Archive for category: NLP

You are here: Home / NLP

Social Market Analytics, Inc.

A Leader in Unstructured Financial Data

, , , , ,

Predictive Power of Sentiment on Earnings Q&A

/by

One of the most closely followed events on the corporate calendar are earnings calls. This gives executives the opportunity to comment on earnings and answer questions from those outside of the company. Using our patented Natural Language Processing, Social Market Analytics scores Earnings Call Transcripts in real time and creates metrics based on sentiment, word count, and section count. For this research, we look specifically at the question-and-answer section of call transcripts. The theory is that isolating the section of the call where executives aren’t controlling the topic of conversation will give a more accurate assessment of the sentiment surrounding earnings results. We use Sum of Sentiment to quantify the positivity of the call. Sum of Sentiment adds all the words and phrases tagged in the section with sentiment. The following histogram shows the distribution of the Sum of Sentiment variable.

The Sum of Sentiment is centered around 3.5 and is roughly normal with a heavy tail skewing right. As executives of companies want to express good things to come, it makes sense that the sum is predominantly positive. Still some earnings calls are more positive than others. Based on the distribution of sentiment, we defined an extremely positive earning call as having a sum greater than 5 and a negative earning call as having a sum less than 1.5. These thresholds give a roughly equal number of instances over the past 14 years. We took these thresholds and compared returns for different time periods following the Earnings Call. Time periods were subsequent Open-to-Close; subsequent Close-to-Close, subsequent week return, subsequent month return, and subsequent quarter return. Since earnings calls are spaced throughout the year, it is difficult to compound the subsequent returns. Instead, we will be looking at average excess returns for each threshold. The excess return for each security is calculated by subtracting the SPY return of the same time frame from the securities return. Our hypothesis is that the average excess returns for the extremely positive earnings calls will be higher than those for negative earnings calls. We calculated returns of all instances since the end of 2009.

For long-term holdings, the average excess return for high sentiment earnings call companies was strongly positive. On the contrary, negative sentiment earning calls company returns were negative for every time frame. Quarterly returns highlight the importance of a positive earnings call as the average excess return is close to .8% higher than negative. The biggest takeaway from different time periods was the large difference in returns between the next Open-to-Close and the next Close-to-Close, especially those with a high sentiment. Entering on the subsequent close rather than open dropped the excess returns by .8% and made them negative. The next close to close returns were negative regardless of the sentiment threshold. Waiting to enter removed the benefit of positivity from high sentiment. We looked at the returns of these two-time frames with high sentiment over the past 13 years.

Looking at the past 14-year performance: The Open-to-Close excess returns were positive 12/14 years and Close-to-Close excess returns were positive 4/14 years. The two negative years for the Open-to-Close also came during an abnormal period of the COVID-19 pandemic. Immediate open to close return benefits from the high sentiment far more than the close to close. Therefore, there is a premium on knowing the sentiment of an Earnings Call in real time and entering the next open to maximize short term returns. Instead of manually reading earnings calls to gain insights, traders can use the sentiment summarized by Social Market Analytics to select positions. Waiting to enter on positive earnings calls generally hurts the short-term returns. Social Market Analytics’ scoring on Earning Calls can give traders the advantage of entering the position as quickly as possible for immediate returns, while also providing a holding option for quarterly returns.

If you are interested in learning more about how SMA’s Earnings data can help your trading strategies, please email us at contactus@socialmarketanalytics.com or schedule a demo using this link.

Explore sentiment on earnings calls and all corporate filings on the SMA Unstructured Data Terminal below.

/by
, , , ,

S-Factor Introduction Video Blog

/by

In this video we explain how our S-Factor feed uses the Twitter firehose to build trading signals, dashboards and widgets to help investors increase alpha. To find out how SMA can help your firm, email us at ContactUs@SocialMarketAnalytics.com or schedule a meeting using our 1 on 1 Meeting Signup.

/by
, , , , , , ,

Alpha Around the World: Insights into Global Annual Reports

/by

Social Market Analytics (SMA) has created the Global Machine Readable Filings (GMRF) dataset in partnership with S&P Global Market Intelligence. GMRF provides parsed textual data broken down to sections along with metrics like word count and sentiment on ­­­­annual and interim reports for companies around the world. Extraneous information such as images, tables, and page numbers are removed for textual analysis purposes.

In this blog we began investigating word count and sentiment metrics related to Annual Reports (AR) in the GMRF dataset. Each document is linked with a company and each company can be linked to multiple securities. These securities could be different class shares on the same exchange, or the same company listed on multiple global exchanges.

Since these reports only come out once a year, there will only be one value of factors per company each year. To construct a calendar-time portfolio, metrics from the most recent Annual Report for each company is extended up to 12 months. Quintile portfolios are constructed at the end of each month based on values of the most recent factors from a company’s Annual Report. In our analysis below, Quintile 1 contains 20% of stocks with the lowest value of the factor. Quintile 5 contains the 20% of stocks with the highest value of the factor.

For time-series purposes and to avoid look-ahead bias, in this analysis we assume that Annual Reports are released publicly 90 days after the fiscal year end date for that company. For example, if a file had a fiscal year end date on 12/31/2011 the word count and sentiment factors associated with that Annual Report would be implemented into the quintile portfolios on 3/31/2012.

After cleaning the dataset by removing documents with low word count, extreme changes in word count, and filings within four months of each other, we map the documents to the securities’ monthly return. Stocks are placed in quintiles based on factor values. Returns are equally weighted within each quintile.

This chart illustrates the relationship between the length of Annual Reports and future price returns. Higher word count in an Annual Report indicates the company has more information to supply to investors. Quintiles 1 and 2 outperform the universe of stocks in the dataset while Quintile 5 severely underperforms. Quintile 5 contains companies that have produced extremely long Annual Reports over the past 14 years. Long Annual Reports typically contain unnecessary language or ‘fluff’ to ease concerns of investors and portray the company in a stable fashion. This text could be information addressing specific risks and business actions or could be new information on products and legal proceedings. The theory is shorter Annual Reports are more forthright with information regarding business affecting the company’s stock.  This is consistent with the classic Harvard Lazy Prices1 paper findings.

Next, we look at the Average Sentiment across an entire Annual Report. Average Sentiment is calculated by taking the mean of sentiment from each Part within an annual report. Each Part is a level one header, or the overarching categories in a Table of Contents. The sentiment of each Part is calculated by taking the sum of each sentence within each Part. Thus, Parts with more textual data (more sentences) can have a higher magnitude of sentiment. A Total Document’s Average Sentiment is slightly correlated to the documents word count.

Here we see an inverse relationship between a document’s word count and price returns. A low Average Sentiment outperforms the rest of the universe while high Average Sentiment underperforms. This can be traced to the company’s transparency with its investors. Theory is Annual Reports that state the company’s proceedings and risks, with no sugarcoating, tend to have a lower Average Sentiment in Annual Reports.

This analysis are preliminary findings on the GMRF dataset. SMA will explore further on how characteristics of Annual Reports can affect future stock prices by sector and region. To trial this product or learn more about other Social Market Analytics datasets please email us at ContactUs@SocialMarketAnalytics.com

 

  1. Harvard Lazy Prices, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1658471
/by
, , , ,

Rare Earth Metals – Machine Readable Filings (MRF) Expose U.S. Domestic Shortage

/by

Overview

Smart phones, electric/ hybrid cars and X-Rays have undoubtedly taken an important role in today’s society. All three essential items depend on a limited resource called Rare Earth Metals. The use of Rare Earth Elements (REE) does not stop there. REEs are also used for refining crude oil, building wind turbines, making televisions and computer screens, and are necessary for production of many other important products. There is no doubt Rare Earth Metals play an important role in the world, therefore, from an investing standpoint, you might wonder who is mining these crucial metals and who is using them? This blog answers those questions by using Machine Readable Filings (MRF) and Social Market Analytics, Inc. (SMA) Complex Topic Model engine. The dataset is run across all 10-K, 10-Q, 8-K, 20-F, 6-K and 40-F filings from SEC Edgar.

Social Market Analytics, Inc. (SMA) has partnered with S&P Global Market Intelligence on ‘Machine Readable Filings’ (MRF). Machine Readable Filings is the first product to provide parsed textual data of SEC Edgar Regulatory Filings at the Total Document, Part, Item, Sub-section, and Notes level with historical baselines back to 2006. The Textual Data, Sentiment NLP, and Word Count values are provided by SMA’s patented technology. Extraneous information such as page numbers, images, and tables are removed. S&P Global and SMA also developed the same structured data values on International Company Reports.

SEC regulatory filings are formal documents reported to the U.S. Securities and Exchange Commission (SEC) that contain important information about companies such as financial statements, forward looking statements, risks associated with the company, etc. Every U.S. publicly traded company is required to submit regulatory filings to the SEC.

Companies discuss their forward focus in these documents, and you can glean a lot about a company’s direction or risks by the topics they discuss in filings. With SMA’s topic modeling capabilities you can quickly scan documents for curated topics. If you are searching for companies investing in a specific topic, this functionality makes the process much more efficient. As an example, if you want to find companies with negative exposure to Covid-19 you can search for Covid-19 mentions in the Risk Factors section of a 10-K or 10-Q. If you would like to find a company’s strategy for handling their business during Covid-19, you may want to search in the Management Discussion & Analysis section.

Machine Readable Filings (MRF) used in conjunction with SMA’s Complex Topic Model engine allows for the programmatic curation of Thematic Investing themes. Our topic modeling allows us to capture ever relevant ‘Rare Earth’ terms within SEC filings and exclude terms that are not relevant. For example, U.S. Rare Earth Minerals, Inc. mentioning their own name in a filing would not be considered relevant. SMA’s proprietary Complex Topic Model engine uses smart synonym and proximity search to make sure you capture all relevant information.

Example of the Dataset

The data is available through an API, CSV reports or other customizable options to be easily read. Below is a picture of part of the data set used for this blog. The circled context reads: “We manufacture and sell an array of plain surface prestressed steel materials and rare earth coated and zinc coated prestressed steel materials…”

In this blog we will be illustrating the power in SMA’s topic searching by identifying companies with the most mentions of the topic ‘Rare Earth’ in their filings. The Topic Model data extraction across ~1.4 Million documents going back to 2006 only takes a few minutes. The MRF textual data structure also enables the creation of baselines.

Full Rare Earth Universe

To answer who is mining and using these Rare Earth Materials, we first must look at the topic model on a macro level. Below is a chart illustrating mentions from each sector. A hit is defined as the frequency of which a term in mentioned within an SEC filing. The topic contains model contains over 40,000 hits from January 1, 2006- May 19, 2021.

The chart demonstrates that 65% of the hits come from the Materials sector. Of the Materials Sector hits, 24,768 (96%) of the hits from materials are from the Metals and Mining Industry. Using that information, we can determine that the large majority of the Materials Sector are the suppliers. The rest of the sectors appear to be the consumers. The top two consumer sectors are Health Care (4,329 hits, 11% of total hits) and Information Technology (2,908 hits, 7% of total hits). We will look in depth at the determined largest suppliers and two largest consumers.

Materials Sector

Zooming in on the suppliers, we see that the three most mentioned Rare Earth Metals are scandium with 5,191 hits, neodymium with 1,266 hits and yttrium with 1,237. The most popular term hit by far was ‘Rare Earth’ with 11,654 hits. Knowing these are the three most talked about metals by suppliers could possibly imply that they are the three most profitable or abundant.

We know which sector the supplier is in and which metals they are discussing the most. Which specific companies are discussing these most frequently? The graph below shows the top 20 companies mentioning Rare Earth terms the most in their filings from January 1, 2006 to May 19, 2021. The bars are color coded by which metal they discuss. Scandium International, a Canadian company is the most popular and they talk about scandium an overwhelming majority of the time. It might be inferred they mention their name a lot in their filings, using by SMA’s Complex Topic Model engine we are able to exclude the mentions of “Scandium International” and only capture the relevant mentions of scandium. Scandium International is very centered on scandium and discusses the metal a lot within their filings. Molycorp (last filed in 2016), although no longer publicly traded is a close second, discusses a variety of metals. The third most popular is Texas Mineral Resources Corp which has the most mentions of any current U.S. publicly traded stock.

The top two suppliers of metals are a Canadian company and an American company that are no longer publicly traded. The next three companies are all trading below $5 a share. This could imply that the U.S. is not a great supplier of Rare Earth Metals. Many articles suggest that China has the majority of Rare Earth Metals. The top companies in this chart support that the majority of the supply of Rare Earth Metals are not mined by American companies.

Consuming Sectors of Rare Earth

Now that we have looked in depth into the suppliers, we will shift to the consumers. The most popularly talked about terms in these filings from January 1, 2006 to May 19, 2021 are yttrium with 800 hits, gadolinium with 708 hits and holmium with 695.  There is a large drop to lutetium with 473 hits. This is vastly different from the suppliers. Scandium and neodymium are the 13th and 8th most talked metals within the consumers. Both suppliers and consumer talk about yttrium in their top three metals, however, the difference in conversation between the rest of the metals could hint at a disconnect.

After looking at the most metals, we can transition to which sectors use which metals. The graph below demonstrates which sectors discuss which metals. The Health Care sector discusses 6 of these metals more than any other sector. Their top being yttrium with 800 hits, gadolinium with 708 hits and holmium with 695 hits. These results are not surprising as yttrium is used in cancer treatments, gadolinium is used in MRI machines and holmium is used in holmium lasers, which are used in surgeries. Information technology talks about 5 different metals more than any other sector. The most popular being ytterbium with 358 hits, erbium with 339 hits and yttrium with 238 hits. These metals are all used in alloys to strengthen or soften metals.

The companies with the most hits from January 1, 2006 to May 19, 2021 within the healthcare sector include Vivos Inc., Trimedyne, Inc. and Immunomedics, Inc. These companies mention yttrium and holmium the most. That could imply that these companies have large dependencies on those metals.

Below shows the top 20 companies within the IT Sector. The top 3 companies include IPG Photonics, Western Digital and II-VI Incorporated. IPG Photonics develops fiber lasers, Western Digital manufactures hard drives and data center equipment, and II-VI Incorporated manufactures optical materials and semi-conductors.

It might be surprising that large companies like Apple or Intel do not make the list for top 20 mentions of rare earth metals, considering they are used in building microchips and cell phones. That could mean that they are not dependent enough on rare earth elements to be mentioned in their Risk Factors or Business portions of their filings. In that manner it is just as important to look at which companies are not mentioning Rare Earth Metals in their filings. A shortage in Rare Earth Metals may not drastically impact their business. However, these other companies are dependent enough that a shortage and an event revolving around Rare Earth Metals could impede their business.

Conclusion

How does knowing who is and who is not mentioning Rare Earth Metals help from an investment standpoint? Investing in the suppliers could be beneficial if you believe the consumer businesses will grow. If the consumers business grows, then the demand will increase, which will allow these mining companies to sell more Rare Earth Metals at higher prices, likely increasing the value of these companies. Knowing which companies are the most and least dependent on Rare Earth Metals could give you an understanding of the risk in these companies if Rare Earth Metals become scarcer due to geo-political risk, mining accidents, civil disputes or many other reasons. Although this data gives the picture of Rare Earth from American, Canadian, and ADR SEC Filings, the data is still very powerful. SMA’s Complex Topic Model engine will soon include Global filings and Sustainability Reports.

In April 2021, SMA released in partnership with S&P Global Market Intelligence ‘Global Machine Filings’ cover +200 Countries Company Annual Reports (AR). SMA will be running a Rare Earth analysis on ARs shortly.

Today’s Macro and Technology Themes are unprecedented and cover a multitude of emerging “Themes” from Drones to ESG to Genome Research to Rare Earth, etc. The investment potential is unlimited. Thematic Investing is becoming a powerful vehicle across investments wrappers such as ETFs and Structured Products, as well as, in Portfolio holdings by Hedge Funds and Asset Management. Prior to Machine Readable Filings (MRF), capturing data within Regulatory Filings and Company Reports was difficult to impossible. The MRF data is broad across Sectors and can include Global and or U.S. listed Companies. The data can be run on any timeframe including Monthly and Quarterly for portfolio reconstitution. As you can see Social Market Analytics’ ability to read and classify documents is a powerful tool for your investment process. To learn how Social Market Analytics (SMA) can help you instantly search for Thematic Investing themes using Machine Readable Filings (MRF) SEC Edgar Filings and SMA’s Complex Topic Model engine, please ContactUs@SocialMarketAnalytics.com.

 

/by
, , , , ,

Analysis of Subsequent returns on Short Squeeze Conversations

/by

In the wake of the recent GameStop (GME) short squeeze, we received a lot of inbound questions on social media’s predictive power for short squeezes. Social Market Analytics (SMA) classifies different types of conversations on Twitter with one topic being Short Squeezes. Looking at subsequent returns of stocks with short squeeze conversations leads to different return characteristics based on when the conversation occurred. This blog will explore subsequent Open-to-Close return characteristics when the short squeeze conversation occurs prior to market open and the subsequent Close-to-Open returns and when the short squeeze conversation occurs prior to the market close.

We generate our social media metrics on 4,000+ US Equities every minute 24/7. Our historical data is out-of-sample since 12/1/2011 making it suitable for quantitative back-testing. In this blog we used all equities in the SMA universe. Morning signals are generated at 9:10am US/Eastern, afternoon signals are at 3:40pm US/Eastern.

Tweets discussing short squeezes are identified and classified by our Natural Language Processing and Topic Modeling technology. The number of Tweets discussing a short squeeze in relation to a specific security are aggregated and ranked. In this analysis we are looking at the ‘squeeze-count’ metrics in our data feed, which is the Tweet volume that contains any ‘short squeeze’ keywords about a certain security. The Top 50 securities with the highest ‘squeeze-count’ are equally weighted in the ‘Top_Squeezes’ portfolio. If there is a tie on the 50th security, all securities tied are included in the portfolio. Therefore, portfolio counts could be greater than 50.

The first graph illustrates the subsequent Open-to-Close performance of securities with the most short squeeze terms prior to market open. The blue line represents the cumulative Open-to-Close performance of the top 50 stocks with the highest ‘squeeze-count’. For comparison, the black line represents the S&P 500. These securities in the portfolio significantly underperform intraday. It appears that conversations prior to market open are more about people wanting a squeeze versus it happening. The GME squeeze is marked in the graph.

Cumulative performance is -71.34% with a -2.62 T-Stat and a p-value of 0.0045. These returns are significant at the 1% level.  One caveat to the performance is there is no accounting for hard-to-borrow securities.

The next chart is more interesting. Short conversations occur during the day and are aggregated at 3:40pm US/Eastern. The average subsequent Close-to-Open returns of the top 50 stocks with the highest ‘squeeze-count’ are shown by the blue. As you can see from this chart, stocks with intraday short squeeze conversations move significantly higher overnight.

Cumulative return for this theoretical portfolio is 497% with a 2.89 Sharpe, 5.47 Sortino, 4.33 T-Stat, and an infinitesimally small P-Value. This proves the significance of this portfolio at the 1% level.

SMA Short Squeeze Data Feed is available through a RESTful JSON and XML API or as FTP files. The data can be packaged at different timestamps throughout the day.

Data Sample:

At SMA we aggregate these short squeeze conversations twice a day and make them available to customers. To access this data or many other SMA services, please ContactUS@SocialMarketAnalytics.com

/by
, , , ,

SMA Short Squeeze Data Feed

/by

Since the now famous GameStop (GME) short squeeze we have received inquiries about Twitter’s ability to identify short squeezes early and do we have any products to help identify these stocks prior to the squeeze?   In fact, Twitter is excellent at identifying early discussions of short squeezes.  We have a pre-market open Short Squeeze Alert Report to notify customers of securities with abnormally significant short squeeze discussions.

At Social Market Analytics we have identified a corpus of words and phrases that identify possible short squeezes.  As we read Tweets in real-time, we identify the topic of conversation. For the last year we have been publishing a list of securities with significant number of short squeeze markers prior to market open.  Clients have found this report to be very helpful for risk and security selection.  We can run this report multiple times a day.

To illustrate the power of this report let us look at GameStop.  GME had been an active member of the SMA short-squeeze list prior to the main squeeze in late January.  GameStop was number one on the short squeeze indicator from January 12th through the 31st.  The Blue bars below represent the number of squeeze indicator phrases for GameStop by day.  Short squeeze indicators are taken 20 minutes before the open.  The orange line represents the subsequent open price.  As you can see from the graph, we started indicating strongly on the 14th with very strong indicator on the weekend of the 23rd.  This report also indicated the recent Silver short squeeze.

SMA Short Squeeze Data Feed

To identify potential short squeezes for customers we built the below report.  Again, this report is sent prior to market open to identify high potential short squeeze stocks for the coming day. SMA’s Short Squeeze data feed is available through a RESTful JSON and XML API or as FTP files. The data can be packaged at different timestamps throughout the day. (e.g., pre-open at 9:10am ET, pre-close at 15:40pm ET, etc.)

Short Squeeze Data Feed Field Descriptions

To learn more about this or any SMA product, please ContactUs@socialmarketanalytics.com.

 

/by
, , ,

Thematic Investing using Machine Readable Filings (MRF)

/by

Thematic Investing seeks to identify macro level trends and the underlying investments that stand to benefit from those trends. This Blog highlights ‘Drones’ as a Thematic Investing model example. Drones are now in the news on almost a daily basis and used by listed Companies across Sectors. Companies like Amazon, Allstate, and Goldman Sachs are all ‘users’ of Drones, but who are the Companies that are Drone manufactures or are providing the Technology in Drones? Extracting Themes, such as Drones, from Regulatory Filings and Company Reports is now possible through Machine Readable Filings (MRF) and Social Market Analytics, Inc. (SMA) Complex Topic Model engine. SMA can build a Thematic Investing model using SEC Edgar Regulatory Filings back to 2006 and process the Textual Data programmatically in hours.

Social Market Analytics, Inc. (SMA) has partnered with S&P Global Market Intelligence on ‘Machine Readable Filings’ (MRF). Machine Readable Filings is the first product to provide parsed textual data of SEC Edgar Regulatory Filings at the Total Document, Part, Item, Sub-section, and Notes level with historical baselines back to 2006. The Textual Data, Sentiment NLP, and Word Count values are provided by SMA’s patented technology. Extraneous information such as page numbers, images, and tables are removed. SMA is currently developing the same structured data values on International Company Reports which will be released in Q1 2021.

Machine Readable Filings (MRF) used in conjunction with SMA’s ‘Complex Topic Model’ engine allows for the programmatic curation of Thematic Investing themes.

SEC regulatory filings are formal documents reported to the U.S. Securities and Exchange Commission (SEC) that contain important information about companies such as financial statements, forward looking statements, risks associated with the company, etc. Every U.S. publicly traded company is required to submit regulatory filings to the SEC.

Companies discuss their forward focus in these documents, and you can glean a lot about a company’s direction or risks by the topics they discuss in filings. With SMA’s topic model capabilities you can quickly scan documents to look for curated topics. If you are searching for companies investing in a specific topic, this functionality makes the process much more efficient. As an example, if you want to find companies with negative exposure to Covid-19 you can search for Covid-19 mentions in the Risk Factors section of a 10-K or 10-Q. If you would like to find a company’s strategy for handling their business during Covid-19, you may want to search in the Management Discussion & Analysis section.

In this blog we will be illustrating the power in SMA’s topic searching by identifying companies with the most mentions of the topic ‘Drones’ in their filings. The Topic Model data extraction across ~1.4M documents going back to 2006 only takes a few minutes. The MRF textual data structure also enables the creation of baselines.

For this “Drone” Thematic blog we:
• Broke the ‘Drones’ thematic Topic Model down by Company, Filing type, and Sector. The data set covers filings from January 2006 – October 2020
• The ‘Drones’ Topic Model captures every relevant ‘Drone’ word/phrase including synonyms (inclusion) within SEC filings
• We exclude mentions of Drones where it does not apply to the Topic Model, e.g. UAS Drone Corp. mentioning their own name in a filing
• The data is run across all SEC Edgar Filings including 6-K ADR. Some of the captured Companies are overseas. Filing types can be excluded as well.

Machine Readable Filings tracks over 20 filing types. The distribution of filings containing ‘Drone’ mentions is below. Majority of the documents containing ‘Drone’ mentions since 2006 tend to be 10-Ks and 10-Qs because they are the most detailed documents. However, in 2020, the amount of 6-Ks mentioning drones almost triple from 2018, making 6-Ks the most hit document for the year so far. This suggests that drone usage for companies could be growing faster internationally than in the United States.

As illustrated by the chart below, ‘Drone’ mentions have grown more than 500% since 2006 even with just 10 months of 2020 documents included. Unique companies discussing drones in their filings has grown 400% from 40 companies in 2006 to over 160 so far in 2020.

*A mention is defined as a unique paragraph that contains one or more words from the Drones Topic Model in any SEC Filing.

Below are the top companies mentioning drones in their regulatory filings. Many of these companies are not the typical Fortune 500 companies that mention drone usage as a buzz word. The companies most frequently hit are companies that build drones and work with the drone AI technology. Amazon, one of the more media mentions of Companies ‘using’ drones through Amazon Prime Air, only mentions a drone related term once in both their most recent 10-K and 10-Q. Prime Air is not mentioned in either document. When making a thematic portfolio for drones, it would not make sense to include Amazon because Prime Air is not a central enough part of their business to mention in their filings and, as of now is not manufacturing drones.

The companies mentioning represent a variety of sectors. The top two sectors are information technology and industrials. Those are the companies that build and develop the technology for drones and use drones for delivery. Other sectors use drones in a variety of ways. For example, Allstate, in the financial sector, uses drones to analyze crash sites for their insurance claims. Goldman Sachs uses Drones in its investment banking arm. It makes sense to include companies from a variety of sectors when creating a thematic portfolio while at the same time having a higher weight on the more relevant sectors.

Today’s Macro and Technology Themes are unprecedented and cover a multitude of emerging “Themes” from Drones to ESG to Genome Research to Biometrics, etc. The investment potential is unlimited. Thematic Investing is becoming a powerful vehicle across investments wrappers such as ETFs and Structured Products, as well as, in Portfolio holdings by Hedge Funds and Asset Management. Prior to Machine Readable Filings (MRF), capturing data within Regulatory Filings and Company Reports was difficult to impossible. The MRF data is broad across Sectors and can include Global and or U.S. listed Companies. The data can be run on any timeframe including Monthly and Quarterly for portfolio reconstitution. As you can see Social Market Analytics’ ability to read and classify documents is a powerful tool for your investment process. To learn how Social Market Analytics (SMA) can help you instantly search for Thematic Investing themes using Machine Readable Filings (MRF) SEC Edgar Filings and SMA’s Complex Topic Model engine please ContactUs@SocialMarketAnalytics.com.

/by
, , ,

Machine Readable Filings (MRF): 8-K Sentiment Alpha

/by

Social Market Analytics, Inc. (SMA) has partnered with S&P Global Market Intelligence on ‘Machine Readable Filings’ (MRF). Machine Readable Filings is the first product to provide parsed textual data of SEC Edgar Regulatory Filings at the Item, Section, Sub-section, and Notes level with historical baselines back to 2006, along with sentiment values provided by SMA’s patented Natural Language Processing (NLP) technology. Extraneous information such as page numbers, images, and tables are removed. SMA is currently developing the same structured data with sentiment values on International Regulatory Filing Reports which will be released in Q4 2020.

SEC regulatory filings are formal documents reported to the U.S. Securities and Exchange Commission (SEC) that contain important information about companies such as financial statements, forward looking statements, risks associated with the company, etc. Every U.S. publicly traded company is required to submit regulatory filings to the SEC.

One type of regulatory filing supported in the Machine Readable Filings (MRF) product is Form 8-K. This broad form is used to notify investors in public companies of specified events that may be important to shareholders or the SEC. This is one of the most common types of forms filed with the SEC. After a significant event like bankruptcy or departure of a CEO, a public company must file a Current Report on Form 8-K within four business days to provide an update to a previously filed 10-Q or 10-K.

We have found predictive power in virtually all MRF analysis across different filing types. Please reach out to us for white papers and research on other document types and sections. For this blog we look at the predictive power of the sentiment values of the 8-K filings. Social Market Analytics’ patented NLP technology reads and assigns fine-grain sentiment scoring to the 8-K document, down to sub-section level. The analysis below demonstrates the predictive power of sentiment values from 8-K textual data. These are not trading strategies by themselves but do illustrate the predictive power of the 8-K sentiment data.

Companies publish an average of eight 8-K filings per year.  For our analysis we looked at approximately 500,000 documents.

For our test:

  • We are analyzing the relationship between sentiment of a company’s 8-K filing and subsequent stock price movement over the next week.
  • If a company releases an 8-K, a position is entered on the Close of the same day of the release and sold 5 days later.
  • The stock universe we tested on includes all companies that filed 8-Ks after 1/1/2010 that mapped to SMA’s pricing source with a Price > $5 filter.

Quintiles:

Stock returns are bucketed into Quintiles daily depending on the value of Average Sentiment of the 8-K filings (Quintile 1 [Q1] has the lowest 20% of sentiment values. Quintile 5 [Q5] has the highest 20% of sentiment values). All quintiles display the Covid-19/Election/Trade War drawdown in March 2020.  As you can see there is a relationship between 8-K filing sentiment and subsequent price returns. Q5 which represents the highest 20% of 8-K sentiment scores significantly outperforms all other quintiles. And we see the sentiment data is monotonic. Included is also the spread between the top and bottom quintiles (Q5 – Q1). This spread is proven significant a 99% confidence level with a T-Statistics of 4.65 and a p-value of 0.00000175.

 

Thresholds:

In this part of the analysis, stock returns are bucketed into Long and Short buckets depending on the sentiment threshold set. In the following example, 8-K filings with an Average Sentiment greater than 1.5 are in the Long bucket and filings with an Average Sentiment less than 0 are in the short bucket. We see companies whose 8-Ks have positive sentiment greater than 1.5 go on to significantly outperform while negative ones tend to underperform. An equally weighted theoretical Long/Short portfolio generates a solid return and nice Sharpe ratio, even during the March 2020 drawdown.

 

Until now there has been no real way to analyze text in corporate filings. SMA’s Machine Readable Filings product allows for rapid identification of the sentiment as well as the predictive power of corporate disclosures. For more information or to start a trial ContactUs@SocialMarketAnalytics.com.

/by
, , , , ,

Crypto Currency Market Index

/by

We are often asked about broad market indicators for different asset classes. Recently Koby finished some research on a Crypto Currency broad market indicator that is predictive of future Bitcoin prices. We call this indicator the Crypto Market Score (CMS).

For the CMS we take once a day snapshots of all coins with a sentiment value. Our universe of covered coins consists of coins that have at one point been in the top 200 market cap of all cryptocurrencies.  Our universe currently stands at approximately 600 coins. We use our patented NLP to read and assign sentiment scores to each Tweet.  Each Tweet is scored from -1 to 1 out to four decimal places.  Because each Tweet is scored independently, we are creating a Tweet count weighted sentiment value.  We aggregate the sentiment from all coins over the last 24 hours by date to aggregate to a daily sentiment value.

We calculate a 14-day rolling sum of the sentiment values and compare it to a 60-day baseline to generate a CMS. Our CMS is a Z-Score with a 14-day summation of Tweet scores for all coins in our universe and compared to a 60-day baseline.

We compared the CMS to forward weekly bitcoin returns. There are a total of 1203 days of data with the CMS ranging from -4.00 to +4.00. In the table below Positive (539) + Negative (664) = our test period 1203 days.

Below are the general return characteristics on the Crypto Market Score. The average 7-day return for BTC was 2.028% for our test period.

As you can see from the return table, days the CMS is positive Bitcoin returns 2.367% over the next 7-days. When the CMS is negative, Bitcoin returns 1.754% over the next 7-days. As the CMS rises the subsequent return of Bitcoin increases; reaching a high of 14.826% for the 8 instances when our CMS is greater than 3. When the CMS is negative, Bitcoin 7-day returns are less than average or in some cases negative. These relationship between the CMS and forward 7-day Bitcoin returns is a positive linear relationship as shown by the monotonic nature of the return characteristics. To learn more about our Market Sentiment Score or Social Market Analytics ContacUS@SocialMarketAnalytics.com.

/by
, , , , , , ,

Filtering Relevant Tweets Using Topic Models

/by

At Social Market Analytics we use proprietary techniques to return the most accurate Twitter volume for topics of interest.  First, we use a topic model not just $Ticker.  Most vendors use the CashTag concept to identify securities.  At SMA we believe only using CashTag excludes a lot of valuable conversation.  We return higher volume and cleaner conversations because we use a machine learning rules based system to return all conversations about a security that are not tagged with $Ticker.  In the diagram below we return about 500 extra Tweets a day for Tesla Motors versus just $TSLA.  Our topic model evolves with the conversation over time.

After applying our topic model filter we additionally filter Tweets based on our proprietary account validation metrics.  Only Tweets from our SMA approved accounts are included.  In the below example, there are about 500 Tweets for Tesla Motors from the certified accounts per day.   These accounts pass our multi-step algorithm.  One metric we use is weighted accuracy over time.  For example, when a Twitter account is bullish on a security what percentage of time does that security subsequently move higher.

Below is a visualization of SMA Twitter filtering process for the Tesla topic.

Below is a time series of Tesla Topic model versus $TSLA with the additional filter for SMA certified accounts.

As you can see from the charts SMA’s proprietary technology provides the truest view of the each securities topic model.  To learn more about our technology or receive a sample data set ContactUs@socialmarketanalytics.com.

 

/by