88th International Atlantic Economic Conference
October 17 - 20, 2019 | Miami, USA

Global market reaction to the 2016 United States presidential election

Saturday, 19 October 2019: 9:20 AM
Seyed Mehdian, Ph.D. , School of Business, University of Michigan-Flint, Flint, MI
Delia-Elena Diaconaşu, Ph.D. , Department of Research, Alexandru Ioan Cuza University, Iasi, Romania
Ovidiu Stoica, Ph.D. , Finance, Money and Public Administration, Alexandru Ioan Cuza University of Iaşi, Iasi, Romania
The paper investigates the reaction of several international stock exchanges to the same news, in this case a major political event, the 2016 U.S. presidential election.

The election of Donald Trump in the 2016 presidential election was a surprise to almost everyone, domestically and internationally. One may argue this surprise rests on several factors such as the manner that he conducted his campaign, the components and slogans of his campaign, and the approach he used to conduct his campaign. Clearly, his promises, sometimes unrealistic, his confrontational behavior during the debates and his exaggerated abilities to solve economic problems, all of these were sufficient reasons to stun many when he was elected. To add to this, he was not, based on the judgement of many, a seasoned politician and was perceived as not part of the establishment arena.

Given the factors that contributed to his election were a surprise, in this paper, we investigate the reaction of the international investors to this supposedly global surprise and shock. More specifically, we use daily returns of stock market indices of different countries to investigate whether investors around the world reacted to this event in the same manner. In order to test for the stationarity of the returns series, we perform the Augmented Dickey-Fuller (ADF) and Kwiatkowski-Philips-Schmidt-Shin (KPSS) tests on each return series. We determine whether important events have occurred in the time series returns data; and identify the market surprises by estimating a GARCH(1,1) model for each index. We perform the F-statistic in order to determine if the variances of post-surprise X-day windows (favorable and unfavorable news) and non-event are significantly different.