Volume 19, Issue 4

This manuscript lists questions from a video interview with Jim Stock at the Norges Bank on 15 March 2015. The interview covers topics ranging from time series analysis and weak instruments to climate policy.

In this paper, we suggest new cointegration tests that can become more powerful in the presence of non-normal errors. Non-normal errors will not pose a problem in usual cointegration tests even when they are ignored. However, we show that they can become useful sources to improve the power of the tests when we use the “residual augmented least squares” (RALS) procedure to make use of nonlinear moment conditions driven by non-normal errors. The suggested testing procedure is easy to implement and it does not require any non-linear estimation techniques. We can exploit the information on the non-normal error distribution that is already available but ignored in the usual cointegration tests. Our simulation results show significant power gains over existing cointegration tests in the presence of non-normal errors.

We develop a present-value asset pricing model with an econometrically useful representation that accommodates a plethora of stylized assumptions about beliefs. Using 20th century S&P500 data we use our model to compare the empirical fit of belief assumptions associated with rational expectations, asymmetic information, learning, behavioral effects and evolution. Among these, asymmetric information with evolution is particularly useful both in terms of statistical criteria and in terms of ability to explain the equity premium, excess volatility and predictability of returns. Our work suggests that popular relaxations of rationality can easily lead to econometric representations that may be impossible to work with in empirical research. Furthermore, replication of stylized facts may be too weak a requirement when evaluating such models. Fortunately, there exist simple relaxations of rationality that are sufficient to drastically improve the empirical fit of models with full rationality.

Intelligent pattern recognition imposes new challenges in high-frequency financial data mining due to its irregularities and roughness. Based on the wavelet transform for decomposing systematic patterns and noise, in this paper we propose a new integrated wavelet denoising method, named smoothness-oriented wavelet denoising algorithm (SOWDA), that optimally determines the wavelet function, maximal level of decomposition, and the threshold rule by using a smoothness score function that simultaneously detects the global and local extrema. We discuss the properties of our method and propose a new evaluation procedure to show its robustness. In addition, we apply this method both in simulation and empirical investigation. Both the simulation results based on three typical stylized features of financial data and the empirical results in analyzing high-frequency financial data from Frankfurt Stock Exchange confirm that SOWDA significantly (based on the RMSE comparison) improves the performance of classical econometric models after denoising the data with the discrete wavelet transform (DWT) and maximal overlap discrete wavelet transform (MODWT) methods.

We use noncausal autoregressions to examine the persistence properties of quarterly US consumer price inflation from 1970:1 to 2012:2. These nonlinear models capture the autocorrelation structure of the inflation series as accurately as their conventional causal counterparts, but they allow for persistence to depend on the size and sign of shocks to inflation as well as the inflation rate. Inflation persistence has decreased since the early 1980s, after which persistence is also greater following small and negative shocks than large and positive ones. At high levels of inflation, shocks are absorbed more slowly before the early 1980s and faster thereafter compared to low levels of inflation.

In this paper we introduce a triple-threshold leverage stochastic volatility (TTLSV) model for financial return time series. The main feature of the model is to allow asymmetries in the leverage effect as well as mean and volatility. In the model the asymmetric effect is modeled by a threshold nonlinear structure that the two regimes are determined by the sign of the past returns. The model parameters are estimated using maximum likelihood (ML) method based on the efficient importance sampling (EIS) technique. Monte Carlo simulations are presented to examine the accuracy and finite sample properties of the proposed methodology. The EIS-based ML (EIS-ML) method shows good performance according to the Monte Carlo results. The proposed model and methodology are applied to two stock market indices for China. Strong evidence of the mean and volatility asymmetries is detected in Chinese stock market. Moreover, asymmetries in the volatility persistence and leverage effect are also discovered. The log-likelihood and Akaike information criterion (AIC) suggest evidence in favor of the proposed model. In addition, model diagnostics suggest that the proposed model performs relatively well in capturing the key features of the data. Finally, we compare models in a Value at Risk (VaR) study. The results show that the proposed model can yield more accurate VaR estimates than the alternatives.

We estimate the dynamic daily dependence between assets by applying the Semiparametric Copula-Based Multivariate Dynamic (SCOMDY) model on intraday data. Using tick data of three stock returns of the period before and during the credit crisis, we find that our dependence estimator better captures the steep increase in dependence during the onset of the crisis as compared to other commonly used time-varying copula methods. Like other high-frequency estimators, we find that the dependence estimator exhibits long memory and forecast it using a HAR model. We show that for out-of-sample forecasts, our dependence estimator performs better than the constant estimator and other commonly used time-varying copula dependence estimators.