U.S. presidential cycles and the foreign exchange market

• Published date: 01 October 2019
• Date: 01 October 2019
• DOI: http://doi.org/10.1002/rfe.1061
• Author: Salil K. Sarkar,Samar Ashour,David A. Rakowski

Rev Financ Econ. 2019;37:523–540. wileyonlinelibrary.com/journal/rfe

|

523

© 2019 University of New Orleans

The policy for my administration is for there to be a strong US dollar…

‐President George W. Bush

(Bloomberg News, 2003)

I believe deeply that it's very important to the United States, to the economic health of the United States, that we main-

tain a strong dollar

‐US Treasury Secretary Tim Geithner

(Bloomberg News, 2009)

Since Robert Rubin's time as Treasury secretary under Bill Clinton, every administration had insisted that a strong

dollar is in U.S. interests.

‐The Wall Street Journal (Mackintosh, 2017)

1

|

INTRODUCTION

For decades, US presidents of both parties have supported the policy of a strong US dollar. The purpose of this study is to evalu-

ate how political events affect the foreign exchange market by focusing on the impact of presidential cycles in the United States

(US), an issue that has been largely overlooked in previous studies. Understanding the potential link between currency markets

and presidential political affiliation is important because the president's political affiliation provides investors with information

about the administration's future economic policies. We show that there exist strong, robust, and meaningful differences in

returns to the US dollar between Democratic and Republican presidencies.

The two major political parties in the United States – Democratic and Republican – are perceived to follow different eco-

nomic policies. Indeed, many previous studies found that Democratic presidencies are associated with expansionist economic

Received: 25 May 2018

|

Revised: 12 September 2018

|

Accepted: 2 January 2019

DOI: 10.1002/rfe.1061

ORIGINAL ARTICLE

U.S. presidential cycles and the foreign exchange market

SamarAshour1,2

|

David A.Rakowski3

|

Salil K.Sarkar3

1Sam M Walton College of

Business,University of Arkansas,

Fayetteville, Arkansas

2College of Business,Tanta University,

Tanta, Egypt

3Department of Finance and Real

Estate,University of Texas at Arlington,

Arlington, Texas

Correspondence

David A. Rakowski, Finance & Real

Estate Department,UTA, PO Box

19449,Arlington, TX, 76019, USA.

Email: rakowski@uta.edu

Abstract

We examine the association between the foreign exchange rate of the US dollar and

US presidential cycles. Results show that Republican presidencies tend to start with

a strong dollar, which then depreciates over the course of the presidency. In con-

trast, Democratic presidencies tend to begin with a weak dollar that then appreciates.

These patterns result in an apparent presidential effect in US foreign exchange rates,

the direction of which depends on whether exchange rates are measured by levels or

by returns.

KEYWORDS

foreign exchange markets, politics and finance, presidential cycles, US dollar

JEL CLASSIFICATION

F310; F400; G150; G180; E650; P480

524

|

ASHOUR etAl.

policies (Alesina, Roubini, & Cohen, 1997; Hibbs, 1986). Hibbs (1986) argues that Republican administrations prefer to adopt

tighter monetary policies than their Democratic counterparts. Tighter monetary policies (i.e., austerity) should be associated

with lower short‐term economic growth, decreasing inflation rates, increasing interest rates, and an appreciation of the US

dollar. On the other hand, the looser (i.e., expansionary) monetary policies perceived to accompany Democratic administra-

tions should be associated with declining interest rates, increasing inflation rates, decreasing unemployment, higher short‐term

economic growth, and a depreciating US dollar. We recognize that political parties’ stances on monetary policy are necessarily

vague, subjective, and evolve over time. We, therefore, take an agnostic approach with regards to party dogma and examine the

extent to which the perspectives presented in the academic literature above are empirically valid with regard to the value of the

US dollar.

Using data from 1957 to 2016 for the US dollar against several major foreign currencies and from 1973 to 2016 for two real

trade‐weighted indices of the value of the US dollar, we find that there is a significant association between the political party

of the US president and the strength of the US dollar. Returns to real trade‐weighted US dollar indices (relative both to a broad

basket of currencies and against only major currencies) are negatively associated with Republican administrations, especially

after correcting for other macroeconomic variables. These findings are similar to what Santa‐Clara and Valkanov (2003), here-

after referred to as SCV, report for US stock returns, but is in contrast to what would be expected from the reasoning of Hibbs

(1986). The depreciation of the US dollar during Republican administrations is strongest relative to the Canadian dollar.

An analysis of the time‐trend of US dollar values over the course of a presidential term indicates that the US dollar tends

to start at a high value for Republican presidents and then depreciates, while the opposite pattern is true for Democrats. These

findings hold for 4‐year presidential terms, entire presidencies, or for contiguous periods when a party holds the presidency.

During Republican presidencies, we document a significant negative time‐trend for the US dollar against the Euro, Australian

dollar, Swedish krona, and Canadian dollar. Democratic presidencies are more mixed, with indications of a significant appre-

ciation of the US dollar against the Euro, Japanese yen, Canadian dollar, and Swiss franc. We thereby extend the analysis of

Kraussl, Lucas, Rijsbergen, van der Sluis, and Vrugt (2014) (hereafter, KLRSV) to consider the course of US dollar exchange

rates over the course of a US presidency. Overall, our results provide a consistent extension and intuitive link between the find-

ings of SCV for the stock market and of Blinder and Watson (2016) for the overall economy.

2

|

LITERATURE REVIEW

The possibility for US dollar exchange rates to be associated with the political affiliation of the US president follows from

a lengthy stream of research on the links between US political cycles and the values of both macro‐economic and financial

variables. Turning first to financial markets, several studies associate US political cycles with stock returns. Niederhoffer,

Gibbs, and Bullock (1970) found that there is no significant difference in stock returns between Democratic and Republican

presidencies, with the exception that Democratic presidencies were associated with higher stock returns in the third year of

the presidency. Herbst and Slinkman (1984) used data from 1962 to 1977 and found that there was a significant association

between 48‐month presidential terms and the stock market cycle. KLRSV also found evidence of patterns in both stock and

credit markets over the course of a presidential term, especially during Democratic presidencies. Hensel and Ziemba (1995)

and Johnson, Chittenden, and Jensen (1999) report that there is a statistically significant difference between Democrats and

Republicans for small stocks. Sabherwal, Sarkar, and Uddin (2017) show that the returns to sin stocks (i.e., those industries that

are supposedly associated with unethical or immoral activities) are systematically associated with US political cycles. Li and

Born (2006) use uncertainty associated with the 2000 US presidential election to provide evidence that the US stock market

is affected by political uncertainty, while Kelly, Pástor, and Veronesi (2016) provide similar evidence from option markets.

SCV provide the most comprehensive study of the association between the US president's political party and the stock

market. They find that Democratic presidential administrations are associated with higher stock market returns than their

Republican counterparts, even after controlling for firm size and business‐cycle variables. The findings of SCV are extended by

Sy and Zaman (2011), who investigate whether risk can explain the difference in market performance between Republicans and

Democrats using conditional multi‐factor asset‐pricing models in which systematic risk can vary across political cycles. They

show that systematic risk varies across political cycles and, consequently, the market risk premium and expected stock returns

are higher under Democratic presidencies. Powell, Shi, Smith, and Whaley (2007) refine the estimation methodology of SCV

to correct for persistence in the explanatory variables used to identify presidential party affiliation. Together, existing studies

in financial variables and US political cycles provide several pathways by which US presidential affiliations may be associated

with demand by investors for US dollar denominated financial assets. Whether this demand translates into actual price pressure

on the US dollar relative to other currencies is one of the empirical questions that we seek to answer.