Surprisingly, the US presidential campaign of 2015-2016 revealed that beyond the progressive increase of cross-border investment and trade, there is a growing link in the social-political sphere between the member states of the North American Free Trade Agreement (NAFTA). The noticeable effects of the campaign over the Mexican financial markets (and to a lesser extent over the Canadian markets) suggest that the political events of the country significantly tainted the financial markets of its trading partners.
Trump tweets and Mexico trembles
Since the pre-campaign for the Republican Party’s nomination, Donald Trump expressed his opposition to the millions of foreign citizens (the majority being Mexican) living and working in the United States illegally, promising to deport them. He also promised to construct his famous wall at the border to stop illegal immigration and drug trafficking. Furthermore, he also frequently talked about cancelling NAFTA if an extensive review could not be made for a “fairer deal for the United States of America”.
Most of the time that Trump put forth proposals against the commercial interests of Mexico or against Mexican citizens, it caused negative effects on the peso / dollar exchange rate and on the performance of the Mexican Stock Exchange. Even when the Trump campaign registered an improvement, in accordance to the electoral polls and betting markets, the Mexican peso depreciated against the American dollar and there was negativity in the Consumer Price Index (CPI). On the contrary, the comments made by the democratic candidate, Hillary Clinton, who was much kinder and more sympathetic towards Mexico, had a positive effect on the same financial variables. Curiously, the speeches made by both presidential candidates did not create the same effect on the Canadian financial market, and at times they were almost irrelevant.
An example of this was when, during the first debate between the republican candidates in august 2015, Trump said that the Mexican government had intentionally sent immigrants (“the bad guys”) to the United States: the Mexican financial markets suffered the consequences of that outburst (Figure 1).
Likewise, the stock and foreign exchange markets, both in Canada and Mexico, were very much affected by the end of 2015 and the start of 2016, as an increase was expected in the reference interest rates by the Fed (Figure 1 and 2). The volatility of the Mexican variables also increased from when Trump took the lead in the republican primaries in March 2016 and up until he was chosen as the republican presidential candidate. All in all, the most dramatic impact for the Mexican financial markets was seen the day after the presidential election and the announcement of Trump’s triumph. The Mexican peso depreciated from $18.32/USD to $20.57/USD, approximately 12% in only a matter of hours.
How do we measure market sentiment?
The observation of these correlations drove us to investigate this phenomenon, posing the main hypothesis that the Mexican stock and foreign exchange markets were significantly more affected by the news of the US electoral process than the Canadian markets. We tested the hypothesis taking measurements from the evolution of the preferences of American voters.
Generally, the way of measuring the subjective probability of the election results is through opinion polls. However, more recently the betting markets (otherwise known as prediction markets) are gaining approval as a trusted source of market sentiment regarding voters’ intentions.
In order to formulate our hypothesis in econometric terms, we obtained the data from the opinion polls and the prices of the prediction markets from November 2014 until November 2016, from FiveThirtyEight and Iowa Electronic Markets respectively. The methodology from Autoregressive Vectors (ARV) and Error Correction Vectors (ECV) was used to test whether the information from the voters’ sentiment is incorporated quicker than the information from the polls; and if the Mexican stock and foreign exchange markets were highly sensitive to campaign news, this is in light of Canadian markets experiencing only marginal effects.
These results are theoretically relevant from the perspective of the Efficient Market Hypothesis, as the Mexican financial variables showed that the hypothesis is met, although surprisingly, the same did not happen for the Canadian variables. Therefore, the administrators and portfolio managers can take a passive position when including the Canadian finance index in their investment strategies.
The most relevant conclusion of this study is that we need to strengthen the Mexican economy in order to reduce its enormous dependency on its only trading partner: The United States. As Economists, Academics and Governmental Authorities, we should all coordinate our efforts and propose strategic plans to achieve this.