18295	17422	17346	16686	15949	15794	14536	...	0	0	0	0	0	0	0	0	0	0
SC 13G/A	18092	16219	15652	15625	15807	16396	17143	16417	15588	15192	...	12589	13022	12008	12400	12757	12483	10461	6482	4318	2499
SC 13D/A	4017	3890	4141	3806	4193	4148	4271	4186	4002	4249	...	5119	5048	4899	5134	5146	5461	5418	3783	1911	807
4/A	3456	3672	3695	3748	3673	3891	4389	5026	4835	4798	...	7392	1334	489	365	247	276	123	78	0	0
S-1/A	1848	3538	1732	1198	1574	1613	1626	2124	2770	2539	...	641	820	869	2873	3061	2132	2490	2204	128	81
8-K/A	1802	1855	1719	1712	1792	2398	1850	2252	2590	2771	...	3590	2985	2713	2876	2831	3233	2824	2136	791	308
C/A	1612	1834	1503	785	718	494	173	0	0	0	...	0	0	0	0	0	0	0	0	0	0
13F-HR/A	1265	1146	974	1102	1041	1026	1022	1278	949	1296	...	1056	1075	958	768	475	0	0	0	0	0
MA-I/A	1169	1173	1101	1429	1397	2083	1366	1367	2	0	...	0	0	0	0	0	0	0	0	0	0
10-K/A	866	1081	905	983	989	914	1129	1210	1477	1726	...	1964	1944	1532	1510	1757	1876	2062	1484	897	562
10-Q/A	692	621	491	571	658	730	860	1089	1412	1817	...	1619	1710	1668	1860	2355	2066	2019	2281	1657	544
ABS-15G/A	673	657	555	505	455	419	385	339	292	248	...	0	0	0	0	0	0	0	0	0	0
NPORT-P/A	666	704	981	119	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
F-1/A	569	537	378	286	301	227	160	233	300	201	...	56	86	27	108	81	26	28	7	0	0
3/A	568	788	581	487	601	649	693	734	791	712	...	909	189	131	171	85	98	82	15	0	0
S-4/A	563	825	376	300	373	426	425	453	472	604	...	1019	867	1033	1019	1414	1891	1562	898	453	238
DRS/A	557	920	584	475	539	398	268	346	458	332	...	0	0	0	0	0	0	0	0	0	0
SC TO-I/A	528	519	588	659	647	720	699	734	664	715	...	819	782	854	280	0	0	0	0	0	0
1-A/A	462	684	606	444	384	353	347	140	103	76	...	54	100	0	0	0	0	0	0	0	0
6-K/A	301	253	255	292	273	228	265	263	264	275	...	203	120	53	28	11	6	3	2	1	0

20 rows × 29 columns

print("Annual filing volume of top 20 amended filing types:")
pd.DataFrame(top_20_types_amended_piv.sum(), columns=['volume']).T

Annual filing volume of top 20 amended filing types:

year	2022	2021	2020	2019	2018	2017	2016	2015	2014	2013	...	2003	2002	2001	2000	1999	1998	1997	1996	1995	1994
volume	60315	59847	54781	52821	52837	54459	53757	54140	52763	52087	...	37030	30082	27234	29392	30220	29548	27072	19370	10156	5039

1 rows × 29 columns

plot_2_figs(top_20_types_amended_piv.iloc[:, ::-1].T, "Filing Volume by Top 20 Most Amended Form Types per Year (1994 - 2022)")

plot_box_plot(top_20_types_amended_piv.replace(0, np.nan)[range(2000, 2023)].T,
              "Top 20 Most Common Amended Filing Types (2000 - 2022)",
              "Form Type",
              "Filing Volume")

top_20_types_per_sector_2022 = pd.DataFrame(top_20_types_amended_piv.T.loc[2022]).sort_values(by=2022, ascending=False)
top_20_types_per_sector_2022['pct'] = top_20_types_per_sector_2022[2022] / top_20_types_per_sector_2022[2022].sum() * 100
top_20_types_per_sector_2022 = top_20_types_per_sector_2022.round(1)
top_20_types_per_sector_2022[:10]

	2022	pct
formType
D/A	20609	34.2
SC 13G/A	18092	30.0
SC 13D/A	4017	6.7
4/A	3456	5.7
S-1/A	1848	3.1
8-K/A	1802	3.0
C/A	1612	2.7
13F-HR/A	1265	2.1
MA-I/A	1169	1.9
10-K/A	866	1.4

Observations:

• The number of filing amendments has consistently increased since 1994, hitting an all-time high of over 60,000 filings in 2022.
• In 2022, Form D/A accounted for 34% of these top 20 amended filings, while SC 13G/A contributed another 30%. Combined, they represented 64% of the total volume, both hitting record highs that year.
• Since its debut in 2009, Form D's annual amendments have steadily climbed, reaching over 20,000 in 2022—up from 12,000 in 2010.
• Form SC 13D/A is the third most prevalent amendment type. Its annual volume has been consistent, hovering between 4,500 and 5,000 filings yearly since 1997.
• Amended Form 4 filings experienced a notable rise from 2002 to 2004 but have consistently decreased since, reaching a historical low in 2022.
• 10-K/A filings show a consistent distribution without any outliers, suggesting steady amendment volumes over the years, even during economic downturns.
• Most filings exhibit low skewness, indicating a regular update frequency over the years.

Form D & SC 13G Excursion

• Form D informs about a private sale of securities to raise capital privately, often from a limited group of investors. It involves a simplified reporting process compared to an IPO.
• SC 13G filings are submitted by institutional investors, like mutual funds and pension funds, to disclose their ownership in a publicly traded company.

---

Next up, the visualization of monthly filing volume over the last three decades. It helps establish a baseline and observe how specific years deviate from this norm, allowing for the identification of patterns or anomalies.

Seasonality vs. Anomalies: Certain months may consistently register higher or lower filing numbers due to elements like fiscal year-ends or regulatory deadlines. Distinguishing between regular seasonal variations and unique patterns is key to spotting anomalies and discovering trends.

Variability and Predictability: The interquartile range and potential outliers offer insights into the consistency of filings for each month over the years. For instance, a month showcasing a vast range indicates its filing volume may be harder to predict.

stats_year_month = volume_by_accNo.sort_values(by=['year', 'month'], ascending=True) \
                                  .drop(columns='formType') \
                                  .groupby(['year', 'month']) \
                                  .sum() \
                                  .reset_index()

stats_year_month_piv = stats_year_month.pivot(index='year', columns='month', values='count')
stats_year_month_piv.fillna(0, inplace=True)
stats_year_month_piv = stats_year_month_piv.astype(int)

plot_2_figs(stats_year_month_piv, "Filing Volume per Month per Year (1994 - 2022)")

plot_box_plot(stats_year_month_piv.loc[2004:],
              "Monthly Filing Volume (2004 - 2022)",
              "Month",
              "Monthly Filing Volume", figsize=(3, 3))

Observations:

• Between 2004 and 2022, February consistently registered the highest monthly filing volume.
• A declining trend in volume is observed from February through the end of each year.
• September sees the sparsest filing activity, with October and July being the next quietest months.
• Notable upticks in filing volume are evident in February, May, August, and November, likely driven by the due dates for 10-Q and 10-K filings.
• The monthly filing volume displays a predictable zigzag pattern: an initial surge from January to February, a subsequent dip from February to April, and this cycle continues throughout the year.

---

Next, we'll analyze the annual submission frequency for each EDGAR filer from 1994 to 2022 using the filing_count_stats dataset. We'll categorize the 2022 data to understand the distribution, such as the count of filers who submitted 1-2 filings, 3-4 filings, and so forth. Additionally, we'll identify the top contributors for 2022 and their respective submission counts.

filing_counts_year = filing_count_stats.groupby(['year', 'cik', 'formType']).sum().reset_index()
filing_counts_year.drop(columns=['month'], inplace=True)
filing_counts_year['year'] = filing_counts_year['year'].astype(int)

print("Number of filings per form type filed by CIK 1000045 in 2022:")
filing_counts_year[(filing_counts_year['year'] == 2022) & (filing_counts_year['cik'] == '1000045')].sort_values(by='count', ascending=False)

Number of filings per form type filed by CIK 1000045 in 2022:

	year	cik	formType	count
7213867	2022	1000045	8-K	18
7213866	2022	1000045	4	13
7213864	2022	1000045	10-Q	3
7213872	2022	1000045	SC 13G	2
7213863	2022	1000045	10-K	1
7213865	2022	1000045	3	1
7213868	2022	1000045	8-K/A	1
7213869	2022	1000045	DEF 14A	1
7213870	2022	1000045	DEFR14A	1
7213871	2022	1000045	SC 13D/A	1
7213873	2022	1000045	SC 13G/A	1

filings_per_cik_year = filing_counts_year.groupby(['year', 'cik']).sum().reset_index()
filings_per_cik_year.drop(columns=['formType'], inplace=True)

print("Number of all filings filed by CIK 1000045 in 2022:")
filings_per_cik_year[(filings_per_cik_year['year'] == 2022) & (filings_per_cik_year['cik'] == '1000045')]

Number of all filings filed by CIK 1000045 in 2022:

	year	cik	count
2898221	2022	1000045	43

filings_per_cik_year_piv = filings_per_cik_year.pivot(index='cik', columns='year', values='count')
filings_per_cik_year_piv.fillna(0, inplace=True)
filings_per_cik_year_piv = filings_per_cik_year_piv.astype(int)
filings_per_cik_year_piv['mean'] = filings_per_cik_year_piv.mean(axis=1)
filings_per_cik_year_piv['median'] = filings_per_cik_year_piv.median(axis=1)

filer_2022_view = filings_per_cik_year_piv[2022]>0
filings_per_cik_2022 = filings_per_cik_year_piv[filer_2022_view][2022]

filings_per_cik_2022.hist(bins=range(0, 100, 2), figsize=(5, 4))
plt.title("Histogram of Number of Filings per CIK in 2022")
plt.ylabel("Number of Filers"), plt.xlabel("Filings filed per Year (2022)"), plt.tight_layout()
ax = plt.gca()
ax.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
plt.show()

print("Number of filers per range of filings submitted in 2022:")
pd.DataFrame(filings_per_cik_2022.value_counts(bins=range(0, 100, 2)).head(10))

Number of filers per range of filings submitted in 2022:

	count
2022
(-0.001, 2.0]	100474
(2.0, 4.0]	26442
(4.0, 6.0]	10429
(6.0, 8.0]	4451
(8.0, 10.0]	2248
(10.0, 12.0]	1412
(12.0, 14.0]	1374
(14.0, 16.0]	990
(24.0, 26.0]	762
(16.0, 18.0]	752

print("Number of filers in 2022: {:,}".format(len(filings_per_cik_2022.index)))

Number of filers in 2022: 158,517

print("Top 5 CIKs that submitted the most filings in 2022:")
filings_per_cik_2022.sort_values(ascending=False).head(5).to_frame()

Top 5 CIKs that submitted the most filings in 2022:

	2022
cik
19617	18288
1665650	17865
895421	14605
1666268	14081
831001	13831

Observations:

• Out of the 158,517 filers in 2022, 100,474 (or 63%) submitted between 1 to 2 filings.
• JPMORGAN CHASE (CIK: 19617) led the pack in 2022 with over 18,000 filings.
• The five most prolific filers of 2022 were all banking institutions: JPMORGAN CHASE (CIK: 19617), JPMorgan Chase Financial (CIK: 1665650), MORGAN STANLEY (CIK: 895421), Morgan Stanley Finance (CIK: 1666268), and CITIGROUP (CIK: 831001).

---

In this section, we'll examine the distribution of filings by sector from 1994 to 2022. To achieve this, we'll extend the filing_count_stats dataset by incorporating sector details. This involves merging the filers dataset with the filing_count_stats dataframe using the cik and year columns. Once we've constructed a pivot table with years and sectors as dimensions and yearly sector filing volume as values, we'll use our plotting functions for visualization.

filers = helpers.get_filers()
filers.head()

/Users/jan/PycharmProjects/research/edgar-patterns/modules/helpers.py:231: DtypeWarning: Columns (6) have mixed types. Specify dtype option on import or set low_memory=False.
  df = pd.read_csv(file_path, dtype={"cik": str, "fiscalYearEnd": str, "sic": str})

	year	fiscalYearEnd	stateOfIncorporation	cik	companyName	sic	irsNo	sector	industry
0	1994	1231	LA	96035	SYSTEM FUELS INC (Filer)	0000	NaN	NaN	NaN
1	1994	1130	NaN	89954	TRUST FOR SHORT TERM U S GOVERNMENT SECURITIES...	NaN	251289316	NaN	NaN
2	1994	0331	MA	901823	PUTNAM MANAGED HIGH YIELD TRUST (Filer)	0000	046733967	NaN	NaN
3	1994	1231	NY	872243	EMPIRE STATE MUNICIPAL EXEMPT TRUST GUARANTEED...	0000	NaN	NaN	NaN
4	1994	1231	NY	872246	EMPIRE STATE MUNICIPAL EXEMPT TRUST GUARANTEED...	0000	NaN	NaN	NaN

filing_count_stats = filing_count_stats.merge(filers[['cik', 'year', 'sector']], how='left', on=['cik', 'year'])
filing_count_stats.head(100)

	cik	formType	count	year	month	sector
0	100240	424B2	1	1994	01	NaN
1	100240	424B3	1	1994	01	NaN
2	100240	8-K	1	1994	01	NaN
3	100712	SC 13G	1	1994	01	NaN
4	100826	424B5	1	1994	01	NaN
...	...	...	...	...	...	...
95	14707	8-K	1	1994	01	Consumer Cyclical
96	14827	DEF 14A	1	1994	01	Industrials
97	14827	S-3	1	1994	01	Industrials
98	14846	DEF 14A	1	1994	01	Real Estate
99	15615	10-K/A	1	1994	01	Industrials

100 rows × 6 columns

# add SIC code, sector, industry columns
filings_per_cik_year_meta = filings_per_cik_year.merge(filers[['year', 'cik', 'sic', 'sector', 'industry']], on=['year','cik'], how='left')
# split sic column into two columns: sic_code, sic_desc by splitting on the first space
filings_per_cik_year_meta[['sic_code', 'sic_desc']] = filings_per_cik_year_meta['sic'].str.split(' ', n=1, expand=True)
filings_per_cik_year_meta['sic_code'] = filings_per_cik_year_meta['sic_code'].str.lstrip('0')
filings_per_cik_year_meta.drop(columns=['sic'], inplace=True)
filings_per_cik_year_meta.head()

	year	cik	count	sector	industry	sic_code	sic_desc
0	1994	100030	3	Consumer Cyclical	Auto Parts	3714	Motor Vehicle Parts & Accessories
1	1994	100063	2	Technology	Scientific & Technical Instruments	3823	Industrial Instruments For Measurement, Displa...
2	1994	100122	2	NaN	NaN	4911	Electric Services
3	1994	100166	9	Consumer Cyclical	Textile Manufacturing	2253	Knit Outerwear Mills
4	1994	100240	26	NaN	NaN	4833	Television Broadcasting Stations

print(len(filings_per_cik_year), len(filings_per_cik_year_meta))

3056738 3056738

filings_per_sector_year = filings_per_cik_year_meta[['year', 'sector', 'count']].groupby(['year', 'sector']).sum().reset_index()
filings_per_sector_year

	year	sector	count
0	1994	Basic Materials	1730
1	1994	Communication Services	1126
2	1994	Consumer Cyclical	2636
3	1994	Consumer Defensive	1522
4	1994	Energy	934
...	...	...	...
314	2022	Healthcare	84840
315	2022	Industrials	58828
316	2022	Real Estate	17435
317	2022	Technology	59326
318	2022	Utilities	9016

319 rows × 3 columns

filings_per_sector_year_piv = filings_per_sector_year.pivot(index='year', columns='sector', values='count')
filings_per_sector_year_piv[-10:]

sector	Basic Materials	Communication Services	Consumer Cyclical	Consumer Defensive	Energy	Financial Services	Healthcare	Industrials	Real Estate	Technology	Utilities
year
2013	24443	20280	40740	18108	27374	96301	56197	46422	20627	61212	10607
2014	24141	19455	41360	18122	29245	100807	63226	45093	20981	59384	10876
2015	22828	18488	40939	17593	26724	98417	66208	45183	20496	59703	11151
2016	21734	18409	38394	16492	25410	93656	64004	43274	20375	55140	11188
2017	22547	17892	38547	15862	25282	101651	65482	43075	20497	52350	9987
2018	20605	18446	37215	15628	23776	103849	67769	45750	19560	50485	10137
2019	19849	17170	36315	14755	21189	107134	71650	43133	18561	50539	9137
2020	20813	19056	38889	16123	19287	125586	83129	45145	19054	52912	8865
2021	20941	20358	43454	17470	18562	144650	91429	64791	20526	65057	9562
2022	20413	17756	38184	17351	17138	148421	84840	58828	17435	59326	9016

plot_2_figs(filings_per_sector_year_piv, "Filing Volume per Sector per Year (1994 - 2022)")

plot_box_plot(filings_per_sector_year_piv.loc[1996:],
              "Annual Filing Volume per Sector (1996 - 2022)", "", "Annual Filing Volume", figsize=(4, 3))

filings_per_sector_2022 = pd.DataFrame(filings_per_sector_year_piv.loc[2022]).sort_values(by=2022, ascending=False)
filings_per_sector_2022['pct'] = filings_per_sector_2022[2022] / filings_per_sector_2022[2022].sum() * 100
filings_per_sector_2022 = filings_per_sector_2022.round(1)
filings_per_sector_2022

	2022	pct
sector
Financial Services	148421	30.4
Healthcare	84840	17.4
Technology	59326	12.1
Industrials	58828	12.0
Consumer Cyclical	38184	7.8
Basic Materials	20413	4.2
Communication Services	17756	3.6
Real Estate	17435	3.6
Consumer Defensive	17351	3.6
Energy	17138	3.5
Utilities	9016	1.8

filings_per_sector_year_piv.loc[2022].sort_values(ascending=True).plot(kind='pie', figsize=(5, 5), autopct='%1.0f%%', startangle=90, fontsize=8)
plt.title("SEC Filings per Sector in 2022"), plt.ylabel(""), plt.tight_layout()
plt.show()

Observations:

• The aggregated annual filing volume across all sectors reached its all-time high in 2021 with over 500,000 filings for the year, and exceeded the last record from 2006, just before the financial crisis.
• The Financial Services sector has dominated in terms of filing volume since 2004. Prior to this, the Technology sector led the pack.
• While the annual filing volume in the Financial Services sector has generally been on the rise, notable exceptions include the periods during the 2008/09 financial crisis and between 2014 to 2016.
• In 2022, the Financial Services sector contributed to 30% of all sector-linked filings, totaling nearly 150,000. The Healthcare sector trailed, with 85,000 filings, accounting for 17% of the year's total volume. Together, the Financial Services and Healthcare sectors constituted 47% of the entire filing volume in 2022.
• The Healthcare sector stands out as the only sector, apart from Financial Services, to display a consistent upward trajectory in filing volume. It began its ascent in 2012 and reached a record high in 2021.
• Contrarily, the filing volume across most other sectors has been waning since 2007.
• While the Technology, Industrials, and Consumer Cyclical sectors saw a modest uptick between 2020 to 2021, their volumes in 2022 were still significantly lower than their 2007 peaks.
• Sectors such as Utilities, Real Estate, Energy, Basic Materials, and Communications Services exhibit minimal fluctuations in their annual filing volumes. This stability might suggest a lower likelihood of unexpected information or surprises, potentially leading to reduced stock price volatility within these sectors.

---

The SEC uses SIC codes to help organize and delegate the oversight and review of company filings based on industry specialization within the Division of Corporation Finance.

The SEC's Division of Corporation Finance is responsible for ensuring that investors are provided with material information to make informed investment decisions, primarily through the review of documents that publicly traded companies are required to file with the SEC.

Assignment of Review Responsibility: Within the Division of Corporation Finance, there are multiple offices responsible for the review of filings. These offices are often organized based on industries. To determine which office reviews which company's filings, the SEC frequently uses the company's SIC code to categorize it into the appropriate industry.

The SIC Codes system classifies industries with a three- or four-digit code. Every publicly traded company has an associated SIC code that represents its primary business activity.

For example, a company primarily involved in the software industry has a different SIC code than a company in the pharmaceutical industry. When these companies file documents with the SEC, the filings of the software company are reviewed by an office specializing in technology companies, while the filings of the pharmaceutical company are reviewed by an office specializing in healthcare companies.

Keep in mind that not every filer has an associated SIC code, so some may not be linked to a specific SEC Office in our analysis. Additionally, a single filing can reference multiple EDGAR entities, each having distinct SIC codes. In these instances, the filing contributes to the annual volume count for each entity, influencing multiple SIC codes simultaneously.

Before diving into the analysis, we'll enhance the filings_per_cik_year_meta dataframe from the previous section by introducing a new office column. This will be achieved by merging it with the sic_codes dataframe.

sic_codes = helpers.get_sic_codes()
sic_codes.head()

	sic_code	office	industry_title
0	100	Industrial Applications and Services	AGRICULTURAL PRODUCTION-CROPS
1	200	Industrial Applications and Services	AGRICULTURAL PROD-LIVESTOCK & ANIMAL SPECIALTIES
2	700	Industrial Applications and Services	AGRICULTURAL SERVICES
3	800	Industrial Applications and Services	FORESTRY
4	900	Industrial Applications and Services	FISHING, HUNTING AND TRAPPING

filings_per_cik_year_meta = filings_per_cik_year_meta.merge(sic_codes[['sic_code', 'office']], how='left', on='sic_code')
filings_per_cik_year_meta.head()

	year	cik	count	sector	industry	sic_code	sic_desc	office
0	1994	100030	3	Consumer Cyclical	Auto Parts	3714	Motor Vehicle Parts & Accessories	Office of Manufacturing
1	1994	100063	2	Technology	Scientific & Technical Instruments	3823	Industrial Instruments For Measurement, Displa...	Industrial Applications and Services
2	1994	100122	2	NaN	NaN	4911	Electric Services	Office of Energy & Transportation
3	1994	100166	9	Consumer Cyclical	Textile Manufacturing	2253	Knit Outerwear Mills	Office of Manufacturing
4	1994	100240	26	NaN	NaN	4833	Television Broadcasting Stations	Office of Technology

filings_per_office_year = filings_per_cik_year_meta[['year', 'office', 'count']].groupby(['year', 'office']).sum().reset_index()
filings_per_office_year

	year	office	count
0	1994	Industrial Applications and Services	2281
1	1994	Office of Energy & Transportation	6079
2	1994	Office of Finance	9432
3	1994	Office of Finance or Office of Crypto Assets	96
4	1994	Office of International Corp Fin	3
...	...	...	...
340	2022	Office of Manufacturing	62537
341	2022	Office of Real Estate & Construction	51960
342	2022	Office of Structured Finance	36211
343	2022	Office of Technology	58047
344	2022	Office of Trade & Services	59321

345 rows × 3 columns

filings_per_office_year_piv = filings_per_office_year.pivot(index='year', columns='office', values='count')
filings_per_office_year_piv.fillna(0, inplace=True)
filings_per_office_year_piv = filings_per_office_year_piv.astype(int)
filings_per_office_year_piv[-10:]

office	Industrial Applications and Services	Office of Crypto Assets	Office of Energy & Transportation	Office of Finance	Office of Finance or Office of Crypto Assets	Office of International Corp Fin	Office of Life Sciences	Office of Manufacturing	Office of Real Estate & Construction	Office of Structured Finance	Office of Technology	Office of Trade & Services
year
2013	43728	2563	67232	104434	4504	1105	30173	73472	41826	10577	59810	71086
2014	44928	2269	66902	107530	4491	1214	35766	69805	41053	11527	58563	70697
2015	44341	2756	62116	104487	4458	1106	37994	68416	40788	13183	57278	67424
2016	43006	2389	57984	100819	4486	1076	37422	63485	40392	13476	54459	61462
2017	41892	1777	55174	113094	2949	1291	39026	60722	40786	14452	52776	61281
2018	40130	1947	53462	119790	3360	1132	41982	59238	40296	16573	54994	57885
2019	39827	1449	49991	124669	3256	1446	45567	56446	38339	19690	52331	55110
2020	42211	1648	47604	156109	3546	1277	52325	57333	47537	29617	52426	58704
2021	44116	1159	47597	186159	4300	1175	55884	61027	73389	33252	59367	61124
2022	42733	1294	45793	196203	5144	892	55581	62537	51960	36211	58047	59321

print("Total number of filings filed in 2022 with entities that are assigned a SIC Code:")
filings_per_office_year_piv.loc[2022].sum()

Total number of filings filed in 2022 with entities that are assigned a SIC Code:

615716

plot_2_figs(filings_per_office_year_piv, "Filing Volume per SEC Office per Year (1994 - 2022)")

plot_box_plot(filings_per_office_year_piv,
              "Annual Filing Volume per SEC Office (1994 - 2022)", "", "Annual Filing Volume")

filings_per_office_2022 = pd.DataFrame(filings_per_office_year_piv.loc[2022]).sort_values(by=2022, ascending=False)
filings_per_office_2022['pct'] = filings_per_office_2022[2022] / filings_per_office_2022[2022].sum() * 100
filings_per_office_2022 = filings_per_office_2022.round(1)
print("Annual Filing Volume and Percentage Distribution per SEC Offices in 2022:")
filings_per_office_2022

Annual Filing Volume and Percentage Distribution per SEC Offices in 2022:

	2022	pct
office
Office of Finance	196203	31.9
Office of Manufacturing	62537	10.2
Office of Trade & Services	59321	9.6
Office of Technology	58047	9.4
Office of Life Sciences	55581	9.0
Office of Real Estate & Construction	51960	8.4
Office of Energy & Transportation	45793	7.4
Industrial Applications and Services	42733	6.9
Office of Structured Finance	36211	5.9
Office of Finance or Office of Crypto Assets	5144	0.8
Office of Crypto Assets	1294	0.2
Office of International Corp Fin	892	0.1

Observations:

1. The majority of SEC Offices have seen a decline in their annual filing volume since the period of 2005-2007.
2. A notable exception is the Office of Finance, which has experienced robust growth in annual filing volume since 1994. Its volume dipped only during 2007-2009 and 2014-2016. In 2022, it was responsible for nearly 200,000 filings, accounting for 32% of the total volume across all SEC Offices.
3. The Office of Structured Finance also deviates from the general trend. After peaking in 2006 with 30,000 filings, it saw a precipitous drop to under 5,000 by 2009. However, it has been on an upward trajectory since then, reaching a new record in 2022.
4. The Office of Life Sciences presents a unique trend as well. After a slight decline from 2006 to 2012, it has consistently grown year-over-year, achieving its highest filing volume in 2022.
5. The Office of Real Estate & Construction exhibited an anomaly in its filing trend. It witnessed a rapid growth from 2019 to 2021, only to experience a decline in 2022.

Office of Structured Finance

The Office of Structured Finance focuses on the oversight of structured finance products. In particular, the office focuses on securitized products, primarily asset-backed securities (ABS). ABS are securities that are collateralized by a pool of assets, such as mortgages (Mortgage-Backed Securities or MBS), auto loans, credit card receivables, student loans, and other types of financial assets. Given the complexity and diversity of structured finance products, and the role they played in the 2007-2008 financial crisis, the office's responsibilities are critical for the stability and integrity of the financial markets.

---

filings_per_sic_year = filings_per_cik_year_meta[['year', 'sic_code', 'count']].groupby(['year', 'sic_code']).sum().reset_index()
filings_per_sic_year

	year	sic_code	count
0	1994		25011
1	1994	100	41
2	1994	1000	59
3	1994	1040	105
4	1994	1044	18
...	...	...	...
12591	2022	8888	611
12592	2022	8900	93
12593	2022	900	134
12594	2022	9721	30
12595	2022	9995	19

12596 rows × 3 columns

filings_per_sic_year_piv = filings_per_sic_year.pivot(index='year', columns='sic_code', values='count')
filings_per_sic_year_piv.fillna(0, inplace=True)
filings_per_sic_year_piv = filings_per_sic_year_piv.astype(int)

top_20_sics_2022 = filings_per_sic_year_piv.T.sort_values(by=2022, ascending=False)[2022].head(20).index.to_list()

plot_2_figs(filings_per_sic_year_piv[top_20_sics_2022], "Filing Volume per SIC Code per Year (1994 - 2022)")

plot_box_plot(filings_per_sic_year_piv.T[range(2000, 2023)].T[top_20_sics_2022],
              "Annual Filing Volume per SIC Code (2000 - 2022)", "", "Annual Filing Volume")

top_20_sic_stats_2022 = pd.DataFrame(filings_per_sic_year_piv[top_20_sics_2022].T[2022])
top_20_sic_stats_2022 = top_20_sic_stats_2022.merge(sic_codes[['sic_code', 'industry_title']], how='left', on='sic_code')
top_20_sic_stats_2022['pct'] = top_20_sic_stats_2022[2022] / top_20_sic_stats_2022[2022].sum() * 100
top_20_sic_stats_2022 = top_20_sic_stats_2022.round(1)
top_20_sic_stats_2022.head(10)

	sic_code	2022	industry_title	pct
0	6021	75614	NATIONAL COMMERCIAL BANKS	19.9
1	6211	55134	SECURITY BROKERS, DEALERS & FLOTATION COMPANIES	14.5
2	2834	39586	PHARMACEUTICAL PREPARATIONS	10.4
3	6189	36211	ASSET-BACKED SECURITIES	9.5
4	6770	21693	BLANK CHECKS	5.7
5	7372	20221	SERVICES-PREPACKAGED SOFTWARE	5.3
6	6022	18699	STATE COMMERCIAL BANKS	4.9
7	6029	17810	COMMERCIAL BANKS, NEC	4.7
8	6798	16965	REAL ESTATE INVESTMENT TRUSTS	4.5
9	2836	12580	BIOLOGICAL PRODUCTS, (NO DISGNOSTIC SUBSTANCES)	3.3

Observations:

• In 2022, the NATIONAL COMMERCIAL BANKS (SIC 6021) sector submitted the most filings. Their annual volume has been on a consistent rise since 1994, with a pronounced surge from 2016 onwards. This sector reached a peak in 2022 with over 75,000 filings, a substantial jump from 27,000 in 2016. This means that the filings from this sector alone constituted 20% of all 2022 filings.
• The second-highest filing volume in 2022 was attributed to SECURITY BROKERS, DEALERS & FLOTATION COMPANIES (SIC 6211), accounting for around 55,000 filings or 15% of the year's total.
• The PHARMACEUTICAL PREPARATIONS (SIC 2834) sector has seen a consistent growth since 2012, moving from just over 20,000 annual filings to 40,000 in 2021.
• The rise of Blank Check companies (often referred to as SPACs or shell companies) was evident in 2021, marking their highest annual filing volume for this SIC code. Starting with fewer than 8,000 filings in 2019, they soared to nearly 40,000 in 2021, before halving to around 20,000 in 2022.
• Asset-backed securities (ABS - SIC 6189) filings stood out in 2006 with 35,000 annual filings, potentially acting as an early warning sign preceding the financial crisis. This category saw the fastest growth rate from 1997 to 2006. However, post the 2006 peak, ABS filings plummeted to 5,000 in 2009 but have been on an upward trajectory since, surpassing their pre-crisis peak in 2022.

---

Registration statements filed on the EDGAR system, such as Forms S-1, S-4 and others, are used by companies and other entities to register new securities before they can be offered to the public. These forms provide detailed information about the issuer and the securities being offered. The main purpose of these registration statements is to provide potential investors with the necessary information to make informed investment decisions.

The following list represents on overview of registration statements and their purposes.

Form Type	Description
S-1	This is the primary registration form for new securities and used for initial public offerings (IPOs). It provides a comprehensive overview of the company's business, financial statements, management, risks, and other relevant data.
S-1MEF	Registration adding securities to prior Form S-1 registration
S-3	Used by companies that meet specific reporting requirements, offering a "short form" registration.
S-4	Used for registration of securities in merger and acquisition situations. When one company acquires another, and as part of the deal, issues its own shares to the shareholders of the acquired company, those shares need to be registered, and the S-4 provides details about the transaction and the combined entity.
S-8	Used to register securities to be offered to employees through employee benefit plans.
S-11	Used to register securities of real estate investment trusts (REITs).
F-1	Used by foreign companies to register securities they intend to list on American exchanges.
F-3	Similar to the S-3 but for foreign issuers that meet the reporting requirements.
F-4	Equivalent to the S-4 but for foreign issuers.
F-6	Used to register American Depositary Receipts (ADRs) issued by a depositary against the deposit of the securities of a foreign issuer.
SB-1	Used by small business issuers to register securities for an aggregate offering price of up to $10 million.
SB-2	Used by small business issuers, allowing them to register securities without a limit on the aggregate offering price.
N-1A	Used by open-end mutual funds and ETFs to register under both the Securities Act of 1933 and the Investment Company Act of 1940. It provides details about the fund's objectives, strategies, risks, performance, expenses, and other information.
N-2	Similar to the N-1A but used by closed-end funds.
N-3	Used by separate accounts organized as management investment companies to register under the Investment Company Act of 1940 and to offer variable annuities.
N-4	Used by separate accounts organized as unit investment trusts to register under the Investment Company Act of 1940 and to offer variable annuities.
N-5	Used by small business investment companies and business development companies to register under both the Securities Act of 1933 and the Investment Company Act of 1940.
N-6	Used by separate accounts organized as unit investment trusts to register under the Investment Company Act of 1940 and to offer variable life insurance policies.

view_s = filing_count_stats['formType'].str.startswith('S-')
view_f = filing_count_stats['formType'].str.startswith('F-')
filing_count_stats_reg = filing_count_stats[(view_s) | (view_f)]

print("List of all registration statements used since 2000:")
filing_count_stats_reg[filing_count_stats_reg['year'] >= 2000]['formType'].sort_values().unique()

List of all registration statements used since 2000:

array(['F-1', 'F-1/A', 'F-10', 'F-10/A', 'F-10EF', 'F-10POS', 'F-1MEF',
       'F-2', 'F-2/A', 'F-3', 'F-3/A', 'F-3ASR', 'F-3D', 'F-3DPOS',
       'F-3MEF', 'F-4', 'F-4 POS', 'F-4/A', 'F-4EF', 'F-4MEF', 'F-6',
       'F-6 POS', 'F-6/A', 'F-6EF', 'F-7', 'F-7 POS', 'F-7/A', 'F-8',
       'F-8 POS', 'F-8/A', 'F-80', 'F-80/A', 'F-80POS', 'F-9', 'F-9 POS',
       'F-9/A', 'F-9EF', 'F-N', 'F-N/A', 'F-X', 'F-X/A', 'S-1', 'S-1.1',
       'S-1/A', 'S-11', 'S-11/A', 'S-11MEF', 'S-1MEF', 'S-2', 'S-2/A',
       'S-20', 'S-20/A', 'S-2MEF', 'S-3', 'S-3/A', 'S-3/A.1', 'S-3ASR',
       'S-3D', 'S-3DPOS', 'S-3MEF', 'S-4', 'S-4 POS', 'S-4/A', 'S-4/A.3',
       'S-4/A.4', 'S-4/A.5', 'S-4/A.6', 'S-4EF', 'S-4EF/A', 'S-4MEF',
       'S-6', 'S-6/A', 'S-8', 'S-8 POS', 'S-B', 'S-B/A', 'S-BMEF'],
      dtype=object)

filing_count_stats_reg_year = filing_count_stats_reg[['year', 'formType', 'count']].groupby(['year', 'formType']).sum().reset_index()
filing_count_stats_reg_year

	year	formType	count
0	1994	F-4	1
1	1994	S-1	41
2	1994	S-1/A	96
3	1994	S-11	7
4	1994	S-11/A	9
...	...	...	...
1327	2022	S-8	2526
1328	2022	S-8 POS	1583
1329	2022	S-B	13
1330	2022	S-B/A	13
1331	2022	S-BMEF	1

1332 rows × 3 columns

filing_count_stats_reg_year_piv = filing_count_stats_reg_year.pivot(index='year', columns='formType', values='count')
filing_count_stats_reg_year_piv.fillna(0, inplace=True)
filing_count_stats_reg_year_piv = filing_count_stats_reg_year_piv.astype(int)

filing_count_stats_reg_year_piv_t = filing_count_stats_reg_year_piv.transpose()
filing_count_stats_reg_year_piv_t['avg'] = filing_count_stats_reg_year_piv_t.mean(axis=1)
filing_count_stats_reg_year_piv_t = filing_count_stats_reg_year_piv_t[filing_count_stats_reg_year_piv_t['avg'] > 50]
filing_count_stats_reg_year_piv_t.drop(columns=['avg'], inplace=True)
filing_count_stats_reg_year_subset = filing_count_stats_reg_year_piv_t.transpose()
filing_count_stats_reg_year_subset.index = filing_count_stats_reg_year_subset.index.astype(int)
filing_count_stats_reg_year_subset[-10:]

formType	F-1	F-1/A	F-3	F-3/A	F-3ASR	F-4	F-4/A	F-6	F-6 POS	F-6EF	...	S-3	S-3/A	S-3ASR	S-3MEF	S-4	S-4/A	S-6	S-6/A	S-8	S-8 POS
year
2013	65	201	99	80	425	75	134	186	178	982	...	1259	1515	2039	43	6745	7535	1207	661	2097	1595
2014	113	303	214	272	52	36	67	350	176	1368	...	1219	979	2849	29	4964	3001	1340	748	2184	1666
2015	79	233	536	1460	97	44	65	164	222	844	...	1254	711	3787	20	4405	2970	1358	755	2144	2084
2016	45	160	62	120	38	40	96	114	296	880	...	1053	518	2338	17	2943	1483	1273	800	1994	2362
2017	79	227	225	45	54	33	77	148	1772	712	...	1178	673	2324	18	3017	1570	1221	776	2045	2071
2018	113	301	252	40	76	51	68	172	288	336	...	838	386	2306	21	1629	619	1113	441	2041	1771
2019	137	288	273	30	61	44	29	158	198	1114	...	800	312	2127	16	1380	533	1010	229	2006	1747
2020	166	388	160	78	65	31	55	140	324	334	...	1037	421	2242	56	1211	1110	1019	213	2253	1866
2021	254	541	160	72	108	89	213	198	178	535	...	1143	357	1983	54	1197	1101	1033	202	2807	1826
2022	205	572	163	200	65	64	184	84	228	348	...	996	345	1504	7	807	815	964	198	2526	1583

10 rows × 26 columns

plot_2_figs(filing_count_stats_reg_year_subset, "Registration Statements per Year (1994 - 2022)", figsize=(10, 5))

plot_box_plot(filing_count_stats_reg_year_subset.loc[1996:],
              "Registration Statements per Year (1996 - 2022)", "", "Annual Filing Volume")

view_s_subset = filing_count_stats_reg_year_subset.columns.str.startswith('S-')
view_f_subset = filing_count_stats_reg_year_subset.columns.str.startswith('F-')

sns.heatmap(filing_count_stats_reg_year_subset.loc[:, view_s_subset], cmap='Blues')
plt.gcf().set_size_inches(4, 4)
plt.title("Local Registration Statements per Year (1994 - 2022)"), plt.ylabel("Year"), plt.xlabel(""), plt.tight_layout()
plt.gca().yaxis.grid(color='lightgrey', linestyle='dashed', linewidth=0.5)
ax = plt.gca()
ax.tick_params(axis='x', rotation=50), ax.set_xticklabels(ax.get_xticklabels(), ha='right')
plt.show()

sns.heatmap(filing_count_stats_reg_year_subset.loc[:, view_f_subset], cmap='Blues')
plt.gcf().set_size_inches(4, 4)
plt.title("Foreign Registration Statements per Year (1994 - 2022)")
plt.ylabel("Year"), plt.xlabel(""), plt.tight_layout()
plt.gca().yaxis.grid(color='lightgrey', linestyle='dashed', linewidth=0.5)
ax = plt.gca()
ax.tick_params(axis='x', rotation=50), ax.set_xticklabels(ax.get_xticklabels(), ha='right')
plt.show()

filing_count_stats_year = filing_count_stats[['year', 'sector', 'formType', 'count']].groupby(['year', 'sector', 'formType']).sum().reset_index()

view_s1 = filing_count_stats_year['formType'] == 'S-1'
filing_count_stats_year_piv_s1 = filing_count_stats_year[view_s1].pivot(index='year', columns='sector', values='count')
filing_count_stats_year_piv_s1.fillna(0, inplace=True)

s1_sums = pd.DataFrame(filing_count_stats_year_piv_s1.sum(axis=1), columns=['sum'])
s1_sums['pct_change'] = s1_sums.pct_change(axis=0) * 100
s1_sums.fillna(0, inplace=True)
s1_sums = s1_sums.astype(int)
s1_sums[-10:]

	sum	pct_change
year
2013	533	56
2014	586	9
2015	445	-24
2016	354	-20
2017	477	34
2018	448	-6
2019	455	1
2020	877	92
2021	1452	65
2022	544	-62

plot_2_figs(filing_count_stats_year_piv_s1, "Annual S-1 Filing Volume per Sector (1994 - 2022)")

plot_box_plot(filing_count_stats_year_piv_s1.loc[1996:],
              "Annual S-1 Filing Volume per Sector (1996 - 2022)", "", "Annual Filing Volume", figsize=(4,3))

view_s1a = filing_count_stats_year['formType'] == 'S-1/A'
filing_count_stats_year_piv_s1a = filing_count_stats_year[view_s1a].pivot(index='year', columns='sector', values='count')
filing_count_stats_year_piv_s1a.fillna(0, inplace=True)

plot_2_figs(filing_count_stats_year_piv_s1a, "Annual S-1/A Filing Volume per Sector (1994 - 2022)")

plot_box_plot(filing_count_stats_year_piv_s1a.loc[1996:],
              "Annual S-1/A Filing Volume per Sector (1996 - 2022)", "", "Annual Filing Volume", figsize=(4,3))

Observations:

1. IPOs: In 2021, the number of IPOs, as gauged by S-1 filings, reached an unprecedented all-time high with over 1,400 submissions. Remarkably, the combined IPO activity in the Industrials and Healthcare sectors accounted for more than half of all IPOs. This volume even surpassed the flurry seen during the pre-dot-com bubble era, marking the highest since the 2000 bubble.
2. Healthcare Sector:
   • The Healthcare sector recorded its most significant growth in S-1 filings between 2008 and 2014, briefly dipped from 2014 to 2016, and then hit a record in 2020 with nearly 350 S-1 filings.
   • The momentum in S-1 filings has been upward since 2008, climaxing in 2020 with over 300 filings.
3. Tech-Boom:
   • The late '90s tech boom is evident with a surge in S-1 filings from 1996 to 1999, which saw a more than 100% growth. However, post the 1999 peak of 350 filings, there was a sharp decline to less than 50 by 2001. A significant resurgence occurred between 2019 to 2021, with filings spiking to over 300 in 2021 from a stable base of 70 in 2019. This uptrend was followed by a dip in 2022, dropping below 100.
4. M&A Activity:
   • S-4 and S-4/A filing volumes suggest heightened M&A activity during 2003-2004 and 2010-2013, possibly reflecting post-crisis consolidations.
   • Overall, M&A activity, as suggested by these filings, has been on the decline post-2013.
5. Additional Filings:
   • 2021 saw a peak in S-1/A filings.
   • F-3/A filings experienced a spike in 2015.
   • F-4/A filings had notable increases in 2004 and 2012.
   • F-6 POS peaked in 2017, while F-6EF had distinct spikes in 2008, 2014, and 2019.
6. SPAC Trend: The rise of Special Purpose Acquisition Companies (SPACs) from 2019 to 2021 can be attributed to the Industrials sector.
7. Post Dot-Com Bubble: After the dot-com bubble burst, S-1 filings plummeted by 61% from 583 in 2000 to 206 in 2001. While the volume remained flat between 2001 and 2003, there was a significant 92% increase from 2003 to 2004.
8. Amendments: The volume of S-1/A filings consistently surpassing S-1 filings underscores the continuous flow of new information post the initial S-1 submission. This highlights the importance of monitoring not just the original S-1 but all subsequent amendments for a comprehensive understanding.

From S-4 Excursion

Form S-4 is used for the registration of securities involved in business combinations, such as mergers, consolidations, or transfers of assets. An increase in the number of S-4 filings in a given year can indicate several things:

• M&A Activity: A rise in the number of S-4 filings typically points to an increase in merger and acquisition (M&A) activity. More companies might be merging, acquiring others, or undergoing asset transfers.
• Favorable Market Conditions: Companies often engage in M&A activities when market conditions are favorable, interest rates are low, and access to capital is easier. An uptick in S-4 filings could signal optimistic market conditions or might reflect optimism about future conditions. Companies might be merging or acquiring in anticipation of future growth.
• Industry Consolidation: It could indicate consolidation within specific industries. For example, if a particular sector is undergoing significant changes, companies within that sector might merge to better compete or achieve economies of scale.
• Tax or Regulatory Changes: Changes in tax laws or regulations can sometimes motivate companies to engage in M&A activities to optimize their structures or operations in light of the new rules.

From F-6 POS Excursion

Form F-6 is used to register American Depositary Receipts (ADRs) issued by a depositary against the deposit of the securities of a foreign issuer. ADRs are a way for U.S. investors to purchase shares of foreign companies without buying the actual shares directly. Instead, they buy the ADRs, which represent shares held in a bank.

The "POS" in "F-6 POS" stands for "Post-Effective Amendment". A Post-Effective Amendment is filed to update, amend, or provide additional information to a previously filed registration statement after it has become effective.

In the context of "F-6 POS," it means that the depositary is amending a previously effective registration statement related to the ADRs. This could be due to various reasons, such as changes in the terms of the ADRs, changes in the underlying agreement, or to provide updated information about the foreign issuer.

---

The "NT" in EDGAR filings stands for "Non-Timely." These filings are used when a registrant knows it will not be able to file a required report on time. The "NT" filings allow the registrant to notify the SEC and investors that there will be a delay and provide reasons for the delay. Here are the specific "NT" forms:

• Form NT 10-K: This form is filed when a company cannot timely file its annual report, Form 10-K. The company is required to file the Form NT 10-K within one business day after the due date of the 10-K.
• Form NT 10-Q: This form is used when a company cannot timely file its quarterly report, Form 10-Q. Similar to the NT 10-K, the company must file the Form NT 10-Q within one business day after the due date of the 10-Q.
• Form NT 20-F: This is for foreign private issuers who can't file their annual report on Form 20-F on time.
• Form NT 11-K: This is used when the required annual report of employee stock purchase, savings, and similar plans (Form 11-K) cannot be filed on time.
• Other forms: 10-D, N-CEN, N-CSR

When filing an "NT" form, the company typically has to state the reasons for the delay and, in some cases, when it expects to file the report. If the report is not filed within the grace period (typically 15 calendar days for Forms 10-K and 20-F, and 5 calendar days for Form 10-Q), the company can face penalties or other regulatory actions from the SEC.

Potential reasons for a company's filing delay, which could influence its stock price, include:

• Pending Financial Restatements: If a company identifies errors in its previously issued financial statements and decides to restate them, it might need additional time to finalize the corrected statements.
• Litigation or Regulatory Issues: Pending litigation or regulatory investigations can introduce uncertainties that affect the financial statements. The company might delay its filing until there's more clarity on these matters.
• Internal Controls Issues: A company might identify deficiencies or material weaknesses in its internal control over financial reporting. Addressing and documenting these issues can delay the preparation of financial statements.

Possible reasons that are less likely to impact a company's stock price include:

• Acquisitions or Mergers: If a company recently underwent a significant merger or acquisition, the complexities of integrating and accounting for the acquired entity could result in delays.
• Management Turnover: If key members of the financial or executive team have recently left the company or if there's been significant turnover, the company might face challenges in preparing its financial reports on time.
• Complex Transactions: The company might have engaged in complex transactions during the period, such as debt restructuring, which requires additional time to account for and disclose appropriately.
• External Factors: Natural disasters, pandemics, or other significant external events might disrupt a company's ability to prepare and finalize its reports.
• System or Technology Failures: Technical issues, such as system outages or failures in financial reporting software, can hinder the preparation of financial statements.

view_nt = filing_count_stats['formType'].str.startswith('NT ')
filing_count_stats_nt = filing_count_stats[view_nt]

print("Late Filing Types:")
filing_count_stats_nt[filing_count_stats_nt['year'] >= 2000]['formType'].unique()

Late Filing Types:

array(['NT 10-Q', 'NT 10-K', 'NT 11-K', 'NT 10-Q/A', 'NT 10-K/A',
       'NT 20-F', 'NT 11-K/A', 'NT 15D2/A', 'NT 20-F/A', 'NT 15D2',
       'NT 10-D', 'NT 10-D/A'], dtype=object)

filing_count_stats_nt_year = filing_count_stats_nt[['year', 'formType', 'count']].groupby(['year', 'formType']).sum().reset_index()

filing_count_stats_nt_year_piv = filing_count_stats_nt_year.pivot(index='year', columns='formType', values='count')
filing_count_stats_nt_year_piv.fillna(0, inplace=True)
filing_count_stats_nt_year_piv = filing_count_stats_nt_year_piv.astype(int)

filing_count_stats_nt_year_piv_t = filing_count_stats_nt_year_piv.transpose()
filing_count_stats_nt_year_piv_t['avg'] = filing_count_stats_nt_year_piv_t.mean(axis=1)
filing_count_stats_nt_year_piv_t = filing_count_stats_nt_year_piv_t[filing_count_stats_nt_year_piv_t['avg'] > 5]
filing_count_stats_nt_year_piv_t.drop(columns=['avg'], inplace=True)
filing_count_stats_nt_year_piv = filing_count_stats_nt_year_piv_t.transpose()
filing_count_stats_nt_year_piv.index = filing_count_stats_nt_year_piv.index.astype(int)
filing_count_stats_nt_year_piv[-10:]

formType	NT 10-K	NT 10-K/A	NT 10-Q	NT 10-Q/A	NT 11-K	NT 20-F
year
2013	1577	19	2767	12	22	62
2014	1467	8	2553	10	11	64
2015	1389	7	2301	8	27	64
2016	1255	18	2183	9	30	60
2017	1074	17	1874	10	18	62
2018	964	12	1671	7	14	77
2019	874	9	1555	6	3	83
2020	714	25	1068	17	7	72
2021	707	10	1880	22	11	104
2022	992	12	1632	14	15	150

plot_2_figs(filing_count_stats_nt_year_piv, "Late Filing Notices per Year (1994 - 2022)")

view_nt_10q = filing_count_stats_year['formType'] == 'NT 10-Q'
filing_count_stats_year_piv_nt = filing_count_stats_year[view_nt_10q].pivot(index='year', columns='sector', values='count')
filing_count_stats_year_piv_nt[-5:]

sector	Basic Materials	Communication Services	Consumer Cyclical	Consumer Defensive	Energy	Financial Services	Healthcare	Industrials	Real Estate	Technology	Utilities
year
2018	20.0	39.0	44.0	25.0	24.0	27.0	93.0	78.0	13.0	89.0	3.0
2019	25.0	29.0	45.0	26.0	33.0	22.0	108.0	92.0	15.0	71.0	4.0
2020	13.0	20.0	45.0	18.0	32.0	25.0	74.0	60.0	10.0	47.0	3.0
2021	24.0	41.0	59.0	24.0	15.0	32.0	131.0	537.0	12.0	106.0	9.0
2022	16.0	25.0	39.0	21.0	16.0	40.0	104.0	320.0	7.0	91.0	4.0

plot_2_figs(filing_count_stats_year_piv_nt,  "Annual NT 10-Q Volume per Sector (1994 - 2022)")

plot_box_plot(filing_count_stats_year_piv_nt.loc[1996:],
              "Annual NT 10-Q Volume per Sector (1996 - 2022)", "", "Annual Filing Volume", figsize=(4, 3))

Observations:

• NT Filing Trends:
  • The peak annual volume for NT filings occurred in 2006 with 7,500 filings, closely followed by 2002 which registered just over 7,000 filings.
  • A consistent decline in NT filings was observed post-2006, reaching a record low of under 2,000 filings in 2020. However, this trend saw a reversal in 2021 and 2022.
• Composition of NT Filings: NT 10-Q and NT 10-K submissions constitute approximately 95% of all NT filings from 1994 to 2022.
• Sectoral Indicators:
  • In the lead-up to the dot-com bubble and the subsequent financial crisis, the Technology sector dominated NT 10-Q filings, serving as a potential early indicator of market turbulence.
  • A similar pattern was evident in the Industrials sector (notably driven by SPACs), which saw a jump from 70 filings in 2020 to over 500 in 2021. Notably, other sectors remained stable during this period.
• Historical Highs: The Industrials sector's NT filings reached an all-time high in 2021, surpassing even the volumes seen during the dot-com and financial crises and outpacing the Technology sector's 2006 figures.
• Sectoral Medians: Over the period from 1994 to 2022, the Healthcare and Tech sectors recorded the highest median annual NT volumes.
• Sectoral Variability: Utilities and Real Estate sectors displayed the least variability in their NT filings, indicating more consistent filing behavior.

---

When a publicly traded company doesn't meet the $1 share price minimum or other listing requirements of a stock exchange, the exchange itself will file a Form 25 on EDGAR. Form 25 is the "Notification of Removal from Listing and/or Registration" and is used by the national securities exchanges to delist and deregister a security.

The process usually goes as follows:

1. The exchange notifies the company that it is out of compliance with listing standards.
2. The company is given a grace period to regain compliance.
3. If the company fails to regain compliance within that period, the exchange may decide to delist the security.
4. The exchange will then file a Form 25 to initiate the delisting and deregistration process.

It's worth noting that while Form 25 is the primary notification of delisting, it's the final step in a longer compliance process. Previous filings, such as Form 8-Ks, might have indicated potential compliance issues. In a Form 8-K filing, Item 3.01, titled "Notice of Delisting or Failure to Satisfy a Continued Listing Rule or Standard; Transfer of Listing," requires public companies to disclose any notifications received from an exchange indicating the company's non-compliance with listing standards.

view_25_nse = filing_count_stats['formType'].str.startswith('25-NSE')
filing_count_stats_25_nse = filing_count_stats[view_25_nse]

print("25-NSE Form Types:")
filing_count_stats_25_nse[filing_count_stats_25_nse['year'] >= 2000]['formType'].unique()

25-NSE Form Types:

array(['25-NSE', '25-NSE/A'], dtype=object)

filing_count_stats_25_nse_year = filing_count_stats_25_nse[['year', 'formType', 'count']].groupby(['year', 'formType']).sum().reset_index()

filing_count_stats_25_nse_year_piv = filing_count_stats_25_nse_year.pivot(index='year', columns='formType', values='count')
filing_count_stats_25_nse_year_piv.fillna(0, inplace=True)
filing_count_stats_25_nse_year_piv = filing_count_stats_25_nse_year_piv.astype(int)
filing_count_stats_25_nse_year_piv[-5:]

formType	25-NSE	25-NSE/A
year
2018	1650	8
2019	1612	44
2020	1794	14
2021	2026	26
2022	1970	18

filing_count_stats_25_nse_year_piv.plot(kind='bar', stacked=True, figsize=(5, 4))
plt.title("25-NSE Annual Filing Volume (2006 - 2022)")
plt.ylabel("Annual Filing Volume"), plt.xlabel("Year"), plt.tight_layout()
ax = plt.gca()
ax.yaxis.set_major_formatter(mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
handles, labels = ax.get_legend_handles_labels()
ax.legend(reversed(handles), reversed(labels), loc='center left', bbox_to_anchor=(1, 0.5), fontsize=8)
ax.grid(), ax.set_axisbelow(True)
plt.show()

view_25_nse = filing_count_stats_year['formType'] == '25-NSE'
filing_count_stats_year_piv_nse = filing_count_stats_year[view_25_nse].pivot(index='year', columns='sector', values='count')
filing_count_stats_year_piv_nse[-5:]

sector	Basic Materials	Communication Services	Consumer Cyclical	Consumer Defensive	Energy	Financial Services	Healthcare	Industrials	Real Estate	Technology	Utilities
year
2018	27	27	29	21	60	126	63	47	39	55	12
2019	28	22	46	23	59	90	70	60	38	70	9
2020	16	41	53	32	52	145	77	59	23	36	15
2021	23	39	63	29	42	181	97	70	71	84	17
2022	28	22	25	30	29	96	92	269	27	75	16

plot_2_figs(filing_count_stats_year_piv_nse, "25-NSE Annual Filing Volume per Sector (2006 - 2022)")

plot_box_plot(filing_count_stats_year_piv_nse,
              "Annual 25-NSE Volume per Sector (2006 - 2022)", "", "Annual Filing Volume", figsize=(4, 4))

Observations:

• Trends in 25-NSE Filings:
  • The annual volume of 25-NSE filings has consistently risen from 2011 through 2022.
  • 2021 marked the record year for delistings with 2,026 instances, closely followed by 2022 with 1,970 cases and 2007 with 1,944.
  • Interestingly, there was a continuous decrease from 2007 to 2010 in the wake of the financial crisis. This post-crisis period, especially 2007, saw accelerated market corrections, as indicated by the high delisting volumes.
• Sectoral Highlights:
  • The Industrials sector, potentially reflective of SPAC activities, registered the highest number of delistings in 2022, with over 250 25-NSE filings.
  • The Financial Services sector has historically been a major recipient of 25-NSE filings, witnessing a peak in 2007/2008 with over 200 filings. After reaching a low in 2014, the sector experienced its second-highest volume in 2021 with 170 filings.
  • The Technology sector ranks second in terms of delisting numbers, maintaining a median annual figure of around 70 delistings.
• Sectoral Lows:
  • From 2006 to 2022, sectors such as Utilities, Communications Services, Consumer Defensive, and Basic Materials consistently recorded the fewest annual delistings.

---

SEC Forms 3, 4, and 5 are designed to provide transparency regarding the transactions of company insiders in their own company's stock. "Insiders" are defined as officers, directors, and beneficial owners holding more than 10% of a class of the company's equity securities. In essence, these forms aim to provide transparency and prevent illegal insider trading by ensuring that the public has timely access to information about transactions by company insiders.

• Form 3 - Initial Statement of Beneficial Ownership: This form is used by insiders to report their initial ownership of the company's securities when they first become an insider. Insiders must file Form 3 within ten days of becoming an officer, director, or beneficial owner.

• Form 4 - Statement of Changes in Beneficial Ownership: This form is used by insiders to report changes in their ownership, such as purchases, sales, gifts, or other acquisitions or dispositions of the company's securities. Form 4 must be filed within two business days following the day on which the transaction occurred.

• Form 5 - Annual Statement of Beneficial Ownership: This form is used by insiders to report any transactions that should have been reported earlier on a Form 4 or were eligible for deferred reporting. It's also used to report transactions that are exempt from immediate reporting requirements, such as those under employee benefit plans. Form 5 must be filed within 45 days after the company's fiscal year-end. If there are no unreported transactions, the insider does not need to file a Form 5.

view_3_insider = (filing_count_stats['formType']=='3') | (filing_count_stats['formType']=='3/A')
view_4_insider = (filing_count_stats['formType']=='4') | (filing_count_stats['formType']=='4/A')
view_5_insider = (filing_count_stats['formType']=='5') | (filing_count_stats['formType']=='5/A')

insider_stats = filing_count_stats[view_3_insider | view_4_insider | view_5_insider]
insider_4_stats = filing_count_stats[view_4_insider]

insider_4_stats_year = insider_4_stats[['year', 'sector', 'count']].groupby(['year', 'sector'], dropna=False).sum().reset_index()

insider_4_stats_year_piv = insider_4_stats_year.pivot(index='year', columns='sector', values='count')
insider_4_stats_year_piv.fillna(0, inplace=True)
insider_4_stats_year_piv = insider_4_stats_year_piv.astype(int)
insider_4_stats_year_piv[-10:]

sector	NaN	Basic Materials	Communication Services	Consumer Cyclical	Consumer Defensive	Energy	Financial Services	Healthcare	Industrials	Real Estate	Technology	Utilities
year
2013	223494	10288	9180	21570	9264	11006	32734	23886	23321	7560	32105	5796
2014	222278	10139	8469	21345	9362	11790	32785	26436	22794	8499	30752	5837
2015	221462	9535	8790	21427	9132	10670	33391	27468	21797	8796	30358	5798
2016	210806	9033	8584	20590	8072	9977	33302	26133	21179	8853	27610	5284
2017	207690	9629	8145	20263	7946	9416	32954	26188	21108	8835	27873	4679
2018	206174	9042	8642	19557	7768	9255	31701	27719	23732	8983	26174	4311
2019	199608	8433	7802	18799	7381	8350	30479	29031	22644	8764	27127	4229
2020	206279	8434	8327	18465	7870	7220	32612	33103	21177	8978	27915	3870
2021	225712	8649	8742	21371	8219	7527	32392	37087	22818	9057	31457	3844
2022	208892	8407	7902	19156	8081	7431	29907	35239	21560	8057	29639	4008

plot_2_figs(insider_4_stats_year_piv.drop(columns=[np.nan]), "Annual Form 4 Volume per Sector (1996 - 2022)")

plot_box_plot(insider_4_stats_year_piv.loc[2003:].drop(columns=[np.nan]),
              "Annual Form 4 Volume per Sector (2003 - 2022)", "", "Filing Volume", figsize=(4, 4))

form_4_2022 = pd.DataFrame(insider_4_stats_year_piv.loc[2022]).sort_values(by=2022, ascending=False)
form_4_2022['pct'] = form_4_2022[2022] / form_4_2022[2022].sum() * 100
form_4_2022 = form_4_2022.round(1)
form_4_2022

	2022	pct
sector
NaN	208892	53.8
Healthcare	35239	9.1
Financial Services	29907	7.7
Technology	29639	7.6
Industrials	21560	5.6
Consumer Cyclical	19156	4.9
Basic Materials	8407	2.2
Consumer Defensive	8081	2.1
Real Estate	8057	2.1
Communication Services	7902	2.0
Energy	7431	1.9
Utilities	4008	1.0

Observations:

• Peak in Insider Trading Filings: A majority of sectors (10 out of 11) experienced their highest volume of insider trading filings (Form 4) in 2004, with a subsequent decline observed in the ensuing years.
• Healthcare Sector Trend: Contrary to the general trend, the Healthcare sector has witnessed a continuous rise in Form 4 filing volume since 2012. Remarkably, in 2021, its filing volume almost matched the peak observed in the Technology sector in 2004.
• Leading Sectors in 2022: The sectors with the highest number of Form 4 filings in 2022 were, in descending order: Healthcare, Finance, Technology, Industrials, and Consumer Cyclical.
• Unassigned Entities: Notably, over half (53%) of the entities mentioned in Form 4 filings in 2022 did not have a designated sector. While not all entities have a specific sector assigned to them, the companies in which the securities transactions are taking place usually have an associated sector.

The lack of sector assignment can be attributed to various factors:

Nature of the Entity: The entities can be diverse, including trusts, limited partnerships, investment vehicles, or other types of organizational structures. These entities might not operate in a specific sector but rather serve a specific purpose, such as estate planning or investment.

Lack of Public Information: Many of the entities might not be public entities. Without public filings, there might be limited information available about the primary sector of the entity.

Diverse Investments: Some entities, especially investment vehicles, might have a diverse portfolio of investments spanning multiple sectors. Assigning them to a specific sector would be inaccurate or overly simplistic.

Form 8-K is known as a "Current Report" and is filed with the SEC by publicly traded companies. The primary purpose of Form 8-K is to inform the public and the SEC of significant corporate events. This form is distinct from the periodic reports that companies are obligated to file (e.g. quarterly 10-Qs and annual 10-Ks) and is triggered by specific events.

Here are some of the key events that might trigger an 8-K filing:

1. Entry into or Termination of a Material Agreement: Major contract or the termination of such an agreement.
2. Completion of Acquisition or Disposition of Assets: Significant acquisitions or sales of assets.
3. Changes in Management: Appointments, resignations, or departures of directors or certain officers.
4. Changes in Financial Condition: Bankruptcy, receivership, or any financial difficulties.
5. Other Events: Stock splits, amendments to the company's articles of incorporation or bylaws, or changes to the company's fiscal year.
6. Notices of Delisting or Failure to Satisfy Listing Standards: If a company receives a warning from a stock exchange about potential delisting due to non-compliance with listing standards, it would file an 8-K.
7. Material Modifications to Rights of Security Holders: Changes that might affect the rights of shareholders or other security holders.
8. Results of Shareholder Votes: The outcome of significant shareholder votes during meetings.

view_8k = filing_count_stats['formType'] == '8-K'
view_8ka = filing_count_stats['formType'] == '8-K/A'
stats_8k = filing_count_stats[view_8k | view_8ka]

count_8k = stats_8k[['year', 'formType', 'count']].groupby(['year', 'formType']).sum().reset_index()
count_8k_piv = count_8k.pivot(index='year', columns='formType', values='count')

plot_2_figs(count_8k_piv, "Annual 8-K Volume (1994 - 2022)")

count_8k_sector = stats_8k[['year', 'sector', 'count']].groupby(['year', 'sector']).sum().reset_index()
count_8k_sector_piv = count_8k_sector.pivot(index='year', columns='sector', values='count')

plot_2_figs(count_8k_sector_piv, "Annual 8-K Volume per Sector (1994 - 2022)")

plot_box_plot(count_8k_sector_piv.loc[1996:],
              "Annual 8-K Volume per Sector (1996 - 2022)", "", "Annual Filing Volume", figsize=(4,4))

Observations:

• Trend in 8-K Filings:
  • The highest volume of 8-K filings was observed in 2005 with over 120,000 filings. Subsequently, there was a decline, with a slight uptick between 2019 and 2021, reaching just under 80,000 filings in 2021.
  • This downward trend could suggest companies are becoming more selective in their disclosures or that fewer events are deemed significant enough for disclosure. However, considering the increased interconnectivity in the digital age, it seems counterintuitive that there would be fewer events of note since 2005.
• Sectoral Insights:
  • The Healthcare sector has seen a consistent rise in 8-K filings since 2012, reaching an all-time high of 15,000 filings in 2022.
  • In 2005, the Tech and Financial Services sectors led in 8-K filings, with Financial Services maintaining its position as the second-largest contributor in 2022, while Tech dropped to fourth place.
  • The COVID pandemic appeared to have a minimal effect on 8-K filing volumes, with only slight increases noted in the Financial Services, Tech, and Consumer Cyclical sectors.
  • The Industrials sector witnessed a notable increase in filings from 6,000 to over 9,000 between 2019 and 2021.
• Potential Indicator of Volatility: The volume of 8-K filings per sector may serve as an indicator of annual volatility within those sectors. A higher frequency of 8-K filings might suggest greater sector volatility, as significant events that warrant disclosure could influence stock prices. Conversely, sectors with fewer 8-K filings might be associated with lower volatility.
• Sectors with Stable 8-K Filing Volumes: Utilities, Consumer Defensive, Communications Services, Basic Materials, Real Estate, and Energy sectors have exhibited stable annual 8-K filing volumes.