Macroeconomic news, public communications, and foreign exchange jumps around U.S. and European financial crises

• Author: Jiahui Wang,Walid Ben Omrane,Robert Welch,Mohamed A. Ayadi
• DOI: http://doi.org/10.1002/ijfe.1747
• Published date: 01 April 2020
• Date: 01 April 2020

RESEARCH ARTICLE

Macroeconomic news, public communications, and foreign

exchange jumps around U.S. and European financial crises

Mohamed A. Ayadi | Walid Ben Omrane | Jiahui Wang | Robert Welch

Department Finance, Operations and

Information Systems, Goodman School of

Business, Brock University, St. Catharines,

Ontario, Canada

Correspondence

Walid Ben Omrane, Department of

Finance, Operations, and Information

Systems, Goodman School of Business,

Brock University, St. Catharines, Ontario

L2S 3A1, Canada.

Email: wbenomrane@brocku.ca

Abstract

Jumps in the Euro, Pound, and Yen, based on 5-minute returns for the period

2004–2015, are shown to be state dependent between recessions and expansions

in their response to macroeconomic news announcements and speeches by trea-

sury and central bank senior officials. We find evidence of large jumps and

cojumps response to the Federal Open Market Committee rate decision consis-

tently over economic states. U.S. news is more important than EU news and

jump magnitude and probability exhibit positive responses. Federal Reserve

senior officials' speeches generate more jumps during the U.S. mortgage crisis

and the EU sovereign debt recession. Although public communications of some

European Central Bank and Bank of England senior officials cause fewer jumps,

they produce significant cojumps of the three major currency markets.

KEYWORDS

foreign exchange markets, high frequency data, jumps, public communications, smooth

transition regression

JEL CLASSIFICATION

F31; G1

1|INTRODUCTION

We examine the effects of U.S. and EU macroeconomic

news announcements and public communications deliv-

ered by senior officials on three major currency's jumps

around the U.S. mortgage and European sovereign debt

crises. Specifically, we explore the Euro/Dollar

(EUR/USD), Pound/Dollar (GBP/USD), and Yen/Dollar

(JPY/USD) jump responses to specific announcements

and explicit speeches affecting the foreign exchange

(FX) market by senior politicians and central banks of

the United States, European Union, United Kingdom,

Germany, and Japan. Any response is analyzed relative

to economic business cycles characterized by the

U.S. financial and EU sovereign debt crises.

For many years, continuous time diffusion has formed

the basis for asset pricing models. However, empirical

research in financial markets has noted some large move-

ments in prices that may violate Gaussian assumptions,

motivating the analysis of discontinuities, or jumps. Huang

and Tauchen (2005) show the jumps account for about

4.5% to 7.0% of the total daily variance of the S&P index,

cash, or futures. The identification of jumps provides a

practical motivation for understanding how these disconti-

nuities affect financial markets. For example, Tauchen

and Zhou (2011) find that reliable identification of jump

dynamics improves the prediction of credit risk premia rel-

ative to interest rates andtypical volatility factors. It is gen-

erally assumed that the existence of jumps would

significantly influence financial and risk management.

The empirical relationship between macro news and

asset price movements provides an important insight into

financial markets.Andersen, Bollerslev, Diebold, and Vega

(2003) report significant news effects with asymmetric

Received: 11 April 2018 Revised: 22 November 2018 Accepted: 3 September 2019

DOI: 10.1002/ijfe.1747

Int J Fin Econ. 2020;1–31. wileyonlinelibrary.com/journal/ijfe © 2019 John Wiley & Sons, Ltd. 1

Int J Fin Econ. 2020;25:197–227. wileyonlinelibrary.com/journal/ijfe © 2019 John Wiley & Sons, Ltd. 197

response patterns on exchange rate movements. However,

their study does not focus on the discontinuous compo-

nent. Without explicitlymodeling the presence of jumps in

the pricing process, researchers note that discontinuities

are irregular and depend on market conditions. In this

regard, Piazzesi (2003) shows that incorporating jumps

related to market information improves pricing models in

bond markets. Moreover, recent studies of different asset

markets incorporating jumps (Ballotta, Deelstra, & Rayée,

2017) and macro news from variouscountries find a signif-

icant relationship between news and discontinuities

(Chatrath, Miao, Ramchander, & Villupuram, 2014; Dél-

èze & Hussain, 2015; Lahaye, Laurent, & Neely,2011).

Several studies assume that the impact of macro news

may depend on the state of the economy. For instance,

Veredas (2006) reports asymmetric responses to positive

and negative news for different phases of the business

cycle and that the Institute for Supply Management Sur-

vey (ISM) index provides a more accurate indication of

business cycles, because it is based on forward-looking

information and expectations.

There are alternative ways to measure business cycles.

Andersen, Bollerslev, and Diebold (2007) assume a reces-

sion begins if nonfarm payroll employment declines for

three consecutive months. However, Laakkonen and

Lanne (2009) and Ben Omrane and Savaser (2017) use

the two-regime smooth transition regression (STR) model

of Teräsvirta (1994) to endogenously define business

cycles. The advantage of this method is that the estimated

transition logistic functions corresponds to the probabili-

ties of being in an economic state, which replaces the ex-

post periods provided by international institutions such

as the National Bureau of Economic Research (NBER) in

the United States or the Center of Economic and Policy

Research (CEPR) in Europe. Although both institutions

are well respected and their approach is used in previous

research (Ben Omrane & Savaser, 2016), others, such as

Melvin and Taylor (2009) and Fratzscher (2009), develop

their own definitions for business cycles because the

NBER and CEPR dates could include a subjectivity bias

as well as differ from ex ante measures.

In our study, we seek answers to the following ques-

tions: Can domestic and foreign macroeconomic news and

public communications explain foreign exchange price

spikes? Are some senior official speeches or types of mac-

roeconomic news announcements prominent than others?

Finally, are jumps and cojumps respond to macroeco-

nomic news and speeches sensitive to economic recessions

such as U.S. mortgage and EU sovereign debt crises? These

questions are tackled through an empirical framework

that involves, first, the extraction of jumps and cojumps

from foreign exchange rates using a modified version of

the test for jumps, initially proposed by Andersen,

Bollerslev, and Dobrev (2007) and Lee and Mykland

(2008), extended by Boudt, Croux, and Laurent (2011) and

applied, among others, by Lahaye et al. (2011) and

Dewachter, Erdemlioglu, Gnabo, and Lecourt (2014); Sec-

ond, the endogenous identification of the business cycles

and more specifically the probability of being in U.S. or

EU crises using smooth transition regression models; And

third, the examination of the size and probability of jumps

and cojumps in response, around U.S. and EU crises, to

macroeconomic news surprises, public communications,

and unscheduled news announcements.

Our study provides evidence that jumps are surpris-

ingly the least sensitive to macroeconomic news during

the U.S. crisis, U.S. news is the most important for all three

currencies, and the EUR/USD is the most sensitive. Nearly

all significant news coefficients are positive indicating an

increase in the absolute value of a jump, and 90% of news

coefficients between the recession and expansion phase

around the U.S. crisis are significantly different from each

other, which is a state dependent effect—a difference in

degree. However, there are six news that exhibit a differ-

ence in kind as their recession and expansion coefficients

have opposite signs—the most extreme case of state depen-

dent (or context specific) news effects. Jumps occur more

frequently duringthe post-announcement period in expan-

sion compared with recession, precisely just after

U.S. macro news release at 8:30a.m. and 2:15 p.m. Eastern

Standard Time. Federal Open Market Committee (FOMC)

rate decision announcement triggers the largest jumps and

cojumps in the three majorcurrency markets. The FOMC's

news effect surges during the U.S. recession period. Fur-

thermore, jumps' response to public communications are

more pronounced through recession periods, particularly

during U.S. crisis. However, the response level depends

upon the senior official who delivers the speech and the

corresponding underlying currency.

The main contributions of this study include, first, the

unstable macroeconomic news effect on jumps between

expansion and recession states involving the U.S. mortgage

and EU sovereign debt crises—a context specific effect

(state dependency); Second, the consistent and pro-

nounced effect of FOMC rate decisions on jumps and

cojumps over different business cycles, where the highest

impact is detected during the U.S. recession; Third, domi-

nance of U.S. news effects relative to European Union and

other countries; Fourth, the preponderance of news caus-

ing larger jump magnitudes (positive coefficients) with

macro news having twice as many significant coefficients

as speakers; Fifth, jump response to speeches is more pro-

nounced from U.S. officials, particularly in the U.S. reces-

sion; Sixth, European Central Bank (ECB) and Bank of

England (BOE) senior officials' public communications

produce relatively fewer jumps but trigger significant

2AYADI ET AL.

198 AYADI ET AL.