A Game of Musical Darts: How Music Charts inform Revenue Predictability and Investment Choices in the Music Industry

Abstract

What determines success in the music industry? What can we learn from music charts and how can we use them to inform our music decisions? Is investing in songs and recordings a game of darts or can we predict future revenues and make informed investment decisions? What does all of this mean for those interested in investing in the industry? Previous research focused on identifying characteristics of the recording themselves that might help forecast the popularity of a song or artist and measured success in terms of the top position held on a music chart or the duration of existing in the charts. This paper attempts to use music charts as a means of information of past success in order to try to predict future generation of revenue. First we take our data and determine revenue trajectories for each song. Second, we extrapolate our data by using these trajectories to estimate future revenue streams. Third, we define a selection criteria for constructing our portfolios of songs. We then propose a prediction metric that compares future revenue streams with past revenue data and calculate this prediction metric for portfolios of songs of different sizes. Finally we use the modern portfolio theory to compare different portfolios with each other to determine if we can use chart information to consistently create high performing portfolios. We find that though chart information does not seem to improve our prediction metric for different portfolios and our forecasting accuracy decreases as duration of past information increases, chart information does have a positive impact on portfolio performance under the modern portfolio theory.