Lasso Regressions And Forecasting Models In Applied Stress Testing

Model selection and forecasting in stress tests can be facilitated using machine learning techniques. These techniques have proved robust in other fields for dealing with the curse of dimensionality, a situation often encountered in applied stress testing. Lasso regressions, in particular, are well suited for building forecasting models when the number of potential covariates is large, and the number of observations is small or roughly equal to the number of covariates. This paper presents a conceptual overview of lasso regressions, explains how they fit in applied stress tests, describes its advantages over other model selection methods, and illustrates their application by constructing forecasting models of sectoral probabilities of default in an advanced emerging market economy.

The Summer 2017 issue of the IMF Research Bulletin highlights new research such as recent IMF Working Papers and Staff Discussion Notes. The Research Summaries are “Structural Reform Packages, Sequencing, and the Informal Economy (by Zsuzsa Munkacsi and Magnus Saxegaard) and “A Broken Social Contract, Not High Inequality Led to the Arab Spring” (by Shantayanan Devarajan and Elena Ianchovichina). The Q&A section features “Seven Questions on Fintech” (by Tommaso Mancini-Griffoli). The Bulletin also includes information on recommended titles from IMF Publications and the latest articles from the IMF Economic Review.

Economic forecasting is a key ingredient of decision making both in the public and in the private sector. Because economic outcomes are the result of a vast, complex, dynamic and stochastic system, forecasting is very difficult and forecast errors are unavoidable. Because forecast precision and reliability can be enhanced by the use of proper econometric models and methods, this innovative book provides an overview of both theory and applications. Undergraduate and graduate students learning basic and advanced forecasting techniques will be able to build from strong foundations, and researchers in public and private institutions will have access to the most recent tools and insights. Readers will gain from the frequent examples that enhance understanding of how to apply techniques, first by using stylized settings and then by real data applications--focusing on macroeconomic and financial topics. This is first and foremost a book aimed at applying time series methods to solve real-world forecasting problems. Applied Economic Forecasting using Time Series Methods starts with a brief review of basic regression analysis with a focus on specific regression topics relevant for forecasting, such as model specification errors, dynamic models and their predictive properties as well as forecast evaluation and combination. Several chapters cover univariate time series models, vector autoregressive models, cointegration and error correction models, and Bayesian methods for estimating vector autoregressive models. A collection of special topics chapters study Threshold and Smooth Transition Autoregressive (TAR and STAR) models, Markov switching regime models, state space models and the Kalman filter, mixed frequency data models, nowcasting, forecasting using large datasets and, finally, volatility models. There are plenty of practical applications in the book and both EViews and R code are available online at authors' website.

Discover New Methods for Dealing with High-Dimensional DataA sparse statistical model has only a small number of nonzero parameters or weights; therefore, it is much easier to estimate and interpret than a dense model. Statistical Learning with Sparsity: The Lasso and Generalizations presents methods that exploit sparsity to help recover the underl

We compared the predictive performance of a series of machine learning and traditional methods for monthly CDS spreads, using firms’ accounting-based, market-based and macroeconomics variables for a time period of 2006 to 2016. We find that ensemble machine learning methods (Bagging, Gradient Boosting and Random Forest) strongly outperform other estimators, and Bagging particularly stands out in terms of accuracy. Traditional credit risk models using OLS techniques have the lowest out-of-sample prediction accuracy. The results suggest that the non-linear machine learning methods, especially the ensemble methods, add considerable value to existent credit risk prediction accuracy and enable CDS shadow pricing for companies missing those securities.

As real estate forms a significant part of the asset portfolios of most investors and lenders, it is crucial that analysts and institutions employ sound techniques for modelling and forecasting the performance of real estate assets. Assuming no prior knowledge of econometrics, this book introduces and explains a broad range of quantitative techniques that are relevant for the analysis of real estate data. It includes numerous detailed examples, giving readers the confidence they need to estimate and interpret their own models. Throughout, the book emphasises how various statistical techniques may be used for forecasting and shows how forecasts can be evaluated. Written by a highly experienced teacher of econometrics and a senior real estate professional, both of whom are widely known for their research, Real Estate Modelling and Forecasting is the first book to provide a practical introduction to the econometric analysis of real estate for students and practitioners.

Applied Predictive Modeling covers the overall predictive modeling process, beginning with the crucial steps of data preprocessing, data splitting and foundations of model tuning. The text then provides intuitive explanations of numerous common and modern regression and classification techniques, always with an emphasis on illustrating and solving real data problems. The text illustrates all parts of the modeling process through many hands-on, real-life examples, and every chapter contains extensive R code for each step of the process. This multi-purpose text can be used as an introduction to predictive models and the overall modeling process, a practitioner’s reference handbook, or as a text for advanced undergraduate or graduate level predictive modeling courses. To that end, each chapter contains problem sets to help solidify the covered concepts and uses data available in the book’s R package. This text is intended for a broad audience as both an introduction to predictive models as well as a guide to applying them. Non-mathematical readers will appreciate the intuitive explanations of the techniques while an emphasis on problem-solving with real data across a wide variety of applications will aid practitioners who wish to extend their expertise. Readers should have knowledge of basic statistical ideas, such as correlation and linear regression analysis. While the text is biased against complex equations, a mathematical background is needed for advanced topics.

Bayesian and Frequentist Regression Methods provides a modern account of both Bayesian and frequentist methods of regression analysis. Many texts cover one or the other of the approaches, but this is the most comprehensive combination of Bayesian and frequentist methods that exists in one place. The two philosophical approaches to regression methodology are featured here as complementary techniques, with theory and data analysis providing supplementary components of the discussion. In particular, methods are illustrated using a variety of data sets. The majority of the data sets are drawn from biostatistics but the techniques are generalizable to a wide range of other disciplines.