The idea for Glass Ticker came after Adam read a news article several years ago announcing that computers had outnumbered humans in quantity of stock market trades. We wrote Glass Ticker to examine the other side of the stock market: humans, and in particular, their susceptibility to news media.

What it does

We sought to determine the impact of a news article on a human's tendency to invest in that company. Our program computes a metric we call the Consumer Investment Coefficient, which is a value ranging from zero to one that represents the correlation that the news coverage of a company on a given day will influence casual investors, or people that do not work exclusively in finance, to invest in the company. The program also computes several other useful metrics that relate to the media's influence over the market.

How we built it

We used the NASDAQ APIs to feed live stock market data into our program. We brainstormed, outlined, and implemented a novel algorithm using several uncommon data transformations. We wrote the code in a variety of programming languages, but the primary mechanisms are implemented in Python.

Challenges we ran into

Visualizing our own processes was a complex undertaking due to the program's great degree of abstraction from reality and the complexity of the algorithm. We were, of course, also limited by the fallibility of our bodies, as two of our team members fell ill at some point, and were not able to make it through the night.

Accomplishments that we're proud of

We built a totally original algorithm that really works!

What we learned

Finance is hard!

What's next for Glass Ticker

We believe that Glass Ticker can be relatively easily validated by grouping NASDAQ-listed companies by sector and verifying that consumer industries are more highly correlated than industrial sectors, in accordance with our premise that news coverage plays a large role in influencing investor decisions.

Share this project: