Position Taking S&P500 and Apple Options through Predicting Implied and Realized Volatility

Abstract

While stock prices are often seen as trending random walks from a quantitative perspective, volatility is proven to have strong predictability from its autocorrelative properties. In this paper, we will create trading strategies for options using a different approach from past public literature; instead of typical neural networks and other ML strategies that predict option prices directly, we will examine delta-hedged SP500 and Apple options to isolate their volatility risk exposures and generate predictions on both realized and implied volatility using close-to and far-from-expiry options. We will utilize these insights to create a position-taking strategy that will reliably outperform long-equity portfolios.