What drives stock prices up and down? The well-known 'efficient market hypothesis' (EMH), which has been widely accepted for decades, claims that markets are driven by rational calculations. Traders, on the other hand, are people, and people are impacted by emotions. Do these feelings influence the stock markets? It's tough to research this topic because people's emotions aren't discernible. However, recent data suggests that the kind of music investors listen to may have a significant influence on the performance of the stock market.
Spotify provides aggregated listening data from throughout the country, as well as an algorithm that categorises each song's positivity (the degree to which the song makes one feel happy) or negativity (the degree to which the song makes one feel sad). Based on these inputs, researchers calculate ‘music sentiment’ - a measure that expresses the country’s sentiment based on the positivity of the songs the citizens listen to.
Researchers have found conclusive evidence to substantiate the concept. Dr. Philip Maymin, assistant professor of finance and risk engineering at NYU’s Polytechnic Institute, found that the public’s song choices may be linked to expected movements within the US equity-based exchanges. People favour softer, calmer musical beats when they predict a more volatile stock market in the future months and faster, livelier music when they anticipate calmer markets, according to Dr. Maymin's results. He came to these conclusions after comparing the standard deviation of returns for the S&P 500 index to the average yearly beat variance in songs tracked by the Billboard Top 100.
Professor Alex Edmans of London Business School along with three co-authors compiled data on the average positivity of songs based on 500 billion streams of 58,000 songs on Spotify by consumers in 40 countries. The data was then compared to the performance of each country's national stock exchange during the same time. The outcome: Equity-based investing avenues tend to outperform when people listen to happy songs.
This research was supplemented by tests run on government bonds. When people are optimistic, they tend to buy fewer bonds, as in comparison to stocks they are less risky, causing bond prices to fall. In investing spheres where people listened to happy songs, they did buy fewer bonds indicating that happy music means a higher average share price.
So, why is this so?
Essentially, music is seen as a proxy for consumer sentiment. This is useful for two reasons, the first being that songs, as well as the mood they portray, are easy to gauge at the individual level, the second being that recent technological developments have made it easier to harvest big data, as the music industry, and by extension it’s consumers have consolidated.
What this leads to is that music becomes in itself a great tool to not only gauge emotions in times devoid of large events. For example, should India win the Cricket World Cup, it would perhaps be intuitive to all that the average sentiment will be one of happiness. Yet, such sentiment, which in turn forms market sentiment, is harder to understand in times of absolute normalcy. This is where music becomes incredibly important, with its easy characterizations of what belongs to which emotions. At the same time, looking at the data for popular songs, such as Spotify top 100 or billboard top music charts, provides a monumental amount of scope, which in turn bolsters the accuracy of the conclusions being drawn.
It is important to note that the researchers went to great lengths for the legitimacy of this study. They controlled for other direct factors, such as macroeconomic policy, general market policy, and others. They also checked extensively to rule out reverse causality, i.e. where the market is governing the song choices of the nation.
Why this matters, is to differentiate between the wacky prima facie idea that might come into the forefront when first reading about this phenomenon, and what this causation is trying to tell. It is not that music somehow has started affecting stock market returns. It is simply that we have found a proxy for consumer sentiment, and harnessing this new data source is revealing groundbreaking conclusions.
What we see at play here is a convoluted mix of emotional congruity, the consistency between the internal emotions of an individual and how they manifest into their dealings with the external environment, behavioural finance, which studies how irrational decisions may be routed in psychological effects, and the butterfly effect, how the mood of a single bad actor, shared over an acceptable number of actors, can dictate market sentiment.
Fundamentals or Sentiments?
Linking sentiment measures with the stock markets, it was observed that higher music sentiment is connected with higher returns on markets during the same week, with lower returns the following week, implying that the initial reaction was sentiment-driven.
It is not only music that can influence returns. A previous research paper by Professor Edmans studies the impact of sporting events. It was discovered that when a country is eliminated from the World Cup, its stock market drops by an average of 0.5% the next day as a result of the negative impact on investor sentiment. Frieder and Subrahmanyam (2004), delved deep into the correlation between festivals and markets and discovered that market volatility always rises high on such important days.
Given how things like music and sports can influence stock market outcomes, the question arises - are markets driven by fundamentals or sentiments? Past research proves that stock markets are affected by emotions. Ultimately, an investor’s frame of mind influences investment choices and hence returns.
Conclusion
In conclusion, we’d like to put out three questions of the rhetorical kind in front of the reader:
1) Is big data going too far? When does its pragmatic use end and its far-reaching capabilities become equipped to be utilised for nefarious purposes? Today, we see this being used for financial purposes. Yet, should one be wary of usage of such data for other purposes, say, oppressive regimes tracking such data to warn themselves about rebellion.
2) What does this say about markets? A lot of investors believe the stock exchanges to be this efficient mechanism dictated by rational actors. The mood of rational actors is not the primary cause of their market activity, rendering music to be emblematic of stock price dynamics as well as the EMH to be a case of unresolvable dissonance.
3) Thirdly, and most importantly, what does this mean for the world of investing? Conventionally, one might look at PE multiples, EBITDA, and other values which help it guess the health of the company, and thus abide by fundamental analysis. Yet, with the advent of sophisticated tools which help discern market sentiment, or at least its proxy indicators, will we perhaps see funds and investing methodologies rise who follow this vastly unknown path?
References
Author: Naman Gambhir and Ketav Rastogi
Illustration by: Kayna Arora and Divya Jha