idw – Informationsdienst Wissenschaft
Nachrichten, Termine, Experten
Nachrichten, Termine, Experten
Ever since the launch of ChatGPT as the fastest-growing consumer app to date, investors worldwide have witnessed substantial growth in the value of technology companies that incorporate artificial intelligence (AI). Meanwhile, other entities have been accused of engaging in AI washing, i.e. overstating their AI focus. This environment has highlighted the difficulty of identifying which companies genuinely invest in AI-based processes—and which merely reference the topic to attract media or market attention and inflate valuations. A new study from Constructor University provides some much-needed guidance on the topic—and created new ETFs, based on its findings.
In his recent paper “Quantifying a Firm’s AI Engagement: Constructing Objective, Data-Driven, AI Stock Indices Using 10-K Filings,” published in the journal Technological Forecasting and Social Change, Constructor University Prof. Lennart Ante, alongside co-author Prof. Aman Saggu from Mahidol University International College, investigated how to increase transparency in the classification of AI-centric firms. By analyzing the content of annual 10-K reports of NASDAQ-listed firms, the article aims to limit undue speculation when organizations disclose AI initiatives in official documents. This approach provides an alternative to subjective practices in which stocks are labeled “AI-related” based on vague descriptions rather than objective practices based on solid evidence.
According to Professor Ante, who recently joined Constructor University as Associate Professor of Entreprenurial Finance, the insights provided by the paper range from practical reasons to economic interests. “Interest in AI has spread beyond software giants to a wide spectrum of industries from manufacturing automation to healthcare diagnostics,” Professor Ante said. He said that these developments have simultaneously created excitement and confusion. “On the one hand, AI can refine business decisions, accelerate product development, and reshape competitive dynamics. On the other hand, uncertainty arises when investors seek to separate companies that contribute to genuine AI innovation from those that provide AI references as a marketing strategy.” Without factual clarity regarding the scope and role of AI, enterprises can quickly be overvalued, opportunities can be missed and unnecessary financial risks incurred. The study responds to this dilemma by establishing systematic metrics for evaluating corporate AI engagement based on disclosures made to regulators.
Professors Ante and Saggu analyzed corporate filings from 3,395 NASDAQ-listed firms between 2011 and 2023 to compile references to AI, tallying the frequency and recording how prominently each company mentioned AI-related concepts. This procedure generated multiple measures and identified whether and how often a filing contained AI terminology. Based on these measures, multiple stock indices were created. To verify the degree of AI in the indices, the study conducted an event study of the launch of OpenAI’s ChatGPT in November 2022, a milestone that prompted widespread public interest in AI. Stocks classified as AI-oriented by the new metrics reacted more strongly in terms of “abnormal returns,” suggesting that the market perceived them as key AI participants. Further statistical testing found that high AI scores among firms correlated with stronger share performance in the aftermath of ChatGPT’s debut. “We tested our framework around the ChatGPT launch because it was a defining moment for AI awareness,” explained Prof. Ante. “The pronounced stock reactions underscored the market’s preference for companies with verifiable AI activity—and confirmed that our measures reflected AI engagement.”
After confirming their classification method, the researchers compared their newly constructed indices to 14 exchange-traded funds (ETFs) advertised as AI-themed using a variety of measures. The results showed that the data-driven indices matched or even surpassed the ETF performance across several timeframes. The indices constructed did not exhibit higher volatility while also delivering favorable returns. This finding indicates that an objective, data-driven framework could provide an effective alternative to thematic investing. By referencing corporate filings, it avoids the subjective criteria or ambiguous buzzwords often found in thematic funds. The study’s evidence suggests that AI references documented in mandatory reports may be a reliable indicator of genuine technology adoption.
“By introducing a statistical approach to classify the AI engagement of firms, we’re aiming to boost fairness and clarity in thematic investing,” said Prof. Ante. Beyond its immediate relevance to AI, the approach can be adapted to other technologies or specialized themes. Analysts interested in identifying firms that engage with blockchain, augmented reality, or climate-related innovations could use similar techniques. The new indices also open the door to heightened market competition. Investors, fund managers, and research institutions can use these transparent metrics to develop AI-themed portfolios with fees potentially lower than those attached to some thematic ETFs. Such developments might encourage more detailed disclosures in corporate annual reports, as companies may recognize the direct impact of thorough AI reporting on market perceptions and index inclusion.
Dr. Lennart Ante, presse@constructor.university
Aerial image of Constructor University Campus in Bremen, Germany (source: Constructor University)
Merkmale dieser Pressemitteilung:
Journalisten, Wirtschaftsvertreter, Wissenschaftler, jedermann
Informationstechnik, Wirtschaft
überregional
Forschungs- / Wissenstransfer, Forschungsprojekte
Englisch
Aerial image of Constructor University Campus in Bremen, Germany (source: Constructor University)
Sie können Suchbegriffe mit und, oder und / oder nicht verknüpfen, z. B. Philo nicht logie.
Verknüpfungen können Sie mit Klammern voneinander trennen, z. B. (Philo nicht logie) oder (Psycho und logie).
Zusammenhängende Worte werden als Wortgruppe gesucht, wenn Sie sie in Anführungsstriche setzen, z. B. „Bundesrepublik Deutschland“.
Die Erweiterte Suche können Sie auch nutzen, ohne Suchbegriffe einzugeben. Sie orientiert sich dann an den Kriterien, die Sie ausgewählt haben (z. B. nach dem Land oder dem Sachgebiet).
Haben Sie in einer Kategorie kein Kriterium ausgewählt, wird die gesamte Kategorie durchsucht (z.B. alle Sachgebiete oder alle Länder).
