The Impact of “Music Emotion” on the Stock Market

The term “market sentiment” describes the views of buyers and sellers towards a specific security or financial market. In simple terms, positive market sentiment is often reflected in rising prices, while falling prices indicate negative market sentiment. It represents the attitude or atmosphere of the market, also known as collective psychology within the industry, which can be inferred from the activity and value changes of securities traded in that market.

As participants in the market, we often observe the impact of emotions fueled by various social events in the market. For example, during periods of political turmoil, people tend to hold a pessimistic attitude towards the market, which includes significant price volatility in the capital market and herd behavior. However, it is difficult to describe the influence of emotions on the market with a specific indicator and understand their relationship.

 “Music is a universal language.” – In a study conducted in 2019, Mehr et al. found that music may reflect universality in human perception worldwide. Following this, researchers have constructed an index to measure investor sentiment based on the key of music within the framework of “emotional consistency” and continued to analyze the influence of emotions on global stock markets.

This article introduces a method of measuring investor sentiment by examining the “positivity” of songs individuals choose to listen to, and further investigates the impact of emotions on stock market returns.

Identifying Music Emotion

To study the impact of emotions on stock market returns, the first and most crucial step is to accurately measure “market sentiment”.

The earliest sentiment indicators were indirect quantification methods based on market transaction data, such as turnover rate, discounts on closed-end funds, implied volatility of options, etc. However, these indicators may be influenced by economic fundamentals rather than emotions. For example, implied volatility also increases when uncertainty rises.

Some researchers have attempted to obtain the indicator of “market sentiment” indirectly, and “music emotion” is one of them. To measure music emotion as a market sentiment indicator, scholars attempted to define an index of “music positivity”. They collected statistical data on the top 200 streamed songs in each country or region every day from Spotify1. The songs listened the most in each region can reflect people’s emotions to some extent, for example, people are more willing to listen to happy songs in joyful situations.

After obtaining the most popular music, the next step is to define the positivity of these songs. Spotify provides an indicator called ‘valence’ to measure the positivity of songs. This indicator rates the positivity of music on a scale from 0 to 1, where higher scores indicate more positive music and lower scores indicate more negative music.

Finally, the authors constructed an index to measure the music emotion of a country or region on a particular day – the “weighted average” of music valence for each country. This weighted average is calculated by taking the daily play count of each song as the weight and averaging the music valence. This weighted average can represent the average “music emotion” of a country or region on that day, which can further reflect “market sentiment”.

The Impact of Music Emotion on Stock Prices

In the previous section, we introduced how to identify music emotion. The next step is to use statistical methods to examine whether market sentiment represented by music emotion is correlated with market volatility to some extent.

In this study, researchers collected data from a total of 40 countries or regions from 1 January 2017 to 31 December 2020. Within these data, there were over 58,000 songs with a total of over 50 billion streams. On average, there were 86 million daily streams per country, providing ample data for analysis.

To balance the influence of other factors on the market in the data, the researchers used control variables to reflect the true effect of “music emotion” more accurately. To control other seasonal emotional fluctuations not captured by music-based emotion measures, the authors used the US Economic Policy Uncertainty (EPU) index2 from the Scott Baker website as a standard for global macroeconomic changes. They also controlled for the implied volatility of the S&P 500 index obtained from the Chicago Board Options Exchange website, along with other methods to control for other effects that could drive changes in stock returns.

The results after conducting regression calculations are shown in the table below. The dependent variable “RET” represents weekly stock market returns. In Group A, the independent variable is the music emotion defined earlier, while in Group B, the lagged one-week music emotion is used as the key independent variable. Other variables are control variables, such as the concurrent implied volatility (VIX) and weekly changes in economic policy uncertainty (ΔEPU). The “***” in the table indicates significance at the 1% level.

Table 1. Regression Estimates of Music Emotion and Stock Market Returns from 2017 to 2020

According to the table, when music emotion increases by one standard deviation, weekly stock market returns increase by 8.1 basis points (6.758 multiplied by the independent variable’s standard deviation of 1.201) and are significant at the 1% level. Simultaneously, in the experiments conducted in Group B, it is evident that there is a “reversal” phenomenon in music emotion’s impact on the next week’s stock returns. That is, when music emotion increases by one standard deviation, next week’s stock returns decrease by 7.1 basis points (annualized return of 3.7%) and are significant at the 1% level. This is a price reversal pattern consistent with temporary mispricing caused by emotions.

The experimental results demonstrate that the emotion index based on music emotion is positively correlated with stock market returns in the current week and negatively correlated with stock market returns in the next week. These opposing results also confirm previous literature’s conclusions regarding “temporary pricing errors caused by emotions”.

The Future and the Challenges

In market analysis, analyzing “market sentiment” can be seen as a supplement to fundamental analysis. This can help in understanding the overall risk preference of the market, assisting in determining market conditions, making decisions at significant turning points in the market, and formulating corresponding quantitative investment strategies. For example, if it is determined that market sentiment is strong and not a rapidly fading trend, it may indicate a sustainable upward trend in the stock market. Conversely, if a significant upward trend is followed by a downturn, it should be considered whether market sentiment has become overheated, and it may be prudent to exit quickly. When “market sentiment” diverges from stock indices, it indicates an impending reversal in the market.

This article discusses an indirect method of measuring “market sentiment,” namely the “music emotion” indicator. Since market participants are not entirely rational, “market sentiment” undoubtedly exists. However, due to its abstract nature, observing and measuring “market sentiment” is highly challenging yet meaningful.

Reference

[1] Edmans, A., Fernandez-Perez, A., Garel, A., & Indriawan, I. (2022). Music sentiment and stock returns around the world. Journal of Financial Economics, 145(2), 234-254.

[2] Mehr, S. A., Singh, M., Knox, D., Ketter, D. M., Pickens-Jones, D., Atwood, S., … & Glowacki, L. (2019). Universality and diversity in human song. Science, 366(6468), eaax0868.

[3] 中國銀河證券研究院. 從市場情緒看交易策略. 15 May 2019,

[4] 楊暘,趙麗. 證券市場投資者情緒指數構建研究. 上海證券交易所, 2019,

The work described in this article was supported by InnoHK initiative, The Government of the HKSAR, and Laboratory for AI-Powered Financial Technologies.
(AIFT strives but cannot guarantee the accuracy and reliability of the content, and will not be responsible for any loss or damage caused by any inaccuracy or omission.)


  1. A legitimate online streaming music service platform. ↩︎
  2. The index is derived by calculating the quantity of articles from US newspapers in the NewsBank Access World News database. ↩︎

Share this content

Address

Units 1101-1102 & 1121-1123,
Building 19W Science Park West Avenue,
Hong Kong Science Park,
Shatin, Hong Kong

Products & Solutions

People

About Us

Address

Copyright © 2024 Laboratory for AI-Powered Financial Technologies Ltd. All Rights Reserved.