The estimation and forecast of the volatility matrix are two of the main tasks of financial econometrics since they are essential ingredients in many practical applications. Unfortunately the use of classical multivariate methods in large dimensions is difficult because of the curse of dimensionality. We present a general semiparametric technique, based on functional gradient descent (FGD) and able to overcome most problems associated with a multivariate GARCH-type estimation. By testing the accuracy of the volatility estimates for the measurement of market risk on real data we provide empirical evidence of the strong predictive potential of the FGD approach, also in comparison to other standard methods.

We develop a multivariate dynamic term structure model, which takes into account the nonlinear (time-varying) relation between interest rates and the state of the economy. In contrast to the classical term structure literature, in which nonlinearities are captured by increasing the number of latent state variables or by latent regime shifts, in our no-arbitrage framework the regimes are governed by thresholds and are directly linked to economic fundamentals. Specifically, starting from a simple monetary policy model for the short rate, we introduce a parsimonious and tractable model for the yield curve, which takes into account the possibility of regime shifts in the behavior of the Federal Reserve. In our empirical analysis, we show the merit of our approach three dimensions: interpretable bond dynamics, accurate short end yield curve pricing, and yield curve implications.

The fast-growing literature on news analytics provides evidence that financial markets are partially driven by sentiments. In contrast with previous studies that have almost exclusively focused on the direct effects of the news related to single companies or sectors, we investigate the time-varying dynamics of news’ cross-industry influences for a set of US and European stocks over a period of 10 years. The graphical Granger causality of the news sentiments-excess return networks is estimated by applying the adaptive lasso. We find significant spillover effects and show the importance of sentiments related to certain sectors for the whole cross-section of stocks.

We propose a new approach based on a generalization of the logit model to improve prediction accuracy in U.S. bank failures. Mixed-data sampling (MIDAS) is introduced in the context of a logistic regression. We also mitigate the class-imbalance problem in data and adjust the classification accuracy evaluation. In applying the suggested model to the period from 2004 to 2016, we show that it correctly classifies significantly more bank failure cases than the classic logit model, in particular for long-term forecasting horizons. Some of the largest recent bank failures in the United States that had been previously misclassified are now correctly predicted.