Trading activity in the Treasury futures market and its role in futures price fluctuations

Does trading by hedgers or speculators destabilize the Treasury futures market? And if it is the case, how do they destabilize the market? I examine in Chapter 1 the empirical relation of volume and volatility---which is central to the market microstructure literature---in the context of Treasury futures trading. Vector autoregressive (VAR) analysis is conducted on the 2-, 5-, 10-, and 30-year Treasury futures contracts traded at the Chicago Board of Trade. With some mixed results, it can be roughly concluded that speculators destabilize the Treasury futures market, causing a more turbulent trading pattern as evident in the increased price volatility. Significance relation can not be said of the hedgers and the price volatility; available evidence at best suggests a weak relation between hedging activity and a decreased price volatility (indicating a market being stabilized). To my knowledge, this study is the first in applying the VAR technique to the context of Treasury futures trading. It is also the first in examining the volume and open interest by constructing two different trading activity series ("aggregate" and "active contract" amounts) in the same study, and the results are compared. On the volume-volatility relation, it is among the few studies that explicitly focus on individual contract instead of an all-as-one approach. The long period of data (from year 1991 to 2006) is applied to the VAR framework. In addition, GARCH volatility specifications are comprehensively tested and the GARCH(1,1) volatility specification---commonly-used in the currency futures market---is conveniently arrived at, so possible cross-market comparisons may be fruitful for future applications.;I review in Chapter 2 the classes of models in theorizing the volume-volatility relation. The strategic trading model of private information by Kyle (1985) and "differences of opinion" model of public information by Harris and Raviv (1993) are discussed. I propose that the "differences of opinion" models are more suitable in studying Treasury futures trading, since virtually all diligent analysts are exposed to the same set of public data (such as Federal Open Market Committee announcements, "Fed watchers" studied guesses, or quarterly GDP growth rates and monthly nonfarm payrolls for general economic outlook), but they nonetheless infer their interest rate expectations with different interpretations of data. Unlike the cases in stock or corporate bond markets where differential accesses to private information are essential, differences in interpretation are more significant in Treasury futures trading and private information is less distinctive for individual contracts. The main predictions of the Harris and Raviv model are compared with my VAR results from Chapter 1; it is especially notable that the positively autocorrelated volume pattern is confirmed by the surprisingly strong VAR evidence. I then conclude with the recent episode (09/2006) of "market squeeze" warnings to financial institutions (essentially primary dealers) in the Treasury market. Implications on Treasury futures trading in terms of regulating the market are discussed.