A.8 Data Sources & Corporate Data

Basic Prerequisites: Financial Statements & Accounting (what the statements are) and Core Financial Ratios (what you will compute from them). (these are the folder-level prerequisites for pages 02–06; page 01 states its own, smaller, entry requirements)

1. Intuition & Practical Objective

Fundamental analysis lives or dies on data quality and timing. Every ratio, screen, and backtest in this area is downstream of a factual claim - “revenue was 1000 in FY2022” - and that claim has a source, a timestamp, and a restatement history. The discipline of this folder is to make those three visible. A screen built on the wrong version of a number is not a rough approximation of a good screen; it is a different screen with a hidden, unknowable bias.

This folder is the hub. It (a) gives you the fast source lookup table below - job #1 of this folder, the single best “where do I get this number, and what will it cost me” page in the whole Accounting & Finance area - and (b) routes you to six sub-pages that walk from raw intuition (what the sources are) through EDGAR/XBRL, commercial providers, insider & ownership data, failure modes, and the advanced extension layer (building a working point-in-time pipeline).

The one-sentence essence. “Every fundamental number has a provenance - a filer, a tag, a vendor, and an as-of date - and the whole game is to keep the as-of date honest, because a backtest that reads today’s database as if it were then’s database is reading the future.”

Audience arc: beginner reads what the sources are and who publishes them; intermediate reads how to pull a number from each provider and reconcile the differences; expert reads point-in-time hygiene, restatement gaps, and pipeline architecture. The sub-pages are ordered for exactly that climb.

2. Mathematical Ground Truth & Lookup Table

The provenance identity. A fundamental data point is a four-tuple, not a scalar:

Drop the fourth component and you have a look-ahead leak. The usable-at time is the report’s filing date, not its period end:

The source lookup table (the hub deliverable). “PIT” = point-in-time / as-originally-reported capable.

Critical scaling caveat - the freshness ladder. Sources do not merely differ in cost; they differ in when they become knowable. A quarterly 10-Q is legally due 40–45 days after period end; a 13-F is filed within 45 days of quarter end; a proxy (DEF 14A) can be months later. If your screen uses “the latest 13-F” as if it were today’s ownership, you have silently injected up to one quarter of look-ahead. Fix the lag convention and hold it constant - the exact discipline 05 · Failure Modes & Practice formalizes.

3. Computational Implementation - auditing the registry

Runs on the standard library only. It turns the lookup table above into a machine-checkable registry and answers the three questions a beginner actually has: what is free, what is point-in-time, and what does the full paid stack cost?

4. Failure Modes & First-Principles Breakdowns

Hub signposts - the folder’s full failure-mode analysis lives in 05 · Failure Modes & Practice. In one line each:

1. Point-in-time bias. Reading a restated database as if it were the as-filed record lets a backtest “know” numbers that did not exist on the decision date - the single largest source of illusory alpha in fundamental screens.
2. Survivorship bias. Screening today’s live tickers silently drops every firm that went to zero - the failure mode that flatters exactly the value/distress screens that are supposed to catch those firms.
3. Restatement gaps. A number can be filed, revised, and revised again; a database that keeps only the latest value has destroyed the history you need to reproduce your own past decisions.
4. Vendor disagreement. Two “standardized” providers can report EBITDA for the same firm/period that differ by several percent, because standardization is a choice - always quote the provider.
5. API & rate limits. Bulk beats per-company pulls; the SEC wants an identifying User-Agent and roughly ≤10 req/s.

5. References

• SEC, EDGAR Application Programming Interfaces documentation (sec.gov/edgar/sec-api-documentation)
• XBRL US (xbrl.us)
• Kenneth R. French, Data Library (Dartmouth)
• Loughran, Tim & McDonald, Bill: “When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks” (JF, 2011)
• WRDS, Compustat Point-in-Time documentation and the Compustat User’s Guide
• Sloan, Richard (TAR, 1996) and Dechow, Ge & Schrand (JAE, 2010)

6. Connected Graph Bridges

• Foundational base: Financial Statements & Accounting (what the filings contain) · Core Financial Ratios (what you compute from them)
• Sibling topic: Accounting Quality & Red Flags (restatements and shenanigans are a data problem) · Quantitative Fundamental Investing (the screens this data feeds) · Fundamental Analysis & Screening (the discretionary workflow that consumes the same sources)
• Sub-pages (in-folder): 01 From Zero · 02 SEC EDGAR & XBRL · 03 Commercial Providers · 04 Insider & Ownership · 05 Failure Modes & Practice · 06 Advanced Extensions

Beginner: start at 01 · Practitioner: start at 05