FinLab supplies daily U.S. stock and ETF/fund prices and quarterly company statements through its Python SDK. Quarterly statement values align to filing dates when combined with daily data. Some former listings are present, but their presence does not establish complete, survivorship-free coverage. FinLab does not supply options chains, ticks or intraday bars for this research workflow.
Which U.S. data fields are available?
The following observations come from data.get() checks on November 2, 2026. The latest price observation was October 30, 2026. A table's overall range is not a promise that every security has data throughout that range. Column counts include stored symbols, not a count of currently tradeable common stocks.
| SDK field | Inspected coverage | Appropriate use and limit |
|---|---|---|
us_price:adj_close |
January 4, 2016 to October 30, 2026; 10,436 stored symbol columns | Daily stock return research; inspect each symbol and corporate actions |
us_fund_price:adj_close |
January 4, 2016 to October 30, 2026; 12,109 stored symbol columns | ETF/fund research; SPY and QQQ both had observations across that sampled window |
us_income_statement:revenue |
Quarterly labels from 2015-US-Q1; 10,472 symbol columns |
Revenue factors; per-company reporting history varies |
us_balance_sheet:total_stockholders_equity |
Quarterly labels from 2015-US-Q1; 10,583 symbol columns |
Balance-sheet factors; missing observations need an explicit rule |
us_cash_flow:operating_cash_flow |
Quarterly labels from 2015-US-Q1; 10,412 symbol columns |
Cash-flow research; distinguish a quarter from trailing twelve months |
us_index_constituents:sp500 |
January 2, 2015 to October 30, 2026; the first row contained 501 members | Inspect dated membership for an S&P 500 universe; this does not establish other index tables' ranges |
us_ratios:current_ratio |
February 20, 2026 to October 30, 2026 | A short-history snapshot series; do not substitute it for a long historical factor |
How do I load U.S. stocks and ETFs in Python?
Start with the AI-assisted setup guide. For a compatible agent, the Claude Code and Codex instructions explain how the skill and SDK work together. Inspect the dataset with code such as:
Show Code
from finlab import data
data.set_market("us")
close = data.get("us_price:adj_close")
revenue = data.get("us_income_statement:revenue")
print(close["AAPL"].dropna().head())
print(close["AAPL"].dropna().tail())
print(revenue[["AAPL"]].loc["2024-US-Q2":"2024-US-Q3"])
data.set_market("us_fund")
fund_close = data.get("us_fund_price:adj_close")
print(fund_close[["SPY", "QQQ"]].dropna().tail())Stock research uses us; ETF and mutual-fund research uses us_fund. Choose the matching market for a backtest's executable portfolio. The current field catalog is in Studio's U.S. data browser; inspect its access labels and the current plan terms before relying on a field.
For a complete, downloadable example, the U.S. momentum study supplies its Python strategy. Its stock signal and ETF execution are distinct stages; inspect the study's own assumptions before adapting it.
When can a backtest use a quarterly filing?
Keep quarterly calculations on FinLab's quarter labels. When combining a quarterly FinLabDataFrame with daily data, the SDK converts it using the statement's availability dates. An explicit index_str_to_date() is useful for inspecting those dates.
Our SDK check of AAPL revenue produced these aligned observations:
| Stored quarter label | Filing-aligned date | Revenue, USD |
|---|---|---|
2024-US-Q2 |
August 2, 2024 | 85,777,000,000 |
2024-US-Q3 |
November 1, 2024 | 94,930,000,000 |
The November observation must not be available to a September decision. Filing-date alignment addresses that timing. It does not certify that every stored value is an immutable, originally reported vintage: later restatements and revised histories are a separate property. This page makes no as-reported-vintage guarantee.
Use an execution time after the information your signal requires is available. Do not add a quarter shift merely because the index contains quarter labels; inspect the SDK's alignment first. The look-ahead-bias guide explains how information timing differs from survivorship.
Does the data include delisted stocks?
Some former listings are present. In the sampled us_price:adj_close table, TWTR had non-null prices from January 4, 2016 through October 27, 2022, and ATVI from January 4, 2016 through October 20, 2023. Both also had historical revenue observations in the inspected statement table.
These examples establish presence, not completeness. They do not prove that every delisting, bankruptcy, ticker change, terminal payout or delisting return is represented. A missing final return can affect a backtest even when earlier prices are present. If exhaustive coverage is a requirement, FinLab cannot substantiate that requirement from these checks alone.
Build a dated eligibility rule rather than selecting only today's surviving securities. Inspect individual exits, missing values and corporate actions. Current company classifications should not silently become historical classifications. For an index-based universe, check the relevant membership table and its provenance separately.
Are ratios and company snapshots safe for a long backtest?
Check each series. The sampled us_ratios:current_ratio history began in February 2026, while the sampled raw statement tables had quarter labels from 2015. That ratio cannot supply a 2016 factor history. Do not infer that similarly named fields share the same range, revision policy or availability dates.
Define a historical accounting ratio from suitable raw statements with explicit reporting periods and missing-value rules. A current profile or analyst view may be useful for today's research while remaining unsuitable for a historical signal.
Which backtesting tool should I use with this data?
FinLab integrates its SDK and portfolio backtest engine. Other tools have different data preparation and simulation models. Compare Python backtesting libraries, U.S. platform workflows and Portfolio123 alternatives. The backtesting guide covers costs, benchmark alignment and report checks; the comparison index connects the options.
Data checks and update: November 2, 2026. Prepared by the FinLab research team, which develops FinLab. The table records SDK observations rather than an independent completeness audit. For education and research only, not investment advice. Historical data and backtests do not guarantee future results.