Every year, thousands of pages of earnings call transcripts are generated and analyzed by equity analysts for signals about how individual companies — and the market — will perform. Our 2018 Word On The Street report demonstrates how Sentieo’s Document Search and Transcript Sentiment tools make that data more accessible to researchers.
How It Works
We used Sentieo’s natural language processing technology to scrape all the earnings transcripts published in the last year. We then processed and cleaned the data to distinguish the keywords in the text. We highlighted the words in each transcript that occurred on the greatest weighted average basis. We also eliminated filler words (like “or”, “and”, etc.) and conducted analytics on the cleaned data, like part-of-speech tagging (picking out nouns and verbs for semantic analysis) and sentiment analysis (quantifying the tone of the text).
Natural language processing powers the Sentieo platform’s document search and transcript sentiment functionality, putting its users at the forefront of financial research technology.
The Most Popular Transcript Keywords Of 2018, By Sector
Our 2018 report spans all sectors – from consumer discretionary to utilities. Here’s a sample page of our report on the Consumer Staples sector. For the full, free report, please download it here.
We want to hear your thoughts on this report, or any of our other whitepapers! Your feedback is welcome at email@example.com. For a free trial of Sentieo or to learn more, get in touch here.
Our third article on the Fed leverages third-party political trend data as well as powerful Sentieo opinion mining to break down past speeches from top contenders for the Fed Chair. We discuss possible 2018 scenarios and delve deeper into the surprising results we come across. Brush up on the previous articles and see what’s coming up next in our series using the FedSpeak lexicon here:
We set out to analyze the historical speeches of the top Fed candidates with Sentieo’s natural language processing capabilities and in the process, we learned something interesting. It doesn’t matter.
The Federal Reserve is not a one-woman organization and while the chair tends to drive policy, the minutes reveal that the entire committee weighs in on decisions. Some subtle changes over the course of this year have changed the makeup of the FOMC into a more hawkish committee. Furthermore, the composition of the FOMC will change when four of the regional bank presidents and voting members rotate out for their peers.
Earlier this year, Daniel Tarullo resigned. And just a little over a month ago, Stanley Fischer, a longtime central banker, resigned from Fed Board of Governors. In their place, Donald Trump has nominated Randal Quarles, a monetary hawk who favors a rule-based approach to monetary policy, as vice chair for bank supervision. Unfortunately, transcripts of Mr. Quarles views on monetary policy are not readily available, so he is not included in the quantitative analysis.
Our second investigation of the Fed’s sentiment discusses the impact Chairwoman Yellen has had on the Federal Reserve since her rise to the Chair in 2014. We created and utilized our ‘FedSpeak’ lexicon to delve into the correlation between the Fed’s intentions and Yellen’s speeches before colleagues, Congress, and the press. Read the previous article and see what’s coming up next in our series here:
This piece kicks off our new series on the analysis of the Federal Reserve using Sentieo’s natural language processing power and flexible Doc Search technology. We will focus on bringing interesting ideas and surprising revelations derived from thousands of public federal reserve documents. Join us as we scrutinize meetings, congressional testimonies, and press conferences with some truly impressive technology; and see what’s coming up next in our series:
The Federal Reserve System’s Federal Open Market Committee (FOMC) meets eight times a year, at 2 p.m. Eastern Time in the basement of a nondescript, Washington, D.C. office building. The terse statements released after those meetings drive the direction of global financial markets and the meeting minutes are carefully scrutinized carefully by the media.
We parsed recent statements and minutes since 2012 using Sentieo’s natural language processing and sentiment analysis and found some interesting trends.
For the most recent statement 9/20, the strongest topic continued to be inflation, as highlighted in the unfiltered word cloud shown here.
The intensity was roughly equivalent to the prior statement, as the Fed continues to be vexed by an inflation shortfall versus expectations. Based on the statements alone, this analysis would suggest that Fed intentions have barely changed. However, when we apply sentiment analysis to the words in the documents using the Loughran-McDonald context-specific lexicon, which assigns a simple positive or negative value to words based on the financial services industry context, the 9/20 statement occurs as much more hawkish.
No other company has done a better job of attracting constant media attention than Amazon ($AMZN). With shares hovering around $1,000 per share, the retail-tech giant now stands as one of the four largest companies in the S&P 500 with a nearly $500 billion market cap. That represents a more than 50,000% return from the $1.73 IPO price two decades earlier. Investors fortunate enough to snatch up shares when it first hit the public market can comfortably call themselves millionaires.
While shares no longer look cheap by any traditional metric, money managers believe ongoing investments will result in even greater future returns. This is because Amazon has shown a remarkable ability to succeed in new spaces that it expands into. This is in many ways, the opposite of conventional wisdom. Large corporations often struggle when they stray outside of their core competencies. Amazon has been able to flip this script.
Amazon’s ability to accomplish this comes in large part from the leadership of its CEO, Jeff Bezos, who has consistently pushed the philosophy of, “Day One.” This excerpt from Amazon’s last letter to shareholders illustrates his commitment:
“Jeff, what does Day 2 look like?”That’s a question I just got at our most recent all-hands meeting. I’ve been reminding people that it’s Day 1 for a couple of decades. I work in an Amazon building named Day 1, and when I moved buildings, I took the name with me. I spend time thinking about this topic.
Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1. To be sure, this kind of decline would happen in extreme slow motion. An established company might harvest Day 2 for decades, but the final result would still come.I’m interested in the question, how do you fend off Day 2? What are the techniques and tactics? How do you keep the vitality of Day 1, even inside a large organization?
Such a question can’t have a simple answer. There will be many elements, multiple paths, and many traps. I don’t know the whole answer, but I may know bits of it. Here’s a starter pack of essentials for Day 1 defense: customer obsession, a skeptical view of proxies, the eager adoption of external trends, and high-velocity decision making.”
Of course, the true measure of success for any public company and its philosophy is how its share price performs. As Amazon’s reach has broadened into new industries, the number of companies who need to mention Amazon as a competitor has broadened as well.
We used Sentieo’s advanced document search to construct a query that uncovers every mention of Amazon as a competitor in public company filings (10Ks, 10Qs, 8Ks, earnings calls, investor presentations, etc.) in the last 10 years. In the chart below, you can see that mentions of Amazon have grown considerably over the past 10 years while the stock price has also grown in lockstep.
Drilling down into specific sectors, the same pattern shows itself. Take Air Freight and Logistics, a nascent segment of Amazon’s business, for example. It was only in 2016 that Amazon first made an announcement to lease 20-40 Boeing jets to augment their distribution capabilities. If we look at the mentions of Amazon in only Air Freight and Logistics company filings, we again see the number of mentions skyrocket.Read More
The last three years have been dismal for fundamental long/short managers, and stock picking at large. However, at Sentieo, our analysis shows that we are currently in the best environment since before the 2008 crash for picking stocks. Now, that isn’t to say that this is the best time to buy stocks, nor is it a prediction of fund performance. But, according to an analysis of one metric, cross-correlation, the current market should provide an unusually ripe environment for stock picking.
First, a bit about what we mean by cross-correlation: The pairwise correlation between two stocks is a value between -1 and 1, that indicates how likely the two securities are to move in the same direction. Over a given time period, two stocks that perform identically will have a value of 1, two stocks have no correlation at all will have a value of 0, and two stocks that are perfectly inversely related will have a value of -1.
We ran the pairwise correlations between every stock in the S&P 500 and every other stock in the index (249,500 computations!) from the 2007-8 financial crisis until now. Averaging all of the correlations provides an indicator of how much stocks move in tandem with each other. If the cross-correlation is 1, there would be no opportunities for stock picking since all stocks would move in tandem with each other. The higher the value of the index, the more difficult it is to make money by selecting individual securities at that point in time.
The graph below shows the cross-correlation for the entire S&P 500 over the past decade. There are a few important takeaways from this chart. First, it is clear that the cross correlations of the S&P 500 are at decade lows. Second, we see a preponderance of large spikes in the data.
As you can see, the spikes correspond with market shocks, the major macro events of the last decade. The jump in cross-correlation following a market shock is to be expected. When this sort of event happens, the entire market tends to turn in one direction as it collectively decides to buy or sell. The most recent inflection point, however, the 2016 election of Donald Trump in the United States, behaves differently.
The 2016 US Presidential election has driven correlations to new lows. Furthermore, correlations in the market actually began dropping prior to the November 8th election day, around the time when then-FBI Director James Comey sent a letter to Congress on October 28th. As opposed to the market shocks where the market all reacts in the same direction, it seems the collective market doesn’t know how to react to Donald Trump with any certainty. In other words, as of today, Donald Trump is an inherently uncertain entity that is creating opportunities for security selection.
Impact on Hedge-Fund Returns
As shown in the chart below, hedge fund monthly returns for long/short equity managers tend to react inversely to cross-correlation, as we would expect. This provides further validation to the idea that cross-correlation is a solid predictor of the overall environment for stock picking.
We can further apply cross-correlation to show the volatility of selected sub-sectors of the S&P 500. Doing so, we can demonstrate which specific sectors may have benefitted the most from the US election, again, purely from a stock-picking perspective.
With the S&P up 7% year to date, is it time to sell in May and go away?
It’s an old Wall Street adage, and the data appear to bear it out. Since 1950, the S&P has returned 3.4% on average for the year up to April, while returns from June to October have averaged only 0.9% over that time.
However, over the last five years, the dynamics of the monthly seasonal trade have not only changed but have become even more pronounced.
Summertime Has Been Producing Good Returns:
Beginning in 2012, January to April returns have averaged 4.9%, similar to the full series from 1950, but June to October returns have also averaged a healthy 3.95%.
Most notably, July has emerged as a very strong month, and June has turned from negative to positive. Also of note, the seasonal weakness in September has pulled forward into August.
This analysis suggests that August, not May, is the real bogeyman for investors.
Volatility Has Been Spiking In August:
Another way to come to the same conclusion is to look at the average returns of the CBOE VIX index shown below.
In Canada, a substantial issuer bid (SIB) is the formal term for a tender offer to repurchase shares. SIBs can be used to buy back an unusually large amount of shares beyond what’s allowed with a typical NCIB buyback program (Normal Course Issuer Bid). Tender offers may be a sign of improving corporate governance or savvy management taking advantage of their stock’s undervaluation. Or, large buybacks might simply be misleading demonstrations of confidence in a company’s prospects.
We’ve compiled a cheat sheet of Canadian stocks that are in the process of buying back a substantial portion of their float. We looked at the past 3 months of filings to find stocks that are:
In the process of a SIB
Have completed a SIB and continue to repurchase shares
The market capitalizations of the 5 stocks we’ve found range from C$133M to C$3,186M, so there should be a reasonable amount of liquidity for the largest stocks in this group. Without further ado, here’s our cheat sheet…
Our CEO, Alap Shah, wrote a guest article that appeared in HedgeWeek this morning on the topic of consumer-grade note-taking apps, and why they don’t work for equity analysts*.
Investment analysts, by and large, are a pretty smart group. If they can find a better way to do their job, they will. So it’s no surprise that an industry that relies so heavily on information has adopted a number of consumer-grade apps to enhance their workflow. And while better than nothing, this practice can create more problems than it tries to solve.
In the analyst community, note-taking apps such as Evernote and OneNote now often serve as the foundation for the research process, in spite of the fact that neither were designed with analysts in mind. Why are these solutions being adopted? First, they are mostly an improvement over the ubiquitous network and folder structure. Second, they are fairly cheap and easy to use. But perhaps most importantly, they bypass internal IT operations that would otherwise express security concerns with such apps. While these solutions do offer an improvement over a network and folder topography, many times they are more like putting square pegs in to a round hole – they might fit, but you’re going to have to smash them in there pretty hard.
On the surface, consumer note-taking applications appear to be a good fit to manage the enormous amount of information—broker research, news, internal notes, SEC filings, call transcripts, etc.—that forms the basis of the fundamental research process. However, there are a number of instances where these generic apps fall short, and, ultimately, inject more problems into the research process than they solve.
We analyzed over 9 million financial documents, covering more than 10,000 companies across the globe, for mentions of the self-driving car theme. We found that interest in self-driving cars has grown 8.5x in the past two years, but suspect that there is much more interest to come. Predictably, car and technology vendors were earliest in bracing for the technology’s impact, but the insurance industry is now beginning to take the threat seriously.
Self-driving cars are approaching quickly. Google unveiled its self-driving project just four years ago, while Tesla shipped the first car with its famous auto-pilot feature just one year ago. Though impressive progress has been made, much more is needed before self-driving cars reach scale. In the meantime, there have been setbacks. Last May, the first person was killed in a car operating on auto-pilot, while Uber ended its San Francisco self-driving project after a week amid permit conflicts with the DMV, along with several sightings of its cars running red lights. (The project continues in Arizona.) Despite the inevitable bumps along the way, self-driving cars—also known as “autonomous vehicles”—will almost certainly become a reality within the next two decades, and their impact will be felt massively across the transportation and logistics industries, among others. Read More
Sentieo is a research platform built by former hedge fund analysts to speed up the research workflow.
Nothing on this website should be considered investment advice. We do not make recommendations (long or short) in any securities. We do not express opinions as to whether any company's accounting practices are in violation of SEC, GAAP, IFRS or other rules/regulations.
Notice: Undefined index: viewedExitPopupWP in /var/www/html/wordpress/wp-content/plugins/exit-popup/exit-popup.php on line 157