How to Look Up ETF Holdings with the Dealcharts API

Every ETF and mutual fund registered with the SEC files quarterly portfolio disclosures called NPORT-P reports. These contain the complete list of holdings — every position, every CUSIP, every dollar value. The problem is that NPORT-P filings are buried in EDGAR as XML documents that are tedious to parse by hand.

The Dealcharts API provides structured, machine-readable access to this data. Two endpoints handle the entire workflow: a search API to find funds by name, and a facts endpoint that returns the full portfolio in clean JSON. No API key required. No authentication. This guide covers both endpoints with working examples.

API Quick Reference

Step 1: Search for a Fund

The search endpoint accepts a free-text query and returns matching entities across funds, CMBS deals, and ABS deals. Each result includes a

facts_url

curl "https://dealcharts.org/.netlify/functions/search?query=ishares+semiconductor"

Response:

{
  "key": "ishares-semiconductor-etf",
  "nameShort": "iShares Semiconductor ETF",
  "sector": "Fund",
  "facts_url": "https://dealcharts.org/llm/facts/fund/ishares-semiconductor-etf.json",
  "page_url": "https://dealcharts.org/capitalmarkets/funds/ishares-semiconductor-etf/"
}

The

facts_url

page_url

Step 2: Fetch the Fund Facts

curl "https://dealcharts.org/llm/facts/fund/ishares-semiconductor-etf.json"

The response is a JSON object with this structure:

{
  "name": "iShares Semiconductor ETF",
  "identifiers": {
    "lei": "5493004SPI3IF1GDIR85",
    "cik": "1100663",
    "fund_id": "S000004354"
  },
  "dates": {
    "filing_date": "2026-02-25",
    "reporting_period_start": "2025-12-31",
    "reporting_period_end": "2026-03-31"
  },
  "summary": {
    "total_positions": 38,
    "total_value_usd": 18107013004,
    "asset_categories": ["STIV", "EC", "DE"]
  },
  "holdings": [
    {
      "issuer_name": "NVIDIA Corp.",
      "cusip": "67066G104",
      "value_usd": 1448021808,
      "pct_val": 8.26,
      "asset_cat": "EC",
      "inv_country": "US"
    },
    {
      "issuer_name": "Advanced Micro Devices, Inc.",
      "cusip": "007903107",
      "value_usd": 1352965866,
      "pct_val": 7.72,
      "asset_cat": "EC",
      "inv_country": "US"
    }
  ]
}

Key fields in each holding:

• issuer_name — Company name
• cusip — CUSIP identifier for the security
• value_usd — Dollar value of the position
• pct_val — Percentage of total fund value
• asset_cat — Asset category code (EC = Equity Common, DBT = Debt, STIV = Short-Term Investment, etc.)
• inv_country — Country of investment

Step 3: Extract and Analyze

Python: Top 10 Holdings

import requests

# Step 1: Search for the fund
search = requests.get(
    "https://dealcharts.org/.netlify/functions/search",
    params={"query": "ishares semiconductor"}
).json()
facts_url = search["facts_url"]

# Step 2: Fetch the full portfolio
fund = requests.get(facts_url).json()

# Step 3: Sort holdings by weight and display top 10
holdings = sorted(fund["holdings"], key=lambda h: h["pct_val"], reverse=True)

print(f"{fund['name']} — {fund['summary']['total_positions']} positions")
print(f"Total value: ${fund['summary']['total_value_usd']:,.0f}")
print(f"Filing date: {fund['dates']['filing_date']}\n")

for i, h in enumerate(holdings[:10], 1):
    print(f"{i:2d}. {h['issuer_name']:<40s} {h['cusip']}  "
          f"${h['value_usd']:>14,.0f}  {h['pct_val']:5.2f}%")

Output:

iShares Semiconductor ETF — 38 positions
Total value: $18,107,013,004
Filing date: 2026-02-25

 1. NVIDIA Corp.                             67066G104  $ 1,448,021,808   8.26%
 2. Advanced Micro Devices, Inc.             007903107  $ 1,352,965,866   7.72%
 3. Micron Technology, Inc.                  595112103  $ 1,221,375,277   6.97%
 4. Broadcom, Inc.                           11135F101  $ 1,180,341,517   6.74%
 5. Applied Materials, Inc.                  038222105  $ 1,031,008,158   5.88%
 6. Qualcomm, Inc.                           747525103  $   952,478,221   5.43%
 7. Texas Instruments, Inc.                  882508104  $   879,246,731   5.02%
 8. Lam Research Corp.                       512807108  $   855,975,219   4.88%
 9. KLA Corp.                                482480100  $   836,261,103   4.77%
10. Marvell Technology, Inc.                 573874104  $   751,925,693   4.29%

Python: Cross-Fund Comparison

One of the more useful applications is comparing holdings across funds. Since the facts endpoint uses the same schema for every fund, you can compare any two directly:

funds_to_compare = [
    "ishares-semiconductor-etf",
    "fidelity-msci-health-care-index-etf"
]

for fundkey in funds_to_compare:
    url = f"https://dealcharts.org/llm/facts/fund/{fundkey}.json"
    fund = requests.get(url).json()
    holdings = sorted(fund["holdings"], key=lambda h: h["pct_val"], reverse=True)
    top5_pct = sum(h["pct_val"] for h in holdings[:5])
    print(f"\n{fund['name']}")
    print(f"  Positions: {fund['summary']['total_positions']}")
    print(f"  Top 5 concentration: {top5_pct:.1f}%")
    for h in holdings[:5]:
        print(f"    {h['issuer_name']}: {h['pct_val']:.2f}%")

curl: Quick One-Liner

If you already know the fund key, skip the search step entirely:

curl -s "https://dealcharts.org/llm/facts/fund/ishares-semiconductor-etf.json" \
  | jq '.holdings | sort_by(-.pct_val) | .[0:5] | .[] | {issuer_name, cusip, pct_val}'

Finding the Right Fund Key

Fund keys are URL-safe slugs of the fund name — lowercase, hyphens for spaces, special characters stripped. If you're not sure of the exact slug, the search endpoint handles fuzzy matching:

# These all work
curl "https://dealcharts.org/.netlify/functions/search?query=ishares+semiconductor"
curl "https://dealcharts.org/.netlify/functions/search?query=vanguard+gnma"
curl "https://dealcharts.org/.netlify/functions/search?query=fidelity+health+care"

For a complete list of available funds, the machine-readable sitemap lists every entity:

https://dealcharts.org/sitemap-llm.xml

Data Source and Freshness

All fund holdings data comes from SEC NPORT-P filings — the quarterly portfolio disclosure required of registered investment companies. The

dates

• filing_date — When the fund filed with the SEC
• reporting_period_start / reporting_period_end — The quarter covered
• Data lag — Typically 60 days after quarter-end (SEC filing deadline)

Holdings data is updated as new NPORT-P filings appear in EDGAR.

Using This with LLMs

The search + facts workflow maps directly to LLM tool use. If you're building an agent that answers questions about fund portfolios, the pattern is:

1. User asks: "What are the top holdings of the iShares Semiconductor ETF?"
2. Agent calls search endpoint with the fund name
3. Agent fetches the

   facts_url

   from the result
4. Agent parses the holdings array and answers with exact data

The facts JSON is designed to be self-explanatory to LLMs — field names like

issuer_name

pct_val

total_value_usd

identifiers

For more on how Dealcharts structures data for AI consumption, see our guide on LLM-Optimized Facts Endpoints for Finance.

Related Resources

• SEC EDGAR API Guide — CompanyFacts endpoint, CIK lookup, rate limits, and Python examples
• LLM-Optimized Facts Endpoints for Finance — Design principles for provenance-first financial data APIs
• BDC Holdings Dataset Download — Bulk BDC portfolio data from NPORT-P filings
• API Documentation — Full Dealcharts API reference with endpoint patterns and data directory