Presidential Elections and the Stock Market: Comparing Markov-Switching and Fractionally Integrated GARCH Models of Volatility

Bumba Mukherjee Department of Political Science, Florida State University, 554 Bellamy Building, Tallahassee, FL 32306 e-mail: smukherj{at}mailer.fsu.edu Existing research on electoral politics and financial marketspredicts that when investors expect left parties—Democrats(US), Labor (UK)—to win elections, market volatility increases.In addition, current econometric research on stock market volatilitysuggests that Markov-switching models provide more accuratevolatility forecasts and fit stock price volatility data betterthan linear or nonlinear GARCH (generalized autoregressive conditionalheteroskedasticity) models. Contrary to the existing literature,we argue here that when traders anticipate that the Democraticcandidate will win the presidential election, stock market volatilitydecreases. Using two data sets from the 2000 U.S. presidentialelection, we test our claim by estimating several GARCH, exponentialGARCH (EGARCH), fractionally integrated exponential GARCH (FIEGARCH),and Markov-switching models. We also conduct extensive forecastingtests—including RMSE and MAE statistics as well as realizedvolatility regressions—to evaluate these competing statisticalmodels. Results from forecasting tests show, in contrast toprevailing claims, that GARCH and EGARCH models provide substantiallymore accurate forecasts than the Markov-switching models. Estimatesfrom all the statistical models support our key prediction thatstock market volatility decreases when traders anticipate aDemocratic victory.