# Herding Unmasked: Insights into Cryptocurrencies, Stocks and US ETFs

**Authors:** An Pham Ngoc Nguyen, Thomas Conlon, Martin Crane, Marija Bezbradica  
**arXiv:** [2407.08069](https://arxiv.org/abs/2407.08069)  
**Date:** 2024-07-10  
**Summary page:** https://www.summarizepaper.com/en/paper/2407.08069v1/

## Key points

- Researchers An Pham Ngoc Nguyen, Thomas Conlon, Martin Crane, and Marija Bezbradica study herding behavior among investors across different types of investment vehicles.
- The study utilizes the Cross Sectional Absolute Deviation (CSAD) model to analyze herding patterns within stocks, US ETFs, and cryptocurrencies individually.
- Similarities in herding behavior are found between stocks and US ETFs, with distinct differences observed in cryptocurrencies.
- Graph-based techniques like Minimum Spanning Tree (MST) and Louvain community detection are used to explore herding behavior at a granular level by partitioning the combined dataset into smaller subsets based on asset similarity.
- Herding behavior is present across all types of investment vehicles at a subset level even if not evident at a broader superset level. Specific events can trigger herding behavior within subsets of assets.
- Understanding patterns of herding behavior within different asset classes is crucial for constructing an effective investment strategy tailored to individual preferences and risk tolerance levels.

## Summary

In the study "Herding Unmasked: Insights into Cryptocurrencies, Stocks and US ETFs," researchers An Pham Ngoc Nguyen, Thomas Conlon, Martin Crane, and Marija Bezbradica delve into the phenomenon of herding behavior among investors across different types of investment vehicles. By utilizing a recent dataset that captures significant events in recent years, the study aims to contribute valuable insights to the existing literature on herding behavior. The researchers employ the Cross Sectional Absolute Deviation (CSAD) model to analyze herding patterns within stocks, US ETFs, and cryptocurrencies individually. They discover similarities in herding behavior between stocks and US ETFs while noting distinct differences when compared to cryptocurrencies. Building upon this initial analysis, the study extends its investigation by combining all three types of investment vehicles. To further explore herding behavior at a more granular level, graph-based techniques such as Minimum Spanning Tree (MST) and Louvain community detection are utilized to partition the combined dataset into smaller subsets based on asset similarity. The findings reveal that herding behavior is present across all types of investment vehicles at a subset level even if it may not be evident at a broader superset level. Moreover, the researchers observe that specific events can trigger herding behavior within subsets of assets. Overall, this comprehensive study sheds light on how herding behavior manifests across various investment vehicles and highlights the importance of considering these dynamics when constructing an effective investment strategy. By understanding and recognizing patterns of herding behavior within different asset classes, investors can make informed decisions tailored to their preferences and risk tolerance levels.