In this month’s research picks, INSEAD faculty offer insights on improving asset pricing models and sales forecasting models, as well as how the extent of differences among banks affects financial stability. On the organisational front, they examine how artificial intelligence (AI) could impact strategic decision-making processes in firms and ways to prevent relational overload in organisations where employees can easily access colleagues and information.
1. The effects of bank heterogeneity on financial stability
In a study, Lin Shen and her co-authors* analyse the interconnectedness and fragility of financial systems. They show that bank run problems within banks and fire-sale spillovers across banks are mutually reinforcing and can lead to a systemic crisis.
The researchers highlight that increasing bank heterogeneity, or the extent of differences among banks, reduces the fragility of financial systems. To achieve this, regulators may reduce asset commonality, bank-specific disclosures, and broad-based policies on asset purchases and liquidity requirements.
* Itay Goldstein from The Wharton School, Alexandr Kopytov from University of Rochester, and Xiang Haotian from Peking University
2. Preventing relational overload in permeable organisations
As rapid organisational and technological change reduce boundaries within workplaces, employees are gaining unprecedented access to new people and information. While this can increase opportunities for collaboration, it can also heighten the risk of attention overload. In highly permeable organisations, members not only have to choose what to attend to, but also whom to attend to – a problem Mark Mortensen and his collaborators* term “relational overload”.
Mortensen et al reveal how organisations can reduce relational overload without sacrificing productive work relationships by strengthening shared norms and mitigating potential misalignments.
*Ethan S. Bernstein from Harvard Business School, Pranav Gupta from University of Illinois Urbana-Champaign and Paul Leonardi from University of California Santa Barbara
3. Improving accuracy of asset pricing models
Traditional asset pricing models of firm’s behaviour often assume that a company’s physical capital – its tangible assets like factories or equipment – is the key factor in determining its stock value. However, research by Frederico Belo and his co-authors* challenges this view. They argue that earlier studies overlooked how stock returns evolve over time, leading to weaker tests of pricing models.
By focusing on the dynamic patterns of stock returns, their new study sets a higher standard for evaluating investment-based models. The findings suggest that to better capture the complexities of stock behaviour, pricing models should include additional inputs, such as intangible capital or human capital, that can contribute to firm value.
*Yao Deng from University of Connecticut and Juliana Salomao from University of Minnesota
4. Enhancing sales forecasting model with search data
Forecasting sales is an essential marketing function. Due to data limitations, data on the full set of competitors are rarely used when forecasting sales. However, that was before the emergence of online search data.
In a recent study, INSEAD’s Abhishek Borah and Rutz Oliver from University of Washington propose a dynamic forecasting model that utilises all available competitive textual search data in a market and addresses the inherent statistical issue when including a large number of competitive effects. Using data from the American automobile industry over a 12-year period, they show that the model forecasts sales better than models that do not include complete competitive search data.
5. How AI affects strategic decision making
Could AI augment strategic decision-making process in organisations? In a study, Hyunjin Kim and her collaborators* analysed the key cognitive processes underlying strategic decision-making, namely search, representation and aggregation. They suggest that AI has the potential to enhance the speed, quality and scale of strategic analysis, as well as enable new approaches such as virtual strategy simulations.
However, AI’s impact on firm performance ultimately depends on competitive dynamics as the technology develops. To shed light on this, the researchers propose a framework connecting AI use in strategic decision-making to firm outcomes.
*Felipe Csaszar from University of Michigan and Harsh Ketkar from University of Texas at Austin
“INSEAD, a contraction of “Institut Européen d’Administration des Affaires” is a non-profit graduate-only business school that maintains campuses in Europe, Asia, the Middle East, and North America.”
Please visit the firm link to site
