IMPACT OF MACROECONOMIC ANNOUNCEMENTS ON INTEREST RATE FUTURES: HIGH‐FREQUENCY EVIDENCE FROM AUSTRALIA

• Published date: 01 September 2013
• Date: 01 September 2013
• Author: Lee A. Smales
• DOI: http://doi.org/10.1111/j.1475-6803.2013.12015.x

IMPACT OF MACROECONOMIC ANNOUNCEMENTS ON INTEREST RATE

FUTURES: HIGH‐FREQUENCY EVIDENCE FROM AUSTRALIA

Lee A. Smales

Curtin University, University of New South Wales

Abstract

I investigate the behavior of Australian interest rate futures around the release of major

scheduled macroeconomic announcements. The adjustment to new information occurs

quickly with the majority of the reaction complete within 30 seconds. The period

immediately before the announcement exhibits high volatility, low levels of volume, and

wide bid–ask spreads. In the 30 seconds following the scheduled announcement there is a

sharp increase in price volatility, significant positive correlation in returns, high levels of

trading activity, and large adjusted returns. The reaction is stronger in shorter maturity

contracts, and in the period surrounding the 2007–2008 financial crisis.

JEL Classification: G10, G14, G15

I. Introduction

I explore the price discovery dynamics of Australian interest rate futures in the period

around major scheduled macroeconomic announcements. High‐frequency data analysis

has become the norm in seeking to explain the economics of such price movements.

Indeed, Fleming and Remolona (1997) suggest that daily data are not sufficient to capture

the market reaction cleanly, or for capturing the microstructure aspects of the price

adjustment mechanism.

The identification of what constitutes a “major”economic release has developed

a literature of its own, and there is evidence of some instability in the market effects of

various macroeconomic announcements. Fleming and Remolona (1997) find that the

behavior of the fixed‐income market has a flavor‐of‐the‐month aspect in which different

announcements are regarded as important in different periods. For instance, studies from

the late 1970s to the mid‐1980s document a significant impact of money supply

announcements, while more recent studies, such as Cook and Korn (1991) and Krueger

(1996), solidify the importance of the nonfarm payrolls number in the monthly

employment report. Ederington and Lee (1993) note seven announcements, including

employment, as having a significant effect on U.S. equity markets.

I am very grateful to my referee, Christopher J. Neely, for his insightful feedback during the review process. I

gratefully appreciate the comments from the associate editor, Louis H. Ederington, which improved the scope and

conciseness of the article. I also thank seminar participants at the University of New South Wales and the 2012

Financial Markets and Corporate Governance conference for useful comments and suggestions. All remaining errors

are mine.

The Journal of Financial Research Vol. XXXVI, No. 3 Pages 371–388 Fall 2013

371

© 2013 The Southern Finance Association and the Southwestern Finance Association

Market efficiency would suggest that the expected portion of an announcement

should have no effect on asset prices, and recent studies (Anderson et al. 2003; Evans and

Lyons 2005) implicitly incorporate this assumption by focusing solely on the surprise

component of macroeconomic announcements. Kuttner (2001) and Fatum and Scholnick

(2008) find that asset returns and volatilities respond primarily to the surprise

component—measured by the difference between the actual announced figure and

the expected figure taken as the median of survey data—in macroeconomic data

announcements. Chen, Jiang, and Wang (2013) note that announcement surprises have a

significant impact on the trading activities and returns of U.S. equity futures.

If markets are efficient, news should be quickly impounded into market prices.

However, if the price adjusts slowly to the new information, it may be possible to earn

excess returns based on the initial market reaction to the news release, thus violating semi‐

strong‐form efficiency. Several studies (e.g., Urich and Wachtel 1981; Kim and

Sheen 2001; Han and Ozocak 2002) find support for at least semi‐strong‐form market

efficiency in the sense that prices react swiftly to news, and profit opportunities disappear

quickly as a result. Ederington and Lee (1993, 1995) find that efficiency holds in the

Treasury bond and foreign exchange markets, and Becker, Finnerty, and Kopecky (1996)

investigate the Treasury bond market and, based on conditional expectations, find

significant price adjustments, but only for Consumer Price Index, nonfarm payrolls, and

the trade balance.

The nature of an efficient market limits the length of the predictable return

response to news but does not place such restrictions on the impact of news on volatility. It

is therefore possible that the volatility response differs from the return response, and as a

result, another branch of analysis has focused on the effect of macroeconomic

announcements on asset price volatility. The general finding is that news releases result in

a rapid increase in return volatility; although the majority of the effects are relatively

short‐lived and subside within the first minute, there are varying degrees of volatility

persistence for periods ranging from 15 minutes to several hours—this is possible even if

the market is efficient. This may be explained by an increase in trading activity as market

participants rebalance their portfolios in light of the news, information traders enter the

market, and liquidity traders benefit from the elevated interest. Fleming and Remolona

(1997) reason that sharp price moves, and the accompanying increase in volatility, are

attributable to changes in expectations shared by investors, and surges in trading activity

are due to a lack of consensus on prices. Ederington and Lee (1993) suggest that volatility

may remain high for some period after an announcement because full information only

arrives gradually owing to the length of some reports and the analysis required in forming

a conclusion on the likely effect of the data release.

In addition, there is mounting evidence that asset prices respond differently to

macroeconomic news depending on market conditions as well as the general state of the

economy, with an impact on both returns and volatility. McQueen and Roley (1993) were

the first to note this effect in the equity market, and Roley and Sellon (1995) find the

relation between news and long‐term rates is likely to vary over the business cycle as

market participants alter their views on the persistence of policy actions. Cook and Korn

(1991) hypothesize that this is due to policy anticipation; the market reaction depends on

how participants expect the central bank to move monetary policy in response to the news,

372 The Journal of Financial Research