The Making of a Crisis in Mexico: An Inductive Analysis of Media Sentiment and Information Cascades on the Value of the Mexican Peso during the 2008 Global Financial Crisis

Abstract

In the two decades prior to the 2008 financial crisis, the Mexican government pursued policies aimed at liberalizing markets, while simultaneously trying to ensure the stability of the peso. These policies consisted of monetary and fiscal controls to keep inflation low and free trade agreements to reduce Mexico's dependence on the United States. The policies significantly reduced the country's public deficit and were implemented in hopes that they would help reduce the country's exposure to currency crises. Yet, despite all provisions the Mexican government put in place, the country's peso still lost two percent of its value in the first three days following the bankruptcy of Lehman Brothers, the US-based investment firm. The loss was significant given the average appreciation of the peso in the months leading up to the crisis was one percent per month, and given that not enough time had passed to fully understand the impact that bankruptcy would have had on Mexico. By the following Monday, the peso recovered all of its lost value, suggesting that investors were uncertain about the true impact the events unfolding in the United States would have on Mexico's economy. It also suggested that the uncertainty and negative sentiment within the market during the initial week of the global crisis played a stronger role in the rapid depreciation and recovery of the peso than changes in market fundamentals. Using an inductive analysis of the historical events, this thesis suggests the circumstances in which sentiment engendered by mainstream media and distributed through digital channels during the financial crisis could have contributed to the dramatic short-term swings in the price of the peso. Specifically, this paper focuses on the new, digital information technologies, their use among investors as a means for financial research, and the role of high-frequency trading (HFT) algorithms in initiating information cascades. HFT algorithms account for nearly 70 percent of daily trading volume in financial markets and can magnify negative market sentiment among rational investors. Utilizing historical trading data for the peso and headlines and tweets published by the Thomson Reuters news group during the crisis, I seek to illustrate the correlations between market sentiment manifest in digital media and the price movements of the peso, indicating possible herd behavior tendencies in the form of information cascades. Though it is not possible to empirically separate the market movements of informed decision-makers from the information cascades of investors and HFT algorithms reacting to media, the fact that information cascades can and do exist as demonstrated by specific examples in this paper has significant implications for the Mexican peso. The existence of information cascades implies that having strong macroeconomic fundamentals is no longer an adequate safe guard against the immediate impacts of external crises. As social media becomes the main source of breaking news and market sentiment for mainstream media and investors, it becomes vital for emerging countries such as Mexico to monitor social platforms for sentiment related to the domestic economy in order to proactively address investor pessimism. Finally, emerging country governments can utilize these platforms to push out relevant and truthful information about the economy in order to diminish investor uncertainty and minimize the impact of externally-induced information cascades.