Robin L. Lumsdaine, Consistency and Asymptotic Normality of the Quasi-Maximum Likelihood Estimator in IGARCH(1,1) and Covariance Stationary GARCH(1,1) Models, Econometrica, Vol. 64, No. 3 (May, 1996), pp. 575-596

This paper develops methods for constructing asymptotically valid confidence intervals for the date of a single break in multivariate time series, including I(0), I(1), and deterministically trending regressors. Although the width of the asymptotic confidence interval does not decrease as the sample size increases, it is inversely related to the number of series which have a common break date, so there are substantial gains to multivariate inference about break dates. These methods are applied to two empirical examples: the mean growth rate of output in three European countries, and the mean growth rate of U.S. consumption, investment, and output.

Ever since Nelson and Plosser (1982) found evidence in favor of the unit-root hypothesis for 13 long-term annual macro series, observed unit - root behavior has been equated with persistence in the economy. Perron (1989) questioned this interpretation, arguing instead that the "observed" behavior may indicate failure to account for structural change. Zivot and Andrews (1992) restored confidence in the unit-root hypothesis by incorporating an endogenous break point into the specification. By allowing for the possibility of two endogenous break points, we find more evidence against the unit-root hypothesis than Zivot and Andrews, but less than Perron.

We examine the reaction of daily Treasury bond prices to the release of U.S. macroeconomic news. These news releases (of employment and producer price index data) are of interest because they are released on periodic, preannounced dates and because they are associated with substantial bond market volatility. We investigate whether these nonautocorrelated announcements give rise to autocorrelated volatility. We find that announcement-day volatility does not persist at all, consistent with the immediate incorporation of information into prices. We also find a risk premium on these release dates.

Regulatory changes that appear comprehensive will have little impact on the functioning of a developing market if they fail to lead to foreign portfolio inflows. We specify a reduced-form model for a number of financial time series and search for a common, endogenous break in the data generating process. We also estimate a confidence interval for the break. Our endogenous break dates are accurately estimated but do not always correspond closely to dates of official capital market reforms. Indeed, the endogenous dates are usually later than official dates, highlighting the important distinction between market liberalization and market integration.

We develop metrics to assess the complexity of a bank holding company (BHC), based on its ownership structure. Large BHCs have intricate ownership hierarchies involving hundreds or even thousands of legal entities that may contribute to increased operational risk and greater opacity. Our measures are mathematically grounded, intuitive, and easy to implement. They may be particularly useful in the context of resolution, where regulators often face significant time pressure and coordination challenges. We use regulatory filing data from the Federal Reserve to validate the measures, demonstrating that they provide a useful complement to balance sheet information in assessing BHC complexity. Notably, the proposed measures are highly correlated with existing complexity indicators that are not based on organizational structure and are less correlated with size than these existing complexity measures. We show that the proposed measures provide additional explanatory power for the regulatory indicators, even after controlling for size.