Empirical modelling of survey-based expectations for the design of economic indicators in five European regions

Abstract

This is a post-peer-review, pre-copyedit version of an article published in Empirica. The final authenticated version is available online at: http://dx.doi.org/10.1007/s10663-017-9395-1”.In this study we use agents’ expectations about the state of the economy to generate indicators of economic activity in twenty-six European countries grouped in five regions (Western, Eastern, and Southern Europe, and Baltic and Scandinavian countries). We apply a data-driven procedure based on evolutionary computation to transform survey variables in economic growth rates. In a first step,\ud we design five independent experiments to derive a formula using survey variables that best replicates the evolution of economic growth in each region by means of genetic programming, limiting the integration schemes to the main mathematical operations. We then rank survey variables according to their performance in tracking economic activity, finding that agents’ ‘‘perception about the overall\ud economy compared to last year’’ is the survey variable with the highest predictive power. In a second step, we assess the out-of-sample forecast accuracy of the evolved indicators. Although we obtain different results across regions, Austria, Slovakia, Portugal, Lithuania and Sweden are the economies of each region that show the best forecast results. We also find evidence that the forecasting performance of the survey-based indicators improves during periods of higher growth.This is a post-peer-review, pre-copyedit version of an article published in Empirica. The final authenticated version is available online at: http://dx.doi.org/10.1007/s10663-017-9395-1”.Peer ReviewedPostprint (author's final draft

Artificial Intelligence in Physical Sciences: Symbolic Regression Trends and Perspectives

“…Early studies have focused on the recognition of those interactions in large datasets that contain different observations of several economic quantities [ 228 ], while succeeding models are centered in the prediction of crude oil price [ 229 ] and economic growth forecasting by investigating expectations of agents. Agents’ expectations regarding the economy’s condition are high-valued for economic modelling due to the fact that they contain explicit, multi-variate information about market [ 230 ] and usually obtained via tendency surveys (business and consumer surveys) [ 231 ]. As a result, approaches are made via SR to form a link between survey data and a successful economic growth model [ 227 , 232 ].…”

The Economics of Walking About and Predicting Unemployment

“…The plethora of data available to forecast and nowcast unemployment rates means analysts have spent increasing time on what is the optimal set of indicators in maximising the accuracy of predictions. In their work Claveria and colleagues (Claveria et al, 2017;Claveria et al, 2019a;Claveria et al, 2019b) use evolutionary computation techniques (a sub-field of Artificial Intelligence) to optimise their unemployment expectations metrics, as well as showing that the degree of correspondence in unemployment expectations across consumers also contains information increasing the predictive power of models estimating unemployment rates (Claveria, 2019a;Claveria, 2019b). 19 There is also a very sophisticated literature, some of which is reviewed above, identifying the predictive power of models, usually based on out-of-sample prediction, accounting for serial correlation, the identification of structural breaks in series and other issues.…”

Review on behavioral economics and behavioral finance

“…To prevent episodes of market failure in any crisis, Breitenfellner and Wagner (2010) develop a formal illustration of a rescue mechanism to distinguish between illiquid but solvent and insolvent banks. Readers may refer to Claveria et al (2019) for other economic indicators.…”