Guides
Stock Market Historical Data for Research
Stock market historical data is the foundation for backtests, academic research and quantitative analysis. This guide covers what to check when sourcing US equity history — delisted names, adjustments, timestamps and corporate actions — and how to get it into pandas with FinzData.
What to check in stock market historical data
When you download stock market historical data for research, survivorship bias is the first risk. Many providers only keep currently listed names, which inflates backtest results. A survivorship-free universe includes delisted names with their final trading days.
Adjustments matter. Prices are stored as traded and adjustments for splits and dividends are applied on read. FinzData supports adjust='none', adjust='splits' and adjust='all'. Timestamps should be exchange time and consistent across the series. Corporate actions such as splits, dividends, mergers and name changes need point-in-time accuracy.
Granularity and coverage also differ. Daily bars are standard for long horizons; hourly, 5-minute and 1-minute bars are used for intraday work. Check the start date of the series and whether the provider offers point-in-time fundamentals alongside prices.
Free vs paid sources in general terms
Free sources are useful for exploration and recent periods for stock market historical data. They typically provide the last 12 months of data, limited rate limits and no historical depth. Check their current pricing page for limits.
Paid sources generally offer longer look-back, survivorship-free universes, bulk downloads and stable APIs. FinzData provides 16,000+ US stocks going back more than 26 years with history starting January 2000 including delisted names, and 231 ETFs. Daily, hourly, 5-minute and 1-minute bars are available. REST API returns JSON or CSV; Parquet and bulk files are on paid plans.
Free tier with FinzData includes last 12 months of every dataset, 1,000 API calls/day, no card. Paid look-back bundles are $195 for the first 5 years, $95 for each further 5 years, plus $19.95/month for daily updates. Institutional is $975 once plus $59.95/month.
Getting US equity history into pandas with FinzData
FinzData mirrors yfinance in Python. Install with pip install finzdata and import as yf. The client returns pandas DataFrames.
Explore live charts at https://finzdata.com/explore. Free 10-row samples are at https://finzdata.com/samples. Docs are at https://finzdata.com/docs and pricing at https://finzdata.com/pricing.
CODE: yfinance-style history and download
import finzdata as yf
import pandas as pd
# Ticker history with start/end and interval
ticker = yf.Ticker("AAPL")
hist = ticker.history(start="2000-01-01", end="2024-12-31", interval="1d")
print(hist.head())
# Adjustments on read via Client
prices = yf.Client().prices(sym="AAPL", start="2020-01-01", adjust="all")
print(prices.head())
CODE: Client API for prices, fundamentals and macro
import finzdata as yf
c = yf.Client()
# Prices with explicit adjustment
prices = c.prices(sym="AAPL", start="2010-01-01", end="2024-12-31", adjust="all")
print(prices.head())
# Point-in-time fundamentals
pe = c.fundamentals(ticker="AAPL", concept="pe_ratio", as_of="2023-12-31")
print(pe)
# Macro series with vintage
fed = c.macro(id="DFF", as_of="2024-06-30")
print(fed)
Related datasets for context
Stock market historical data is often used with complementary datasets. FinzData provides CFTC Commitments of Traders reports since 1986 for 950+ futures markets with COT index 0-100. SEC fundamentals include XBRL company facts point-in-time by filing date, financial statements, earnings release dates from 8-K Item 2.02 since 2004, and daily point-in-time valuation from 2009.
Additional coverage includes 13F institutional holdings, FINRA short interest since 2017, SEC Form 4 insider trades, 18,000+ US macro series with every revision, and dealer gamma levels nightly.
Practical checklist before you backtest
Confirm survivorship-free coverage and delisted names are present. Verify adjustment method and whether dividends and splits are applied consistently. Align timestamps to exchange close and handle missing sessions.
Check corporate actions and name changes for point-in-time accuracy. Validate fundamentals are point-in-time by filing date, not restated. Document data vintage for macro series.
Start with a small universe and compare a few tickers against a known source. Use the free 10-row samples to inspect schema before committing to a plan.
Frequently asked questions
What is the best source for stock market historical data?
There is no single best source. For research you want survivorship-free coverage, long history, clear adjustment rules and point-in-time fundamentals. FinzData offers 16,000+ US stocks back to January 2000 including delisted names with adjustments applied on read.
How far back does US stock market historical data go?
FinzData history starts January 2000 for US equities, more than 26 years, including delisted names. Free tier provides last 12 months; paid look-back bundles extend history in 5-year blocks.
Do I need adjusted prices for backtesting?
Yes for total return calculations. FinzData stores prices as traded and applies splits and dividends on read via adjust='none'|'splits'|'all'. Choose the method that matches your research design.
Can I get delisted stocks data?
Yes. FinzData includes delisted names in its survivorship-free universe, which is important to avoid survivorship bias in backtests.
How do I load stock market historical data into pandas?
Use import finzdata as yf. yf.Ticker('AAPL').history returns a pandas DataFrame. The Client API c.prices also returns DataFrames and supports adjust parameter.
Is there a free stock market historical data API?
FinzData free tier includes last 12 months of every dataset with 1,000 API calls/day, no card. Check the pricing page for current limits.