Will Meme Stocks Come Back?
So far in 2023 meme stocks have powered to an enormous 50%+ return. So it seems that our old friends have come back to play.
Over the last few years Rebellion Research ran a study on whether or not ‘The Meme Stock Index Act As An Equity Risk Forecasting Mechanism?’
Firstly, what is a meme stock?
A term referring to stocks that experience significant, rapid price increases due to the influence of social media platforms like Reddit and Twitter. A surge in value is typically not grounded in the fundamental performance metrics of the underlying companies, often resulting in subsequent price corrections or crashes. The study aims to explore the potential of meme stock behavior as a predictor for broader market movements, specifically the S&P 500, and the 2022 stock market crash.
Methodology:
Our team adopted a novel approach by creating a meme stock index, comprising major meme stocks such as GameStop and AMC. This index forms the basis of a machine learning model designed to predict buy and sell signals. The model’s predictive power is primarily assessed through its beta, a measure of volatility and correlation relative to the broader market.
Key Findings?
The paper’s findings are twofold. First, the meme stock index demonstrates a moderate predictive accuracy of 57.38% in forecasting buy and sell signals. This level of accuracy, while above chance, suggests a degree of unpredictability and risk inherent in using meme stock behavior as a market indicator. However, more notable is the model’s ability to anticipate market sell-offs, achieving an 88.89% accuracy in predictions for downturns occurring 9 days later. This high degree of accuracy in a specific timeframe is particularly significant and suggests that meme stock behavior may indeed have predictive value for certain market movements.
Implications and Contributions?
This paper contributes to the understanding of how new forms of social media-driven investment behavior can impact broader financial markets. The study highlights the growing influence of non-traditional market participants and the potential for unconventional data sources in market prediction models.
Acknowledgment!
We extend our gratitude to Cornell Financial Engineering & Balyasny’s Dr. Giuseppe Paleologo, whose guidance was instrumental in the conceptualization and execution of this study.
