[Paper Review] Facebook drives behavior of passive households in stock markets

Milla Siikanen, Kęstutis Baltakys+4 more|RePEc: Research Papers in Economics|Sep 21, 2017

FinTech, Crowdfunding, Digital Finance4 citations

TL;DR

This study examines how Facebook activity influences investor trading behavior using unique Finnish investor transaction data for Nokia stock. It finds that passive households and nonprofit organizations significantly adjust their buy/sell decisions based on Facebook data, while sophisticated financial institutions remain unaffected, suggesting social media acts as a biased information channel for less informed investors.

ABSTRACT

Recent studies using data on social media and stock markets have mainly focused on predicting stock returns. Instead of predicting stock price movements, we examine the relation between Facebook data and investors' decision making in stock markets with a unique data on investors' transactions on Nokia. We find that the decisions to buy versus sell are associated with Facebook data especially for passive households and also for nonprofit organizations. At the same time, it seems that more sophisticated investors---financial and insurance institutions---are behaving independently from Facebook activities.

Motivation & Objective

• To investigate whether Facebook data influence individual investor trading decisions, particularly distinguishing between sophisticated and passive investors.
• To analyze how different investor groups—households, nonprofits, financial institutions—respond to social media activity related to Nokia stock.
• To determine if Facebook serves as a secondary, potentially biased information source that affects less sophisticated investors' behavior.
• To explore the role of social media as an alternative information channel compared to official exchange announcements.

Proposed method

• The study uses a unique dataset of investor-level shareholding registrations from Euroclear Ltd., covering all Finnish investors from 2010 to 2016.
• Trading dates are reverse-engineered from registration dates using T+3 and T+2 settlement conventions to enable weekly aggregation.
• Facebook data—including posts, comments, likes, and shares—are collected and linked to investor transactions for Nokia stock.
• Multivariate regression models are applied to test the association between Facebook activity and investor decisions to increase or decrease shareholdings.
• The analysis distinguishes between investor types: households (passive vs. active), nonprofit organizations, and financial/institutional investors.
• The study controls for potential reverse causality by noting that trading position changes are not publicly disclosed in real time, making it unlikely that companies post in response to investor activity.

Experimental results

Research questions

• RQ1 Do Facebook data influence the buy/sell decisions of passive households in the stock market?
• RQ2 How do nonprofit organizations and financial institutions differ in their responsiveness to Facebook activity?
• RQ3 Is there a correlation between investor sophistication and sensitivity to social media information on Facebook?
• RQ4 To what extent does Facebook serve as a substitute or complement to official company announcements in shaping investment behavior?
• RQ5 Are the observed associations between Facebook activity and trading decisions driven by causality or reverse causality?

Key findings

• Passive households show a statistically significant association between Facebook activity and their trading decisions, particularly for posts, comments, and likes.
• Nonprofit organizations also exhibit a strong and significant relationship between Facebook data and their buy/sell behavior.
• Financial and insurance institutions do not show any significant association with Facebook data, indicating they trade independently of social media activity.
• Active households, despite being part of the broader household category, do not show significant responsiveness to Facebook data, suggesting that activity level moderates social media influence.
• The signs of regression coefficients for Facebook variables are consistently positive for passive households and nonprofits, indicating that increased Facebook engagement is linked to increased trading activity.
• The study finds no evidence of reverse causality, as real-time trading position changes are not publicly available, making it unlikely that companies post in response to investor movements.