RDP 9808: What Moves Yields in Australia? 1. Introduction

At the broadest level, asset prices and interest rates are determined by saving and investment, and the return on capital. These, in turn, depend to an important extent on underlying economic conditions and expectations about economic performance. Thus, news which influences these expectations can influence prevailing yields, as investors adjust their portfolios in response to any reassessment about the economic outlook. This paper examines how the arrival of new information affects yields for short- and long-term securities in Australia. We identify ex ante what information might be expected to affect interest rates, and measure announcement effects associated with the release of this information. The analysis is based on high-frequency data for market yields, and precise release times for relevant announcements. The focus is the regular stream of macroeconomic indicators as well as policy announcements.^[1]

The paper does not seek a general explanation of interest rate behaviour. We simply attempt to measure how the fixed-interest market in Australia assesses and reacts to new economic information, the release of which is known in advance and the content of which can be forecast. We acknowledge that the market might respond to other economic and financial information or announcements – for example, an unforeseen move in the exchange rate or in bond yields abroad, or, say, an announcement by a rating agency. The current exercise is, nevertheless, of interest since it helps describe how Australian markets respond to an important set of economic information. It also sheds light on whether traders tend to react in a ‘rational’ way to new information – that is, are their responses in accordance with generally accepted views about how the economy and monetary policy work? Put differently, if these reactions were physical phenomena, it would be reassuring to find that they were consistent, say, with Newton's laws rather than chaos theory. The Australian case might also be of more general interest since, although the market is relatively small in global terms, it is open, active and highly deregulated.

The paper examines developments over the period from January 1994 to September 1997. This period is determined primarily by data availability but is also of interest because it was marked by a number of phases in interest rates. As well as investigating what news has been important for the period as a whole, we examine whether market reactions to particular items of news have varied between these different phases.

The paper is arranged along the following lines: Section 2 contains a brief description of how the literature on this topic has developed; Section 3 defines the data, episodes and news that are of interest; the relative influence of local and overseas news on Australian yields is discussed in Section 4; Section 5 refines the estimates for Australian news; and Section 6 draws some conclusions.

Footnote

The paper parallels the approach taken by Fleming and Remolona (1997), of the Federal Reserve Bank of New York, who examined the announcement effects of new information on 5-year bond prices in the United States. The current study extends Fleming and Remolona's approach to search for effects on short-term yields (90-day bills) as well as long-term yields (10-year bonds). [1]