Extract Subsidiaries from 10-K Exhibit 21

Retrieve structured subsidiary data from SEC 10-K Exhibit 21 filings using the Datastream API. Includes curl, Python, and JavaScript examples.

Extract Subsidiaries from 10-K Exhibit 21

Public companies list their significant subsidiaries in Exhibit 21 of their 10-K annual filings. This data is useful for understanding corporate structure, mapping jurisdictional risk, and identifying undisclosed relationships. The API extracts and structures this data so you do not need to parse the raw exhibit.

Prerequisites

• An Omni Datastream API key (set as OMNI_DATASTREAM_API_KEY)
• Basic familiarity with REST APIs
• (Optional) Python 3.8+ or Node.js 18+ for SDK examples

Step 1 — Retrieve subsidiaries for a company

Use /v1/companies/subsidiaries to get the structured subsidiary list from the most recent 10-K Exhibit 21.

curl

curl -H "x-api-key: $OMNI_DATASTREAM_API_KEY" \
  "https://api.secapi.ai/v1/companies/subsidiaries?ticker=AAPL"

Python

from omni_datastream_py import OmniDatastreamClient

client = OmniDatastreamClient(api_key="your-api-key")

subs = client.companies.subsidiaries(ticker="AAPL")

print(f"Apple Inc — {len(subs.data)} subsidiaries")
print(f"Source: {subs.accession_number} (filed {subs.filed_at})")
print()

for sub in subs.data[:10]:
    print(f"  {sub.name} — {sub.jurisdiction}")

JavaScript

import { OmniDatastreamClient } from "@omni-datastream/sdk-js";

const client = new OmniDatastreamClient({
  apiKey: process.env.OMNI_DATASTREAM_API_KEY!,
});

const subs = await client.companies.subsidiaries({ ticker: "AAPL" });

console.log(`Apple Inc — ${subs.data.length} subsidiaries`);
console.log(`Source: ${subs.accessionNumber} (filed ${subs.filedAt})`);
console.log();

for (const sub of subs.data.slice(0, 10)) {
  console.log(`  ${sub.name} — ${sub.jurisdiction}`);
}

Expected output

Apple Inc — 87 subsidiaries
Source: 0000320193-24-000081 (filed 2024-11-01)

  Apple Asia LLC — Delaware
  Apple Asia Limited — Hong Kong
  Apple Canada Inc. — Canada
  Apple Distribution International Ltd. — Ireland
  Apple Insurance Company, Inc. — Arizona
  Apple Japan, Inc. — Japan
  Apple Korea Ltd. — Republic of Korea
  Apple Operations International Limited — Ireland
  Apple Pty Limited — Australia
  Apple Sales International Limited — Ireland

Step 2 — Analyze jurisdictional distribution

Map subsidiary counts by jurisdiction to understand geographic exposure and tax structure.

Python

from collections import Counter

subs = client.companies.subsidiaries(ticker="AAPL")

jurisdictions = Counter(sub.jurisdiction for sub in subs.data)

print("Subsidiary Jurisdiction Breakdown:")
print("-" * 40)
for jurisdiction, count in jurisdictions.most_common(10):
    bar = "#" * count
    print(f"  {jurisdiction:<25} {count:>3}  {bar}")

JavaScript

const subs = await client.companies.subsidiaries({ ticker: "AAPL" });

const jurisdictions: Record<string, number> = {};
for (const sub of subs.data) {
  jurisdictions[sub.jurisdiction] = (jurisdictions[sub.jurisdiction] || 0) + 1;
}

const sorted = Object.entries(jurisdictions).sort((a, b) => b[1] - a[1]);

console.log("Subsidiary Jurisdiction Breakdown:");
console.log("-".repeat(40));
for (const [jurisdiction, count] of sorted.slice(0, 10)) {
  const bar = "#".repeat(count);
  console.log(`  ${jurisdiction.padEnd(25)} ${String(count).padStart(3)}  ${bar}`);
}

Expected output

Subsidiary Jurisdiction Breakdown:
----------------------------------------
  Ireland                    12  ############
  Delaware                    9  #########
  California                  7  #######
  United Kingdom              6  ######
  Japan                       5  #####
  China                       5  #####
  Germany                     4  ####
  Singapore                   3  ###
  Australia                   3  ###
  Hong Kong                   3  ###

Step 3 — Compare subsidiary structures across peers

Pull subsidiary data for multiple companies to compare corporate complexity.

Python

TICKERS = ["AAPL", "MSFT", "GOOG", "AMZN", "META"]

print(f"{'Company':<10} {'Subsidiaries':>14} {'Jurisdictions':>15}")
print("-" * 42)

for ticker in TICKERS:
    subs = client.companies.subsidiaries(ticker=ticker)
    jurisdictions = set(sub.jurisdiction for sub in subs.data)
    print(f"{ticker:<10} {len(subs.data):>14} {len(jurisdictions):>15}")

Expected output

Company    Subsidiaries  Jurisdictions
------------------------------------------
AAPL               87              32
MSFT              214              48
GOOG              156              41
AMZN              312              55
META               78              29

Step 4 — Search subsidiaries by jurisdiction

Filter subsidiary data to find entities in a specific jurisdiction.

curl

curl -H "x-api-key: $OMNI_DATASTREAM_API_KEY" \
  "https://api.secapi.ai/v1/companies/subsidiaries?ticker=GOOG&jurisdiction=Ireland"

Python

subs = client.companies.subsidiaries(ticker="GOOG")

ireland_subs = [s for s in subs.data if "Ireland" in s.jurisdiction]

print(f"Google subsidiaries in Ireland: {len(ireland_subs)}")
for sub in ireland_subs:
    print(f"  {sub.name}")

Next steps

• Map supply chain risk: Combine subsidiary jurisdiction data with geopolitical risk indices.
• Tax structure analysis: Identify companies with high concentrations of subsidiaries in low-tax jurisdictions.
• Track changes year over year: Compare Exhibit 21 across multiple filing years to detect new or dissolved entities.
• Corporate hierarchy: Cross-reference subsidiary names with the /v1/entities/resolve endpoint to find publicly traded subsidiaries.

See the API Reference for the full subsidiaries endpoint specification.