Towards an early warning system for sovereign defaults leveraging on machine learning methodologies

In this study, we address the topic of credit risk stemming from central governments from a technical point of view. First, we explore various econometric and machine learning techniques to build an enhanced sovereign rating system that effectively differentiates the risk of default among countries. Our empirical results indicate that the machine learning method of XGBOOST has a superior out-of-sample and out-of-time predictive performance. Then, we use the models developed to calibrate a sovereign rating system and provide useful insights into the set-up of a parsimonious early warning system. Our results provide a more concise view of the most robust method for classifying countries’ default risk with significant regulatory implications, given that the efficient assessment of sovereign debt is crucial for effective proactive risk measurement.     

Forecasting using variational Bayesian inference in large vector autoregressions with hierarchical shrinkage

Many recent papers in macroeconomics have used large vector autoregressions (VARs) involving 100 or more dependent variables. With so many parameters to estimate, Bayesian prior shrinkage is vital to achieve reasonable results. Computational concerns currently limit the range of priors used and render difficult the addition of empirically important features such as stochastic volatility to the large VAR. In this paper, we develop variational Bayesian methods for large VARs that overcome the computational hurdle and allow for Bayesian inference in large VARs with a range of hierarchical shrinkage priors and with time-varying volatilities. We demonstrate the computational feasibility and good forecast performance of our methods in an empirical application involving a large quarterly US macroeconomic data set.     

Machine learning for US cross-industry return predictability under information uncertainty

This paper investigates the association between industry information uncertainty and cross-industry return predictability using machine learning in a general predictive regression framework. We show that controlling for post-selection inference and performing multiple tests improves the in-sample predictive performance of cross-industry return predictability in industries characterized by high uncertainty. Ordinary least squares post-least absolute shrinkage and selection operator models incorporating lagged industry information uncertainty for the financial and commodity industries are critical to improving prediction performance. Furthermore, in-sample industry return forecasts establish heterogeneous predictability over US industries, in which excess returns are more predictable in sectors with medium or low uncertainty.     

Time-variation,multiple testing,and the factor zoo

Bayesians circumvent the need for significance threshold correction when multiple testing and we recommend controlling the Type-S (sign), rather than the Type-1, error rate because it yields more reliable frequency properties for inferences. Our unified Bayesian framework, with theory-informed priors, identifies two breaks (2001 and 2008) in our 1980–2018 sample period. After each break the set of characteristics changes, and only market beta is selected in all regimes. In a portfolio application, the method generates significantly larger Sharpe ratios after transaction costs than a range of benchmark methods, including the same model that uses a Type-1 (not Type-S) error framework.     

Improving risk classification and ratemaking using mixture-of-experts models with random effects

In the underwriting and pricing of nonlife insurance products, it is essential for the insurer to utilize both policyholder information and claim history to ensure profitability and proper risk management. In this paper, we apply a flexible regression model with random effects, called the Mixed Logit-weighted Reduced Mixture-of-Experts, which leverages both policyholder information and their claim history, to categorize policyholders into groups with similar risk profiles, and to determine a premium that accurately captures the unobserved risks. Estimates of model parameters and the posterior distribution of random effects can be obtained by a stochastic variational algorithm, which is numerically efficient and scalable to large insurance portfolios. Our proposed framework is shown to outperform the classical benchmark models (Logistic and Lognormal GL(M)M) in terms of goodness-of-fit to data, while offering intuitive and interpretable characterization of policyholders' risk profiles to adequately reflect their claim history.     

Cross-sector comovements and policy impact in the COVID-19 stock market: A dynamic factor approach

This study investigates the remarkable comovements in U.S. equity returns during the COVID-19 pandemic. It constructs a dynamic factor model (DFM) to illuminate the sources of the comovements and their implications. Using the Markov Chain Monte Carlo (MCMC) estimation method, the study finds that the comovements had a weak daily oscillation pattern during the pandemic. With that pattern, the study also finds significant monetary policy effects on the equity returns of several key sectors. In addition, it estimates the impact of news shocks, including monetary policy news, fiscal stimulus news, and unemployment news, on cross-sector equity returns. For any given sector, the conventional and unconventional monetary policy news shocked the sector in opposite directions. Among the positive monetary news shocks, the strongest were from interest rate policy surprises. Conversely, fiscal stimulus news had the most substantial positive impact and triggered all sectors to rebound from the bear market at the end of March 2020. Furthermore, by applying Natural Language Processing (NLP) sentiment analysis, this study sheds light on the positive correlation between comovements and news sentiment.