Hi everyone,
We are Jennifer Jiang and Varisara Pongpairoj. We've been using Quantopian for the Quant Finance lectures run at the BU Finance and Investment Club.
Recently, we made a hypothesis test to see whether gold price return and market volatility have positive relationship or not We made an assumption that GLD tracks actual gold price. We also decided to calculate variance of S&P 500 returns within each 11 market opening days instead of VIX because we want more flexibility, for instance, we can freely choose any number of days to calculate market volatility. We partially referenced Preston Yadegar's Python code for "Interest Rates v. Market Volatility" (https://www.quantopian.com/posts/interest-rates-v-market-volatility). In addition, we also built a linear regression model for GLD price and market volatility, which is significant at alpha =0.1. We would like to get some feedback on our project.
We first did a regression with rolling GLD mean returns and rolling SPY variance over 1400 days. Then we tried method to test whether there is a relationship. We divided our data into two groups based on whether the gold price is increasing or decreasing and used the Levene Test to compare the variances of S&P 500 (market volatility) for GLD_returns>0 and GLD_returns<0.
Thanks a lot!
Jennifer and Varisara