Easier SEC Disclosure NLP with Standardized HTML

Ingesting SEC disclosures for algorithmic natural language processing (NLP) is difficult because the HTML is poorly formed.  Now Calcbench API users can access standardized disclosure HTML.

For instance, Microsoft's Contingencies note looks like this  -




but the HTML looks like this -



everything is a paragraph, there is no hierarchy, the headers are not headers.


Calcbench's standardized HTML looks like this -



The hierarchy of headers headers is correct and they are in sections with the text to which they refer.


To get the standardized HTML use the disclosure API (Calcbench API access required) and pass the standardized=True to the DisclosureSearchResults objects returned by the disclosure_search method , documentation.



Comments

Post a Comment

Popular posts from this blog

On Amazon and Server Lifespans

So there we were the other day, skimming the latest from Deep Quarry , a newsletter on financial statement analysis written by accounting aficionado Olga Usvyatsky.  She had been picking apart Amazon.com’s latest quarterly report , and noted that the e-commerce giant had revised the useful life of certain servers and networking equipment downward from six to five years. That change, Usvyatsky noted, will have the practical effect of reducing 2025 operating income by $700 million. Moreover, Amazon ($AMZN) also decided to retire some equipment early; that led to $920 million in accelerated depreciation in Q4 2024 and will reduce 2025 operating income by another $600 million. Wait a minute. The useful life of computer servers, why does that sound familiar…  Because Calcbench wrote about that exact issue in 2021!  Back then, we noted that Google ($GOOG) and Microsoft ($MSFT) had both extended the estimated life of their computer servers from three years to four, which had t...
Read more

McDonalds versus Chipotle Earnings and KPIs

Image
This morning (July 29) , McDonalds reported Q2 earnings.  Last week (July 24th), Chipotle Mexican Grill did the same.   Since we have the financial data catchers mitt, we thought we would roll up the sleeves and get to work a little.   First question: 'How did the firms fare?'  A quick as-filed income statement should suffice: MCD: Sales Flat.  Profits Down.  But levels are high meaning they make a lot of Big Macs consistently.     CMG: Sales way up (18.3% YoY) and Profits as measured by Net Income up 33.3% YoY.  The business is objectively smaller than MCD though.  There are other ways to compare these firms and we love to talk about our ability to showcase the tools that allow for that comparison.   Using our earnings press-release toolkit allows us to compare things like Net Profit Margins in real time for these firms.  Like this quarterly since 2010:   In the last 10 plus quarters, McDonald's is 2-3 times more pro...
Read more

Inventory as a Non-Current Asset

Image
Most financial analysts think of inventory as a current asset, and in most cases you’d be correct — except for those cases where you’re not.  Turns out, more than a few companies disclose inventory as a non- current asset. The tricky part is that in many of those cases, the non-current inventory is added to “other assets.” You, the financial analyst on the outside, need to dig into the footnotes if you want to find that information.  Typically that would be a painstaking exercise — but fear not! XBRL is coming to the rescue with a dedicated XBRL tag, InventoryNoncurrent . That allows you to use Calcbench to find the item simply and quickly. So what is non-current inventory, anyway? One example of non-current inventory comes from Pfizer (PFE). The pharma giant reported $4.57 billion in non-current inventory at the end of 2023, compared to $10.19 billion of current inventory. The information was included in the Other Financial Information note, and included the following table ...
Read more