From Crypto to Interest Rates: A Sentiment Analysis of Q1 2018 Earnings Calls

Today we published our quarterly Sentiment Analysis Report, which summarizes last quarter’s top keyword searches and provides detailed sentiment analysis across all industries. We used Sentieo’s Transcript Sentiment Analysis feature to analyze earnings call transcripts and discover which topics companies discussed the most last quarter, versus the same quarter in 2017.

We also compared the sentiment of management and analyst sections of transcripts, and graphed these data points so you can easily see trends or discrepancies between the two. We publish these reports every quarter, so you can stay updated on information that could impact your investment decisions this year. Here are some interesting themes that came up in our research:

 

Sentiment Analysis

Management versus investor sentiment is diverging.

Our sentiment analysis on transcripts shows that a decoupling is taking hold between the language from company management and market participants. Management continues to be upbeat during earning calls and presentations, while sell-side analysts and investors are taking a more cautious stance. To learn more, download the full, free report.

 

Keyword Mentions

Two themes we look at closely in this report are Cryptocurrency and Trade Tariffs. When analyzing the number of mentions of crypto by company, the companies that are leading the conversation are currently Nvidia (crypto chip vendor), Visa (payment company), CBOE (professional crypto derivatives exchange), and Salesforce (CRM and enterprise SaaS vendor). 

Talks of trade wars have unsurprisingly become central to many companies, with a marked edge towards uncertainty, caution and even fear for the effect it will have on business. Read more on page 4 of the report.

With these quarterly reports, we are starting down the path of quantifying linguistic data. This report is a real use case of the exciting new features we recently released, like our Transcript Sentiment Report function, which is part of Sentieo Document Search.

Below is a sneak peek of the report: a sample page about cryptocurrency.

To learn more about the companies, industries, and regions where crypto and other themes are being most discussed, download the full report, which covers this sector and many more. To find out more about how to run your own sentiment analysis with Sentieo, sign up here for a free trial.

Food Tickers & Social Media: How Dine Brands ($DIN), Wendy’s ($WEN), and Domino’s ($DPZ) Are Heating Up

Dine Brands ($DIN), which owns IHOP and Applebee’s, is up 18% since the start of this month. You might have heard about their headline-making name change from IHOP (International House of Pancakes) to IHOB (International House of Burgers). See their tweet announcing the change below:

 

So did this branding change cause Dine Brands’ stock price movement? Did a successful social media campaign contribute to a forward earnings prediction?

We used Sentieo’s alternative data tool, Mosaic, to dig deeper and understand if social media traffic drove the increase. Alternative datasets like Google Trends, Website Visits (Alexa), Twitter Mentions, Instagram Followers, and Instagram Likes are offered in Mosaic and can provide an edge in analyzing consumer-facing businesses, as they often have a high correlation with revenue growth and are available ahead of traditional financial metrics for the period. As consumer behavior shifts more and more towards digital, indicators like these have become more predictive of tech and consumer company results.

Mosaic also allows users to visualize this alternative data. For example, we used Mosaic to create the chart below. Below the chart is a link to an interactive version of the graph.

In the chart below, we plotted the $DIN stock price (black line), IHOP Twitter Mentions (blue), IHOP.com website views, IHOP Instagram likes, and IHOP Instagram followers (gray). With these social media channels skyrocketing after the announcement of “IHOB” on June 11, the stock price (black line) surged.

Interactive graph

$DIN isn’t the only food company that experienced a social media boom and a price reaction following it. Last year, Wendy’s ($WEN) challenged a teenager to get 18 million retweets to receive the prize of free chicken nuggets for a year. His #nuggsforcarter hashtag set the all-time record for most retweets, and of course he received the prize.

In the Sentieo chart below, we mapped Twitter mentions of #nuggsforcarter (purple line), as well as Wendy’s stock price (gray line). We can see that the stock price increased and stayed high after the social media campaign peaked.

Sentieo

Interactive graph

But we can’t forget about pizza. Domino’s ($DPZ) has had several successful social media campaigns, beating out even Wendy’s in Q1 2018 by having 11 times their total number of total Instagram engagements (1.4 million). We plotted DPZ stock price (purple), Instagram Followers (red), and Instagram likes (black) below, and the correlation is clear.

Sentieo

Interactive graph

We cannot stress the importance of keeping a close eye on alternative data — especially social media — by using an alternative data tool like Sentieo. Stay ahead of the game with access to as many sizzling stats as possible.

Look Who’s Talking: How Active Are CEOs on Transcripts, and Why Does It Matter?

What CEOs say — or don’t say — on a transcript can mean a lot. A recent study by the University of Chicago Booth School of Business revealed the following about CEO activity on calls:

[Research] indicates that the chief executives who spoke more — presumably because they were more knowledgeable — also tended to earn more relative to the rest of the executive team. What’s more, firms that had mismatched compensation strategies (i.e., chief executives who talked more but got paid less) had lower firm value than those where talk time and compensation were aligned….In the study, the researchers used how much and how often the chief executive spoke in contrast to other firm employees as an indicator of how knowledge is dispersed throughout the firm.  (Booth School)

Thus, investors can learn a lot from CEO activity on calls, which seems to be correlated with firm financials. Sentieo can aid with this research by helping you search and highlight CEO language on calls.

We ran a search in Sentieo Doc Search across the S&P 500, using IN:TR to narrow down our searched documents to only transcripts. From there, we filtered down our search to CEO speech only.

We took a look at the results, which include a summary of the companies that have the most instances of their CEO talking on transcripts:

 

From the summary above, we see that General Electric has had its CEO talking the most frequently on its calls. We can click into these highlighted instances to specifically look at the language used by the CEO.

 

Management calls will typically seem optimistic, but you can look for clue words to get a better picture of the underlying meaning. For example, the Booth School article notes the following:

When a chief executive tells analysts that “lumpiness,” “headwinds,” or a “wait-and-see” period is on the way, it may be time for an adjustment in your portfolio. Euphemisms such as these can obscure the details of bad news, and research suggests that greater use of them is associated with lower stock prices in the quarter ahead, even after taking into account the already disclosed financial results.

We narrowed down our search to see if the GE CEO mentioned “headwinds,” and if so, how many times.

 

Here is a closer look at  the most recent mention:

 

We then exported mention results to see exactly how many times the GE CEO mentioned “headwinds” (34 times since July 2016), and also within which contexts.

Euphemisms aren’t the only phrases that analysts should track. Other linguistic indicators to watch for include:

  • Executives who are uncomfortable with what they are saying and who often use “distancing language,” changing pronouns from “I” and “we” to “the company.”
  • Openness about risks (generally thought of as a sign of genuine enthusiasm)
  • Comments on liquidity: trends in cash flow (or lack there-of) on balance sheet
  • Use of credit lines (are they too dependent on these?)
  • New product announcements, focus on research and development, or R&D

In the same vein, Harvard’s Ian D. Gow, Chicago Booth’s Steve Kaplan and Anastasia A. Zakolyukina, and David F. Larcker of Stanford researched how CEO personalities (as seen through call language) were correlated with firm performance:

[The] companies of CEOs who have extroverted personalities, as observed in the words used in the transcripts, were more likely to have lower cash flows and lower returns on assets. Top executives deemed to be more conscientious tended to run slower-growth companies, while those whose dominant personality trait was openness had companies with a greater focus on R&D.

Lastly, we used Sentieo’s sentiment analysis engine on the GE transcript to better understand the management and analyst sentiment, and any discrepancies in between them. This report the reveals the following output from past quarters’ conference calls:

 

The Booth School report and our sentiment analysis engine show that the language of CEOs should not be discounted. Words allow investment professionals to quickly gain insight into firm’s future performance, and make decisions based off this knowledge.

You May Be Losing Money By Not Redlining Language Changes in 10Ks

As 10Ks get longer and more complex, it gets harder and harder to find the material financial information within them. In fact, the average number of pages in a 10K has grown almost 34% over the past 8 years. We used Sentieo data to create the graph that shows this increase below.

Data Source: Sentieo

The HBS finance paper Lazy Prices shows that this rich information is hidden in 10Ks and investors are missing it (or at least taking their time incorporating it into stock prices).

The HBS paper explains how:

 

Changes to the language and construction of financial reports also have strong implications for firms’ future returns: a portfolio that shorts “changers” and buys “non-changers” earns up to 188 basis points in monthly alphas (over 22% per year) in the future. Changes in language referring to the executive (CEO and CFO) team, regarding litigation, or in the risk factor section of the documents are especially informative for future returns. We show that changes to the 10-Ks predict future earnings, profitability, future news announcements, and even future firm-level bankruptcies.

 

Finding changes across 10Ks seems like a time-consuming chore. However, with a redlining tool like Sentieo, you can see these changes in seconds. Let’s use the same example that the authors used in the report: bioscience/medical product company Baxter International (BAX). Follow this two-step process:

1. Enter the ticker BAX and the query IN:10K to pull up all of Baxter’s 10Ks.

2. Focus on the years where the biggest changes in 10K language were said to occur: the 10Ks from 2008-2009. Click the redlining icon, checkmark “All changes,” checkmark “Compare with (YoY) 10K – q4 2008,” and click the “Apply” button.

See the 10k below, after we applied Sentieo redlining. As the HBS report discusses, mentions of the words, “scrutiny,” “regulatory,” “laws and standards governing” were added to this 10k when compared with the one from the previous year. The purple text shows new language:

 

There was also a change in language regarding “stopped shipments” (versus “hold shipments” in previous 10K) of the company’s COLLEAGUE infusion pump product. Another addition was “substantial additional charges, including significant asset impairments” (versus just “additional charges” in the last 10K). See the revised copy in purple below.

Many news headlines flooded the media in the months following this 10K’s release in 2010, possibly resulting from the addition of these more regulatory-related and overall more negative words to the 10K. The HBS report explains this in more detail, including the fact that the stock price ultimately plummeted as a result:

 

For instance, a New York Times article published on April 24, 2010 reported that the FDA was clamping down on medical devices – in particular, on automated IV pumps used to deliver food and drugs. From the article: “The biggest makers of infusion pumps include Baxter Healthcare of Deerfield, Ill.; Hospira of Lake Forest, Ill.; and CareFusion of San Diego.” The article went on to quote an FDA official commenting that the new, tighter regulations would slow down the FDA approval process for automated pumps. Then, on May 4th (just 10 days later) the New York Times reported that the FDA had imposed a large recall on Baxter: “Baxter International is recalling its Colleague infusion pumps from the American market under an agreement with federal regulators that sought to fix problems like battery failures and software errors.” Moreover, the stock returns of Baxter International moved substantially surrounding the New York Times articles. In the two-week period around the articles, Baxter’s price burned down more than -20%…the price remained depressed, not reverting over the subsequent 6-month period. In contrast, we see no significant reaction to Baxter’s own disclosure of its 10-K on February 23, 2010, nearly two months before the news articles were published.

 

Looking at the stock price over time using Sentieo Plotter, one can see the drastic dip in 2010 after the 10K was released in February.

The HBS report also showed that companies with 10-K content changes generally demonstrated substantially lower future stock returns versus other companies without changes:

 

We first show that firms that change their reports experience significantly lower future stock returns. In particular, a portfolio that goes long “non-changers” and short “changers” earns a statistically significant 34-58 basis points per month – up to 7% per year (t=3.90) – in value-weighted abnormal returns over the following year. These returns continue to accrue out to 18 months, and do not reverse, implying that far from overreaction, these changes imply true, fundamental information for firms that only gets gradually incorporated into asset prices in the months after the reporting change. As all publicly traded firms are mandated to file 10-Ks (and 10-qs), the sample over which we show these abnormal returns is truly the universe of firms (not a small, illiquid or otherwise selected subset).

 

The importance of following 10K language changes — and using a redlining tool to quickly spot them — is clear. Seeing these linguistic clues ahead of time can help you predict the direction a price will move.