تاثیر اخبار اقتصاد کلان بر معاملات سلف فلزی

This paper uses intra-day data for the period 2002 through 2008 to examine the intensity, direction, and speed of impact of US macroeconomic news announcements on the return, volatility and trading volume of three important commodities – gold, silver and copper futures. We find that the response of metal futures to economic news surprises is both swift and significant, with the 8:30 am set of announcements – in particular, nonfarm payrolls and durable goods orders – having the largest impact. Furthermore, announcements that reflect an unexpected improvement in the economy tend to have a negative impact on gold and silver prices; however, they tend to have a positive effect on copper prices. In comparison, realized volatility and volume for all three metals are positively influenced by economic news. Finally, there is evidence that several news announcements exert an asymmetric impact on market activity variables.

The relationship between information arrival and asset price movements is of central importance to price formation and price discovery in financial markets, and is a topic that has been extensively investigated in the literature. For instance, the mixture of distributions model relies on “news” to explain movements in asset returns (see Tauchen and Pitts, 1983). Among the various sources of information the role of public information is frequently examined because they are easily identifiable, and also carry implications for the canonical model of weakly efficient markets which posits that security prices reflect all available information. Chen et al. (1986) highlight the importance of macroeconomic factors such as industrial production and measures of unanticipated inflation on stock returns. In a related study, Flannery and Protopapadakis (2002) relate equity returns to macroeconomic variables. More recently, research attention has shifted to an examination of intra-day data that provides additional insights into the market microstructure behavior relating to trading and pricing variables. As a case in point, Adams et al. (2004) show that unanticipated inflationary news has distinguishable effects on intra-day equity returns, and that most of this information is incorporated within minutes of the news release. This paper explores the price formation process and trading volume activity in the metals futures market around the release of new macroeconomic information. Four important questions are addressed. First, what is the impact of macroeconomic news on the return, realized volatility and volume of gold, silver and copper futures? Second, does the release of macroeconomic news affect the three metals in different ways? Third, how long does it take for the impact of macroeconomic news shocks to be fully absorbed by the market? Finally, does the metals market respond asymmetrically to the release of unexpected macroeconomic news? The answers to these questions are important for several reasons. First, our analysis is based on high-frequency intra-day data, which allows us to detect patterns of market reaction that may not be easily discerned in lower frequency daily data. In this regard, it is important to point out that the empirical literature using daily data finds only mixed or relatively weak evidence of the link between macroeconomic announcements and commodity prices (see Roache and Rossi, 2010; Hess et al., 2008), thus lending support to the argument that, unlike other assets, commodity prices are predetermined with respect to US macroeconomic aggregates such as real output, consumption and investment variables (Kilian and Vega, 2010). Therefore, an investigation of how an important class of commodities, specifically metals, responds to macroeconomic news at intra-day frequencies provides a meaningful contrast with existing studies. Based on Andersen’s (1996) argument that different types of news may have different stochastic arrival processes and therefore convey varying impacts on pricing behavior, we evaluate the impact of 19 different types of macroeconomic news. These announcements are sorted by the time of each news release, with the aim of identifying those announcement times that have the largest impact on the metals market. Furthermore, taking into account evidence from related asset markets such as equities (Koutmos and Booth, 1995), we also examine whether or not metal futures respond asymmetrically to economic news. We contend that a study of how metal futures prices react to positive versus negative economic surprises would be informative not only in terms of market efficiency and information processing, but may also provide an explanation as to why previous studies that do not account for potential asymmetries may have been unsuccessful in documenting a significant relationship between economic news and commodity prices. Second, compared to financial assets, there is a relative paucity of studies that examine the role of information in the metal futures markets. This is especially noteworthy considering that in recent years there has been a steady increase in the amount of investor attention given to these markets. In general, the popularity of commodities, and in particular metals, stems from the belief that these assets act as a hedge against inflation, offer valuable diversification opportunities to investors, serve as a monetary medium during times of market uncertainty, and have a wide range of manufacturing and industrial applications. It is therefore not entirely surprising that these products are one of the most heavily traded in organized futures exchanges. Finally, our study is comprehensive in scope in that it evaluates the responsiveness of several market activity variables including return, volatility and trading volume. We construct a realized volatility measure that accounts for intra-day price information within each particular time interval. Therefore, to summarize, our sample period, research design, and empirical methods allow us to investigate more fully the high-frequency dynamics of three important commodities. The remainder of the paper is organized as follows. In the next two sections we provide a literature review and discuss theoretical considerations, respectively. Section 4 explains the data sources, summary statistics and cleaning procedures. Section 5 provides a brief description of the research design and empirical methods, and discusses the results. Section 6 concludes the paper.

The question of whether commodity prices are influenced by macroeconomic fundamentals or whether they are predetermined with respect to monetary aggregates is a topic that has yet to find consensus. Although there are strong economic reasons to expect commodity prices to be sensitive to macroeconomic fundamentals, prior empirical studies have not been very successful in documenting a strong relationship. We suggest that this might be due to several reasons including measurement of data at low (daily or monthly) frequencies and research methods that do not control for asymmetric impacts. We undertake a comprehensive examination of the response of return, volatility and trading volume for gold, copper and silver to 19 different types of macroeconomic announcements. Notably, we rely on an improved analytical framework that uses high-frequency data for a sample period spanning 7 years from 2002 to 2008. The announcements are classified by time and direction (i.e., whether the news release signals better-than-expected-economic news or worse-than-expected news) in order to: (a) identify the most important set of announcements; (b) trace the persistence of macroeconomic shocks; and (c) allow for asymmetry in the relationships. Our analysis reveals that news releases have a strong and instantaneous impact on all three metals. First, the 8:30 set of announcements appear to have the largest impact on prices, realized volatility and trading volume during the immediate post-announcement 5 min interval. For instance, univariate regression models show that surprises in nonfarm payroll (released at 8:30 am) explain about 35% and 23% of the returns of gold and silver, respectively during the 8:30–8:35 time interval. Copper returns, on the hand, are more sensitive to the 9:15 set of announcements which include capacity utilization and industrial production. The evidence suggests that the behavior of commodity markets is quite similar to other asset markets in terms of their responsiveness to economic information. Our results for volume are most consistent with informed trading models where informed traders tend to have an advantage in processing, rather than predicting, new information. Second, our results indicate that the metals market respond in an economically predictable manner. Unexpected improvement in economic growth has a negative impact on gold and silver prices, but registers a favorable effect on copper returns. For instance, improvements in real economic activity (e.g., advance retail sales), consumption (e.g., new home sales) and investment (e.g., durable goods orders) are found to negatively influence gold and silver prices. The performance of copper, on the other hand, is consistent with its status as an important industrial metal that benefits primarily from unexpected growth in the economy. Finally, our evidence indicates that the effect of macroeconomic news dissipates quickly, within about 60 min of the news release, and notably, and that several announcements have an asymmetric impact on market activity variables. These results provide an insightful contrast to previous studies that use daily data to examine the relationship between macroeconomic news and commodity prices.