Majid Majzoubi
The findings are contained in the paper "Going Beyond Optimal Distinctiveness: Strategic Positioning for Gaining an Audience Composition Premium," which was published in the Strategic Management Journal. The paper was co-authored by Majid Majzoubi, assistant professor in the strategic management area at the Schulich School of Business, and Eric Yanfei Zhao, associate professor of strategy and entrepreneurship at the Kelley School of Business, Indiana University.
The study looked at how companies can strategically position themselves to gain more attention from audiences with positive predispositions toward them - what the researchers call an "audience composition premium."
The researchers state that "recommender systems" - a family of machine learning algorithms - can be used to discern an audience member's idiosyncratic predisposition toward a firm, even before they have evaluated it. The insights and predictions gained from these machine learning models can then be used to optimize a firm's positioning strategy.
"The study's predictions are tested in the context of security analysts' investment recommendations of publicly listed U.S. firms," said Majzoubi.
The study examines 152,312 investment recommendations issued by security analysts for public U.S. firms between 1997 and 2018, and uses topic modelling to analyze the textual content of firms' 10-K filings and 297,931 earnings call transcripts. "The empirical analysis we conducted strongly supports the study's predictions," said Majzoubi.
The researchers have publicly shared the code for their models and empirical analysis here.
Adds Majzoubi: "To gain higher aggregate evaluations, a firm should influence its audience composition so that it is evaluated primarily by audiences with favourable predispositions toward it - and evaluated less by those with unfavorable predispositions."
This story was originally featured in YFile, York University's community newsletter.
