Jiahui Wang

Inference on a Structural Parameter in Instrumental Variables Regression with Weak Instruments

In this paper we consider the problem of making inference on a structural parameter in instrumental variables regression when the instruments are only weakly correlated with the endogenous explanatory variables.

A Bayesian Time Series Model of Multiple Structural Changes in Level, Trend, and Variance

We consider a deterministically trending dynamic time series model in which multiple structural changes in level, trend, and error variance are modeled explicitly and the number, but not the timing, of the changes is known. Estimation of the model is made possible by the use of the Gibbs sampler. The determination of the number of structural breaks and the form of structural change is considered as a problem of model selection, and we compare the use of marginal likelihoods, posterior odds ratios, and Schwarzs Bayesian model-selection criterion to select the most appropriate model from the data. We evaluate the efficacy of the Bayesian approach using a small Monte Carlo experiment. As empirical examples, we investigate structural changes in the U.S. ex post real interest rate and in a long time series of U.S. real gross domestic product.

Inference on Structural Parameters In Instrumental Variables Regression With Weak Instruments

The authors consider the problem of making asymptotically valid inference on structural parameters in instrumental variables regression with weak instruments. Using local-to-zero asymptotics, they derive the asymptotic distributions of likelihood ratio (LR) and Lagrange multiplier (LM) type statistics for testing simple hypotheses on structural parameters based on maximum likelihood and generalized methods of moments estimation methods. In contrast to the nonstandard limiting behavior of Wald statistics, the limiting distributions of certain LR and LM statistics are bounded by a chi-square distribution with degrees of freedom given by the number of instruments.

Vector Autoregressive Models for Multivariate Time Series

The vector autoregression (VAR) model is one of the most successful, flexible, and easy to use models for the analysis of multivariate time series. It is a natural extension of the univariate autoregressive model to dynamic multivariate time series. The VAR model has proven to be especially useful for describing the dynamic behavior of economic and financial time series and for forecasting. It often provides superior forecasts to those from univariate time series models and elaborate theory-based simultaneous equations models. Forecasts from VAR models are quite flexible because they can be made conditional on the potential future paths of specified variables in the model.

Rolling Analysis of Time Series

A rolling analysis of a time series model is often used to assess the model’s stability over time. When analyzing financial time series data using a statistical model, a key assumption is that the parameters of the model are constant over time. However, the economic environment often changes considerably, and it may not be reasonable to assume that a model’s parameters are constant. A common technique to assess the constancy of a model’s parameters is to compute parameter estimates over a rolling window of a fixed size through the sample. If the parameters are truly constant over the entire sample, then the estimates over the rolling windows should not be too different. If the parameters change at some point during the sample, then the rolling estimates should capture this instability.

Unit Root Tests

Many economic and financial time series exhibit trending behavior or non-stationarity in the mean. Leading examples are asset prices, exchange rates and the levels of macroeconomic aggregates like real GDP. An important econometric task is determining the most appropriate form of the trend in the data. For example, in ARMA modeling the data must be transformed to stationary form prior to analysis. If the data are trending, then some form of trend removal is required.

Multivariate GARCH Modeling

When modeling multivariate economic and financial time series using vector autoregressive (VAR) models, squared residuals often exhibit significant serial correlation. For univariate time series, Chapter 7 indicates that the time series may be conditionally heteroskedastic, and GARCH models have been proved to be very successful at modeling the serial correlation in the second order moment of the underlying time series.

State Space Models

The state space modeling tools in S+FinMetrics are based on the algorithms in SsfPack 3.0 developed by Siem Jan Koopman and described in Koopman, Shephard and Doornik (1999, 2001)1. SsfPack is a suite of C routines for carrying out computations involving the statistical analysis of univariate and multivariate models in state space form. The routines allow for a variety of state space forms from simple time invariant models to complicated time-varying models. Functions are available to put standard models like ARMA and spline models in state space form. General routines are available for filtering, smoothing, simulation smoothing, likelihood evaluation, forecasting and signal extraction. Full details of the statistical analysis is provided in Durbin and Koopman (2001). This chapter gives an overview of state space modeling and the reader is referred to the papers by Koopman, Shephard and Doornik for technical details on the algorithms used in the S+FinMetrics/SsfPack functions.