St. James Ballroom
New GARCH Methods for Financial Analysts: Using Industry Knowledge to Develop Better Forecasts for Common Financial Products (303877)
*Vishruti Ganesh, Dougherty Valley High School
Michael Kotarinos, University of South Florida
Chris P. Tsokos, University of South Florida
Keywords: Exchange Traded Funds, Time Series Analysis, k-th Moving Average, k-th Weighted Moving Average, Time Series Forecasting, Volatility, Risk
The following project modifies the traditional GARCH procedure by applying two new methods that are meant to mimic the opening and closing of stock exchanges: the k-th moving average and the k-th weighted moving average. In order to observe the effectiveness of the methods, they are used to model the volatility of the returns from 9 different ETFs (Exchange-Traded Funds). The 9 ETFs used are SPDR S&P 500, Financial Select Sector SPDR, iShares Russell 2000, iShares Investment Grade Corporate Bond, iShares 1-3 Year Treasury Bond, iShares Emerging Markets, iShares China, Vanguard Total Bond Market Index Fund, and SPDR S&P transportation. In using these 9 different ETFs, data is captured from various sectors of today's economy, creating useful insights for financial economists and asset managers across disciplines looking to manage risk. After analyzing the AIC values of each method applied to the ETFs, it can be seen that the proposed models, for all nine ETFs, have a lower AIC value than the traditional GARCH procedure. Therefore, it can be contended that the proposed models not only more accurately measure risk but are structurally similar to ideas and methods used across finance.