To what extent do amateur investors trust social media?

By Eric Tham
Posted on July 19, 2018

Household participation in the stock market has changed considerably since the 1970s. Back in the days of Fisher Black and his journal article Noise, household investors were perceived as noise traders. Compared to professional investors, household investors lacked sophistication and access to trustworthy financial information.

The rise of Big Data and the huge popularity of social media in the last two decades has changed all this. Various social media investing platforms (like SeekingAlpha and, StockTwits) have arisen to level the playing field for household investors by increasing their financial literacy. How has this affected the decision-making process of the household investors? To what extent do they trust social media for investing purposes?

All about beliefs

I use Prospect Theory to answer the above questions. Prospect Theory (a seminal paper written by Daniel Kahneman and Amos Tversky) is a descriptive model of decision-making under risk. A key principle of this theory is that an investor derives value from investment losses or gains to a reference level.

Prospect theory states that people go through two different stages when making a decision between choices that carry uncertainty. First, there is the editing phase. Here people simplify the complicated decisions into simpler ones by just making a belief of gains versus losses per option. In the second valuation phase, people examine the edited options and act on the one they believe has the highest value.

The success of social media and Big Data allows to distinguish between the two phases. Investors reflect their beliefs on the stock market by writing about it on social media. This is the first editing phase. However, they might not necessarily act on their beliefs; they may have just lost their job or need to save their money to keep their kids in college. This gives insights on the second valuation phase. These two phases are distinct although dependent on each other.

AI meets NLP

I use Thomson Reuters MarketPschy indices (TRMI) to determine the investors’ sentiment beliefs on the social media. TRMI uses artificial intelligence to obtain these beliefs from various social media platforms around the world, applying Natural Language Processing (NLP) to the textual postings on these platforms.

I also use another widely cited index, the Baker-Wurgler sentiment index. Unlike TRMI, this sentiment index is obtained from trading activities whereby investors actually put their money where the mouths are. The Baker-Wurgler index thence encompasses both the editing and valuation phases unlike the TRMI index that reflects only the household investors’ beliefs.

I test for the hypothesis that household investors increasingly trust social media for their investment decisions. To do this, I apply a decomposition method to the TRMI and Baker-Wurgler index data that allows separating investors’ beliefs from their fundamental financial valuations.

Increasing trust in social media

Belief or trust as a measure is hard to quantify. Yet, it is the very fabric that holds the economic system together. This belief is the commonality between the social media sentiment and the Baker-Wurgler index, since only the belief editing phase is found in both. The results show that a time series of this belief significantly increases in value from 1998 to 2016, meaning that trust in social media for financial investments has increased among households during the same period. Additional analyses also show that this pure belief has a significant power to explain stock factor returns. This suggests that household investors are actually acting on these beliefs to trade on the stock market or what is the same: households are increasingly trusting social media information to make their decisions about financial investments.

This has important implications. Trust in social media seems to lower the costs for households to acquire financial knowledge or financial advice. It could then increase household participation rates (especially amongst the lower income families) and their wealth allocation in the stock markets. Without this social media trust, household investors would just invest more into the savings markets as per status quo.

In the wider economy, it may cause gyrations in the stock markets since household investors are more subjected to psychological biases, as typified by the Chinese stock market crash of 2016, which was brought about by household investors.

Are you curious to know more about these research outcomes? Keep a close look to the TFI website and read the full report that will be published in August 2018.

Eric Tham is a PhD candidate in Finance at Edhec Business School. His research interests are in behavioural and household finance as relating to asset pricing. His PhD supervisor is Prof. Laurent Calvet. The research project that was funded by TFI last February looks at how social media influence household investors’ decision making.