Navigate unstructured Financial Intelligence

Insights into the momentum & challenges of a firm's strategy & operational effectiveness

Graphs and charts are commonly used to present financial information. Behind every number, there is a story. Stories are communicated in press releases, SEC filings, or earnings conference calls. Being able to timely analyze and capture what is relevant is important. Analyzing the text of a few documents is easy. Reading and anlyzing documents of 4000+ public companies at scale is time consuming and impossible. The solution is Qualitative Business Analysis (QBA).

CIF Qualitative Business Analysis (QBA)

As companies report earnings, a review and Q & A session on business operation with investors and analysts yields valuable insights. Since these communications are conversations not easily understood using quantitative analysis, it remains untapped until an experienced analyst reads it for insights. This human dependency is now being disrupted by an Artificial Intelligence system, namely, Qualitative Business Analyst (QBA), created by SiteFocus Inc.

Armed with Natural Language Understanding technology, QBA works without meta knowledge. It is a fully automated analytics machine capable of learning and understanding new subjects without prior training. The premise behind this development is derived from Symbolic Logic and propositional calculus. QBA learns and understands by discovering entity-context relationships embedded in unstructured text and categorized it into challenges, momentum, and works-in-progress. It uses entity-relationships that it discovers as labels, and then connects labels with context according to the category each context belongs. The end-result is a report of visual graphs, charts, tables, and narratives that goes to the heart of issues that matters.

QBA ingesting every publicly available earnings call transcript as soon as it is available. As part of the validation process, researchers have been tracking companies with large momentum and/or challenges reported by QBA to equity price histories. The result is amazingly consistent.

We surveyed the AI industry to find that QBA is the only commercially AI application capable of analyzing unstructured text without the need of constant update of meta knowledge such as ontology, dictionary, taxonomy, word vectors, or pre-defined entity definitions or relationships. QBA goes to work 24X7, ingesting new earnings call transcripts and identifying value targets. Just about every publicly exchange traded companies are being tracked for qualitative performance.

With the advent of QBA, it creates opportunities for many business applications. Here are a few:

  • Company executives getting feedbacks on its operations
  • Competitive products and niches among peer companies
  • Go-to-market strategy
  • Investors and hedge funds to determine quality of candidate investments
  • Sales team to identify sales targets based on company segment, service, and business activities like CAPEX
  • Evaluating vendors for procurement in supply chain
  • Identification of property transactions target
  • Identification of M & A or LBO targets
  • Public relation firms to find new sales leads

How Does It Work?

  • Quarterly earnings calls transcripts are submitted to QBA for processing
  • QBA automatically discovers subject matters found in these documents
  • Assesses context for relevancy of subjects, identify fallacies, strength and weakness
  • Delivers QBA reports comprise of visualization chart and graph, heat-map, tables with labeled supported excerpts associated with its analysis

Case Study: QBA Analysis of Kraft Heinz Earnings Calls

To demonstrate the effectiveness of this technology, we have analyzed Kraft Heinz Company's earnings calls with QBA.

A quick summary:

  • Kraft Heinz Company’s stock (KHC) dropped 27% following its earnings release on Feb 21, 2019
  • QBA evaluated 5 consecutive quarters of KHC’s earnings call transcripts at each of the points in time (from Q4 2017 to Q4 2018)
  • This demonstration provides a walk through of QBA’s analysis
  • Should management team have the benefit of this analysis a year ago, they would have spotted the warning signals of their business execution before it arrived at the undesirable outcome a year later