Success stories: Finding financial signals in text from SEC filings
The SEC Edgar database has more than 20 million documents. Each document can be very long with some exceeding 500 pages. There are also many new documents filed each day. On a busy day there can be more than 500 new filings.
How do you keep track of all the information and find the updates you need when you need to track companies and stocks, but you cannot possible read all of them in a timely manner? More importantly, how can you get the relevant information served to you proactively?
Here is are some examples of signals that have been using Data Simply software mining company SEC filings:
- Aetna (AET) disclosed that projection operating earnings would be at the top of their range. They disclosed this in advance of their analyst meeting. Within a day, their stock had gone up 4%.
- Lumber Liquidators (LL) began a long slide of difficulty with some accusations about chemicals in their flooring materials. Their filings actually showed problems a full 7 days ahead of a big price drop. Subsequently their stock hit a three year low.
- Macy's (M) disclosed that their earnings estimates were down. They disclosed this in a 32 page 10Q when many were wondering if they could recover. Soon after, they closed stores, and their stock was down more than 40% in three months.
- IBM (IBM) disclosed an SEC investigation into their revenue recognition accounting. These legal issues can be read about in detail in the filings.
- Square (SQ) disclosed an investor concentration by investor BlackRock. This was in their form 425 where it must be disclosed when more than 5% of a share class is held in concentration by a single entity owner. Such activist investor concentration often preclude action upon the company.
These are some of the success stories that demonstrate Data Simply software in action mining text from SEC filings as financial signals.