This was an R script used in the Boston Music Hackday for a danceability market index for investing. The idea was to create something similar to the hemline index using the danceability rating of songs from Echo Nest.
Musichackday is a weekend hackathon for music tech and the idea is to create something fun related to music. This project was for fun, not real investing, and is obviously no longer profitable!
2 Minute Presentation:
For write-ups see:
The process:
-
use Last FM to find the most popular songs in NYC historically of a period of time
-
use the Echo Nest API to get the danceability rating of each song
-
create a moving average of how danceable the popular music was in NYC at each time period
-
data mine (overfit) to find profitable trading rule (this is not a serious endeavor)
-
simulate trading on S&P 500 going long when the indicator is above a threshold and short when below
-
profit