| April 23, 2014
Companies do not operate in isolation and are dependent on other suppliers and customers for their operations.
In this study, we use data from FactSet Relationships to build a monthly stock selection model centered on customer and supplier companies in the Russell 3000 universe. We first study customer‐to‐supplier momentum and vice versa in modified FactSet Revere industry groups. We find that the customer‐to‐supplier lead‐lag effect (or vice versa) is not only contingent on the industry group in which the underlying company exists, but is also dependent on the directionality of the source of the information which changes over time.
We then build a dynamically‐updating model, which incorporates our empirical observations, to find that our Sharpe ratio drastically increases to 2.2 with annualized returns of 11%. Finally, we test the limited investor attention theory by studying how media attention on an individual firm itself impacts economically‐linked price momentum. Incorporating asymmetrical media reaction improves the annualized returns on the underlying strategies to over 15% with a Sharpe of 1.5. Our results are consistent even when restricting our universe to companies with a market capitalization of over $1 billion dollars.
Over the last couple of years, numerous studies have explored the relationship between customers and their associated suppliers. From Cohen and Frazzini’s seminal paper (2008) on the lead‐lag relationship between customers and their associated suppliers (and vice versa) to the more recent paper by Cahan and Xing (2013), in which they incorporate a vector auto‐regressive model to identify the clusters of economically‐related stocks to forecast future returns, supply chain relationships and a localized momentum effect between economically‐linked stocks have become an important factor in stock selection models.
In this paper, we aim to build on the research that companies do not operate in isolation. They are economically‐linked to other industries and companies. More importantly, the direction of the lead‐lag effect must take into account the industry in which they operate. For this entire study, we restrict our analysis to the Russell 3000 universe. In future iterations, we aim to expand this universe to include the entire U.S. tradable universe, as well as European stocks.
The first question we address in this study is: “Do customer returns drive supplier returns or is the inverse effect stronger?” In Section (3) we look at companies operating within the Russell 3000 universe, and we find that the customer‐to‐supplier effect is more pronounced than the supplier‐to‐customer effect across the entire universe. In Section (4) we analyze both approaches in major industry groups and find that the return profiles and performance of both effects vary widely by industry. In fact, for most of the industries in which the customer‐to‐supplier effect is stronger, the supplier‐to‐customer effect is non‐existent. We incorporate our new information to create a hybrid model that takes into account the widely differing return profiles in certain industries, and the changing lead‐lag nature of information from suppliers or customer portfolios. The resulting strategy shows drastically improved return profiles and information ratios. Finally, in Section (5), we look at how media attention surrounding companies with linked portfolios impacts the future performance of those companies. We build double‐sorted portfolios incorporating media attention, and we find that significant media attention improves the short signals while lesser media attention improves our longs.
We use FactSet Relationships for our study, specifically for the customers and suppliers. Companies are economically‐linked to other companies and industries. The entire picture, however, might not be obvious by simply looking at a single stock’s own disclosure. In a networked economy, many companies are disclosing relationships to each other in both directions, and one can argue that no single analyst can fully analyze the ecosystem around a company without looking at all other companies and their disclosures. FactSet Relationships analyzes all companies under coverage and stitches their respective disclosures together to gain a complete perspective on how companies are doing business together.
A key distinction made by FactSet Relationships is in recognizing that the customer‐supplier relationships are transitive in nature; for example, if Company A supplies its goods or services to Company B, this implies that Company B is a customer of Company A, regardless of whether Company A explicitly disclosed Company B as a customer. We use the bidirectional nature of information disclosure as captured within FactSet Relationships to construct our customer and supplier portfolios at each cross section of time. An example of the final dataset used in portfolio construction (for supplier portfolios) is depicted below:
Please use your business email. If you don't have one, please email us at firstname.lastname@example.org.
We will process your personal data with the purpose of managing your personal account on
RavenPack and offering our services. You can exercise your rights of access, rectification,
erasure, restriction of processing, data portability and objection by emailing us at email@example.com. For more information, you can
Your request has been recorded and a team member will be in touch soon.
RavenPack is introducing a multidimensional approach to news analytics that captures the evolving complexity of news, enabling traders to build more dynamic and transparent strategies.
New research using RavenPack Job Analytics to quantify tech adoption in hiring posts proves that companies hiring for novel tech skills outperform peers from an investment perspective.
Our multi-asset allocation strategy based on inflation nowcasts and RavenPack sentiment analytics outperforms the S&P 500 and stabilizes volatility. With real Sharpe ratios of up to 1.25 and a risk control approach, it delivers a compelling inflation hedge for investors.