Get Company Financial Statements with Python

The tutorial explains how to fetch income statements, balance sheets, and cash flow statements for public companies in Python, load them into pandas, calculate key financial metrics, including profit margins, and liquidity ratios. All examples use the FinancialData.Net API, either through the official fdnpy SDK or plain requests.

Prerequisites

• Python 3
• An API key from FinancialData.Net
• A subscription that includes financial statements: Standard for US companies and Premium for international companies, as outlined in the API documentation.

pip install fdnpy pandas requests
export FDN_API_KEY="your_api_key"

The three statement endpoints

All three share the same parameters:

Authentication: Include your API key in the key query parameter (?key=API_KEY).

Quickstart with fdnpy

import os
from fdnpy import FinancialDataClient

client = FinancialDataClient(api_key=os.environ["FDN_API_KEY"])

income_statements = client.get_income_statements(identifier="MSFT", period="year")
print(income_statements[0])

Every SDK method returns a list[dict] parsed straight from the API JSON. The client uses requests under the hood and paginates automatically.

The latest annual record for MSFT:

{
  "trading_symbol": "MSFT",
  "central_index_key": "0000789019",
  "registrant_name": "MICROSOFT CORP",
  "fiscal_year": "2026",
  "fiscal_period": "FY",
  "period_end_date": "2026-06-30",
  "revenue": 331839000000.0,
  "cost_of_revenue": 106374000000.0,
  "gross_profit": 225465000000.0,
  "research_and_development_expenses": 35562000000.0,
  "general_and_administrative_expenses": 7956000000.0,
  "operating_expenses": 70228000000.0,
  "operating_income": 155237000000.0,
  "interest_expense": 3051000000.0,
  "interest_income": 3301000000.0,
  "net_income": 133749000000.0,
  "earnings_per_share_basic": 18.0,
  "earnings_per_share_diluted": 17.95,
  "weighted_average_shares_outstanding_basic": 7429000000,
  "weighted_average_shares_outstanding_diluted": 7453000000
}

Full field lists and sample responses: income statements, balance sheets, cash flow statements.

Load all three statements into pandas

The three endpoints return one record per reporting period. Join them on (trading_symbol, fiscal_year, fiscal_period, period_end_date) to get a single row per period.

import time
import pandas as pd

KEYS = ["trading_symbol", "fiscal_year", "fiscal_period", "period_end_date"]


def load_statements(client, ticker, period="year"):
    """Fetch income, balance sheet and cash flow statements as sorted DataFrames."""
    fetchers = {
        "income": client.get_income_statements,
        "balance": client.get_balance_sheet_statements,
        "cashflow": client.get_cash_flow_statements,
    }
    out = {}
    for name, fetch in fetchers.items():
        df = pd.DataFrame(fetch(identifier=ticker, period=period))
        df["period_end_date"] = pd.to_datetime(df["period_end_date"])
        out[name] = df.sort_values("period_end_date").reset_index(drop=True)
        time.sleep(0.1)

    return out


def build_metrics(s):
    """Join the three statements on period and derive margins, growth, liquidity."""
    inc = s["income"][KEYS + ["revenue", "gross_profit", "operating_income",
                              "net_income", "earnings_per_share_diluted"]]
    bs = s["balance"][KEYS + ["total_assets_current", "total_liabilities_current",
                              "total_liabilities", "total_shareholders_equity"]]
    cf = s["cashflow"][KEYS + ["cash_from_operating_activities",
                               "acquisition_of_property_plant_and_equipment"]]
    df = inc.merge(bs, on=KEYS).merge(cf, on=KEYS)

    df["gross_margin"] = df["gross_profit"] / df["revenue"]
    df["net_margin"] = df["net_income"] / df["revenue"]
    df["revenue_yoy"] = df["revenue"].pct_change()

    df["current_ratio"] = df["total_assets_current"] / df["total_liabilities_current"]
    df["liabilities_to_equity"] = df["total_liabilities"] / df["total_shareholders_equity"]
    return df

Run it:

df = build_metrics(load_statements(client, "MSFT", period="year")).set_index("fiscal_year")

print((df[["revenue", "gross_profit", "net_income"]] / 1e9).round(1))
print(df[["gross_margin", "net_margin", "revenue_yoy"]].mul(100).round(1))
print(df[["current_ratio", "liabilities_to_equity"]].round(2))

Output for MSFT, last three fiscal years (USD billions, margins in %):

Quarterly statements

Pass period="quarter":

q = client.get_income_statements(identifier="AAPL", period="quarter")
for r in q[:3]:
    print(r["fiscal_year"], r["fiscal_period"], r["period_end_date"], r["revenue"])

2026 Q3 2026-06-27 109417000000.0
2026 Q2 2026-03-28 111184000000.0
2026 Q1 2025-12-27 143756000000.0

Things to know:

• fiscal_period is Q1, Q2, Q3 for quarters and FY for annual. In the AAPL response, each fiscal year contains Q1–Q3 only, which matches 10-Q filings. Derive Q4 as FY − (Q1 + Q2 + Q3).
• fiscal_year and period_end_date differ. Apple's 2026 Q1 ends 2025-12-27. Always sort and join on period_end_date, not on calendar year.
• If you omit period, annual and quarterly records come back mixed. Filter on fiscal_period == "FY" if you do that.

Pagination

Statement endpoints return at most 50 records per call. A call to period="quarter" can fill that page, so older history needs offset.

The SDK handles this for you. With plain requests:

import os
import requests

BASE = "https://financialdata.net/api/v1"
API_KEY = os.environ["FDN_API_KEY"]
PAGE = 50


def fetch_all(endpoint, **params):
    rows, offset = [], 0
    while True:
        r = requests.get(
            f"{BASE}/{endpoint}",
            params={**params, "offset": offset, "key": API_KEY},
            timeout=30,
        )
        r.raise_for_status()
        page = r.json()
        rows.extend(page)
        if len(page) < PAGE:
            return rows
        offset += PAGE


income_statements = fetch_all("income-statements", identifier="AAPL", period="quarter")

Equivalent cURL:

curl "https://financialdata.net/api/v1/income-statements?identifier=MSFT&period=year&key=$FDN_API_KEY"

# next 50 (older) records
curl "https://financialdata.net/api/v1/income-statements?identifier=AAPL&period=quarter&offset=50&key=$FDN_API_KEY"

Screen multiple companies

tickers = ["MSFT", "AAPL", "AMZN", "NVDA"]

latest = []
for t in tickers:
    metrics = build_metrics(load_statements(client, t, period="year"))
    latest.append(metrics.iloc[-1])  # rows are sorted ascending, last = most recent FY

screen = pd.DataFrame(latest).set_index("trading_symbol")
print(
    screen[["fiscal_year", "net_margin", "revenue_yoy", "current_ratio"]]
    .sort_values("net_margin", ascending=False)
)

Fiscal year-end dates differ by company, so compare period_end_date as well as fiscal_year before ranking.

International companies

International statements live on separate endpoints (Premium plan) and cover Toronto, London, Frankfurt, Euronext Paris and Amsterdam, Tokyo, Hong Kong, Singapore, Indonesia, Malaysia, Korea, Brazil, Mexico, India, and Bombay exchanges.

GET /api/v1/international-income-statements?identifier=SHEL.L&period=year
GET /api/v1/international-balance-sheet-statements?identifier=SHEL.L&period=year
GET /api/v1/international-cash-flow-statements?identifier=SHEL.L&period=year

Differences from the US endpoints:

• identifier is the exchange-suffixed ticker (SHEL.L). CIK is not accepted.
• Records include currency_code (for example USD for Shell). Do not sum or compare across companies without checking it.

Export to CSV

Add format=csv:

import io
import pandas as pd
import requests

r = requests.get(
    "https://financialdata.net/api/v1/balance-sheet-statements",
    params={"identifier": "MSFT", "period": "year", "format": "csv", "key": API_KEY},
    timeout=30,
)
r.raise_for_status()
df = pd.read_csv(io.StringIO(r.text))
df.to_csv("msft_balance_sheet.csv", index=False)

Rate limits

From the API documentation:

FAQ

How do I get a company's balance sheet in Python? Call client.get_balance_sheet_statements(identifier="MSFT", period="year") from fdnpy, or GET /api/v1/balance-sheet-statements?identifier=MSFT&period=year&key=API_KEY with requests.

Can I look up companies by CIK instead of ticker? Yes, for US statements. Pass the 10-digit Central Index Key as identifier, for example 0000789019 for Microsoft.

Is there a free way to get financial statements? The free plan (300 requests/day) covers symbol lists, stock prices, and commodity data. US financial statements require a Standard plan. See pricing.

Does it work with AI agents? Yes. FinancialData.Net publishes an MCP server exposing the same endpoints as tools.

Can I get ratios without computing them myself? Yes. There are dedicated endpoints for liquidity, solvency, efficiency, profitability, and valuation ratios computed from the statements, plus a Key Metrics endpoint with P/E, P/B, EBITDA, and free cash flow.