The Impact of Presidential Elections on Currency Values in Latin America

This paper employs event study methodology to investigate the impact of presidential elections on nominal currency values for a group of Latin American countries between 1980 and 1996. The results indicate that presidential elections are associated with a statistically significant decline in nominal currency values, which appears strongest around the time of inauguration. The results suggest that the stage of the political cycle in a country is an important factor influencing exchange rates and should be recognized when making trade and investment decisions. INTRODUCTION This paper examines the impact of presidential elections on nominal currency values for a group of fifteen Latin American countries between 1982 and 1996 using event study methodology. A total of forty-one separate presidential elections are examined. For each election in the sample, the monthly percentage change in nominal currency value is calculated for each of the sixty-one months surrounding the election itself and the inauguration of the winner. The cross sectional average monthly change in currency value is then examined statistically to investigate whether currency values experience abnormal declines in the months surrounding a presidential election. The statistical results of this study suggest that presidential elections often have a significant impact on nominal currency values. In general, the approach of a presidential election leads to larger than normal monthly decline in nominal currency value, with the largest declines occurring close to the time of inauguration. After inauguration, the rate of currency depreciation tends to diminish. METHODOLOGY In the past, the most common application of event study methodology has been to examine the impact of various events on common stock prices. For example, many studies have investigated the impact of dividend announcements, mergers, etc. on the share price of the announcing firms. However, the methodology is potentially useful in a wide range of cases where the intent is to observe the effect of some event on a market rate or price. Previous studies using the methodology with exchange rates include those by Kwok and Brooks (1990) and Cosset and de la Rianderie (1985). MacKinlay (1997) provides a good introduction to the use of event study methodology in the areas of Economics and Finance. The first step in an event study is to define the event of interest and identify a sample of such events. In the present case, we are interested in the impact of the presidential election process and we identify two separate events for each election cycle: the election itself and the inauguration of the winner. A separate statistical analysis is completed for each of these two events. In each case, the month in which the event occurs is designated as the event month and is given a time index of t=0. Months before and after the event month are identified by their relative place in time. For example, the month before the event month is month -1 and the month after is +1. The sample examined consists of forty-one presidential elections in fifteen Latin American countries. Countries in the sample include: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Paraguay, Peru, Uruguay and Venezuela. Of the Latin American countries located in Central or South America, only Nicaragua was excluded (due to lack of required data). For each of the fifteen countries included, all regular democratic presidential elections occurring between 1980 and 1996 were identified. The sample includes one election from Paraguay (`83); two elections each from Brazil (`89, `94), Chile (`88, `93), Honduras (`89, `93), and Uruguay (`89, `94); three elections each from Argentina (`83, `89, `95), Bolivia (`85, `89, `93), Ecuador (`84, `88, `92), El Salvador (`84, `89, `94), Guatemala (`85, `90, `95), Mexico (`82, `88, `94), Peru (`80, `85, `90) and Venezuela (`83, `88, `93); and four elections each from Colombia (`82, `86, `90, `94) and Costa Rica (`82, `86, `90, `94). …