Sparse structures with LASSO through principal components: Forecasting GDP components in the short-run

This paper aims to examine the use of sparse methods to forecast the real, in the chain-linked volume sense, expenditure components of the US and EU GDP in the short-run sooner than the national institutions of statistics officially release the data. We estimate current quarter nowcasts along with 1- and 2-quarter forecasts by bridging quarterly data with available monthly information announced with a much smaller delay. We solve the high-dimensionality problem of the monthly dataset by assuming sparse structures of leading indicators, capable of adequately explaining the dynamics of analyzed data. For variable selection and estimation of the forecasts, we use the sparse methods - LASSO together with its recent modifications. We propose an adjustment that combines LASSO cases with principal components analysis that deemed to improve the forecasting performance. We evaluate forecasting performance conducting pseudo-real-time experiments for gross fixed capital formation, private consumption, imports and exports over the sample of 2005-2019, compared with benchmark ARMA and factor models. The main results suggest that sparse methods can outperform the benchmarks and to identify reasonable subsets of explanatory variables. The proposed LASSO-PC modification show further improvement in forecast accuracy.

[1]  Kurt Graden Lunsford,et al.  Forecasting residential investment in the United States , 2015 .

[2]  Julius Stakenas,et al.  Generating short-term forecasts of the Lithuanian GDP using factor models , 2012 .

[3]  P. Bickel,et al.  SIMULTANEOUS ANALYSIS OF LASSO AND DANTZIG SELECTOR , 2008, 0801.1095.

[4]  A. Belloni,et al.  Square-Root Lasso: Pivotal Recovery of Sparse Signals via Conic Programming , 2011 .

[5]  Gonzalo Camba-Mendez,et al.  Short-Term Forecasts of Euro Area GDP Growth , 2008, SSRN Electronic Journal.

[6]  Peng Zhao,et al.  On Model Selection Consistency of Lasso , 2006, J. Mach. Learn. Res..

[7]  Katja Heinisch,et al.  Bottom-up or direct? Forecasting German GDP in a data-rich environment , 2018 .

[8]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[9]  A. Belloni,et al.  Least Squares After Model Selection in High-Dimensional Sparse Models , 2009, 1001.0188.

[10]  F. Ziegelmann,et al.  LASSO‐Type Penalties for Covariate Selection and Forecasting in Time Series , 2016 .

[11]  Zi Zhen Liu,et al.  The D oubly Adaptive LASSO Methods for Time Series Analysis , 2014 .

[12]  Francis X. Diebold,et al.  Comparing Predictive Accuracy, Twenty Years Later: A Personal Perspective on the Use and Abuse of Diebold–Mariano Tests , 2012 .

[13]  Rob J Hyndman,et al.  Automatic Time Series Forecasting: The forecast Package for R , 2008 .

[14]  H. Zou.  The Adaptive Lasso and Its Oracle Properties , 2006 .

[15]  R. Tibshirani,et al.  Sparse Principal Component Analysis , 2006 .

[16]  Jianqing Fan,et al.  Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties , 2001 .

[17]  C. Granger.  Long memory relationships and the aggregation of dynamic models , 1980 .

[18]  Tommaso Proietti,et al.  Dating Business Cycles : a Methodological Contribution with an Application to the Euro Area 2 0 0 3 E D IT IO N , 2003 .

[19]  Hussein Hazimeh,et al.  Fast Best Subset Selection: Coordinate Descent and Local Combinatorial Optimization Algorithms , 2018, Oper. Res..

[20]  Serena Ng,et al.  Determining the Number of Primitive Shocks in Factor Models , 2007 .

[21]  D. Giannone,et al.  Now-Casting and the Real-time Data Flow , 2012, SSRN Electronic Journal.

[22]  R. Tibshirani,et al.  Least angle regression , 2004, math/0406456.

[23]  Cun-Hui Zhang,et al.  Adaptive Lasso for sparse high-dimensional regression models , 2008 .

[24]  Victor Chernozhukov,et al.  On cross-validated Lasso in high dimensions , 2020 .

[25]  J. Stock,et al.  Macroeconomic Forecasting Using Diffusion Indexes , 2002 .

[26]  Brandyn Bok,et al.  Macroeconomic Nowcasting and Forecasting with Big Data , 2017 .

[27]  Halbert White,et al.  Tests of Conditional Predictive Ability , 2003 .

[28]  David H. Small,et al.  Nowcasting: the real time informational content of macroeconomic data releases , 2008 .

[29]  D. Cox,et al.  An Analysis of Transformations , 1964 .

[30]  P. Vermeulen,et al.  Business cycle duration dependence and foreign recessions , 2018, Scottish Journal of Political Economy.

[31]  The duration of economic expansions and recessions: More than duration dependence , 2010 .

[32]  J. Bai,et al.  Determining the Number of Factors in Approximate Factor Models , 2000 .

[33]  Norman R. Swanson,et al.  Mining Big Data Using Parsimonious Factor, Machine Learning, Variable Selection and Shrinkage Methods , 2016 .

[34]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[35]  Domenico Giannone,et al.  Comparing Alternative Predictors Based on Large‐Panel Factor Models , 2006 .

[36]  Wenjiang J. Fu,et al.  Asymptotics for lasso-type estimators , 2000 .

[37]  J. Bai,et al.  Forecasting economic time series using targeted predictors , 2008 .

[38]  Massimiliano Marcellino,et al.  Forecasting economic activity with targeted predictors , 2015 .

[39]  Eduardo F. Mendes,et al.  L_1-Regularization of High-Dimensional Time-Series Models with Flexible Innovations , 2015 .