How Does Intrinio Build Standardized Fundamental Data?

Chelsea Caltuna
June 24, 2020

The financial statements that publicly-traded companies file with the SEC can be a goldmine of information. And, like gold, it can take quite a bit of digging and refining to get real value out of those statements. You can spend weeks mapping out this data on your own (which you probably don’t have time for) or pay a firm with thousands of analysts who pore over the filings for you (which can be prohibitively expensive, especially for smaller companies).

Fortunately, there is a third option: machine learning. Intrinio’s fundamental data is powered by our proprietary machine learning technology, which helps us maintain a high level of quality while keeping costs low (which could mean five-figure savings for your business every year).

Read on to learn how we process fundamentals, what’s included in our fundamentals feed, and how you can test the data for free in our developer sandbox.

How Intrinio Processes Fundamentals

The SEC requires publicly-traded companies to file certain forms each year. Most investment firms (and other companies that make decisions using financial data) are interested in the 10-K and 10-Q filings, which represent annual and quarterly financial statements, respectively.

Companies file these forms in two formats. The first is an unstructured PDF that can really only be analyzed manually – great for full-time analysts, not so convenient for everyone else. The second is based on eXtensible Business Reporting Language, or XBRL. This format uses eXtensible Markup Language (XML) to help businesses report on their activities in a more consistent way.

XML can be, to put it lightly, difficult to work with. For example, if you look up the 2019 10-K for Apple, you’ll get something like this:

A bit of a headache, right? That’s why the SEC will usually parse this information out and show a cleaner version that’s easier to digest:

If you’re looking at one company, for one to three years at a time, the SEC interface is fine. If you want to go back more than three years, you’ll have to comb through multiple filings, but it’s doable. If you want to compare filings between multiple companies – well, that’s where it gets tricky.

Even using the XBRL taxonomy, no company files statements the same way. If you want to, say, compare Apple’s R&D expenses to Microsoft’s, you’ll quickly find that companies label these financials differently. That makes a direct comparison basically impossible.

To enable this comparison, Intrinio offers standardized financial statements. We parse out the information that is particularly relevant and useful to our customers and use propriety machine learning algorithms, such as multi-class classification, to map out the reported data against a standard set of terms that we use for all of the companies we cover. We’ve spent years training our algorithm with correct data to help it detect hidden patterns that are often not intuitive to humans.

Traditional data firms don’t use algorithms – they have teams of 100 to 1,000 people who map financial statements by hand. As we mentioned in the outset, this can be very expensive for you as the user. Our proprietary XBRL processing engine allows us to read these complex XBRL files at scale and extract relevant information. Any potential issues are flagged, then reviewed and fixed by a (human) team member. Our accuracy rate is roughly 99.8% - comparable to a team of analysts, but without the steep data costs.

In addition to the standardized financials, we calculate more than a hundred investment metrics and performance ratios to add value to your analysis. All calculations are comparable across filings and companies. We also calculate several fiscal periods that are not reported. For example, companies are required to report Q1, Q2, Q3, and FY, but not Q4. We calculate the last quarter of the fiscal year to give you a more complete picture.

We have our own database of companies and SEC filings, so you can access as-reported financials at any time. When new filings are submitted to the SEC, that data is available to you, on average, within fifteen minutes.

Inside Our Fundamentals Feed

Our US Company Fundamentals data feed includes standardized and as-reported 10-K and 10-Q statements for all US public companies. Get income statements, balance sheets, cash flow statements, and metrics such as market capitalization, EBITDA, and ROI. OTC securities, ETFs, and ADRs/foreign issuers are not covered.

This feed offers fundamental history back to 2007 and company news history back to May 2016. Visit the product page to learn more about pricing and see our frequently asked questions.

For an in-depth dataset to analyze US equities, we recommend US Core Market Data, which includes our Fundamentals data feed as well as real-time, intraday, and EOD stock prices.

Testing and Accessing Fundamental Data

Intrinio’s fundamentals are available via API, Excel, and CSV. Our tools are built by developers, for developers, and we strive to provide the most advanced data delivery platforms.

We encourage everyone to test our data in our developer sandbox. If you don’t already have a free Intrinio account, you can make one here and get your sandbox API key from the account page. (Learn more about how to navigate the sandbox in this blog.)

Here are a few useful links to get you started:

API Documentation

API Endpoints

All Filings by Company

All Fundamentals by Company

Lookup Fundamental by Company

Software Development Kits

R

Java

Ruby

Python

JavaScript

C#

Data Tags

We hope you enjoy exploring our fundamentals! If you have questions, our team is available via live chat seven days a week at Intrinio.com.