Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

The University of British Columbia

$31,000.00

Inference for long cycles in economic data

EF 2024-25

Résumé
Recurrent boom-and-bust cycles are a salient feature of economic data. The durations of booms and busts have important policy implications, and measuring the cycle length is a long-standing problem in statistics and economics. Stochastic cycles in economic data are often long: they span substantial fractions of observed data and are highly persistent. The project seeks to develop a statistical methodology for inference on the length of cycles in persistent multivariate time series data when only a small number of peaks and troughs are observed. Further, the project will develop a detection, estimation, and inference methodology for steady-state relationships between cyclical processes. In a recent paper "Modeling Long Cycles" (with N. Kang, accepted for publication in the Journal of Econometrics), we showed that existing methods for inference on the length of cycles are distorted when cycles are long. We developed a new inferential procedure for the periodicity of cycles specifically designed to account for the possibility of long, persistent stochastic cycles. Our procedure produces a confidence interval for the cycle length and has the correct asymptotic size uniformly. In other words, it is uniformly valid regardless of whether the cycle is long or short. A significant limitation of that paper is that it was restricted to univariate time series only. This proposal seeks to extend the methodology to multivariate time series and presents a long-term research program that will result in novel statistical methods for analyzing cyclical processes. The generalization is far from a mechanical extension of the previous paper, as multivariate time series processes can have more complex dynamics. Most importantly, cyclical behavior may arise due to interactions between variables and thus shed light on deeper data-generating mechanisms behind data. The proposal will result in several academic papers that will be submitted for publication to leading journals in statistics and econometrics. I will start by developing a uniformly valid inference procedure for the cycle length in multivariate processes. The next part of the program is concerned with detecting common long cycles in multivariate processes, their estimation, and inference. Identifying such cycles will enable examining long-run relationships between variables, particularly their impacts on each other in the steady state. A longer-term objective of the proposal is to study the implications of the proposed long-cycle framework for leading indicators that play an essential role in forecasting and planning. The proposed approach presents a novel probabilistic framework for analyzing the data-generating process behind leading, coincident, and lagging indicators and developing novel superior indicators. The long-term goal of this research program is to merge the long-cycle framework with structural approaches to improve our understanding of the nature and causes of business cycles.
Programme
Discovery Grants Program - Individual
Emplacement
British Columbia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
RGPIN-2024-04877
Name nom
Marmer, Vadim
Department département
Vancouver School of Economics
Organization id
2
Institution établissement
The University of British Columbia
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
31000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
1200
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Advancement of knowledge
Area of application fr
Avancement des connaissances
Research subject code
3002
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Time series analysis
Research subject fr
Analyse des séries chronologiques
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Inference for long cycles in economic data
Keywords
time series; nonstationary time series; local-to-unity modeling; business cycles; cyclical processes; vector autoregressions; stochastic cycles; robust inference; uniform inference; statistical inference
Application summary
Recurrent boom-and-bust cycles are a salient feature of economic data. The durations of booms and busts have important policy implications, and measuring the cycle length is a long-standing problem in statistics and economics. Stochastic cycles in economic data are often long: they span substantial fractions of observed data and are highly persistent. The project seeks to develop a statistical methodology for inference on the length of cycles in persistent multivariate time series data when only a small number of peaks and troughs are observed. Further, the project will develop a detection, estimation, and inference methodology for steady-state relationships between cyclical processes. In a recent paper "Modeling Long Cycles" (with N. Kang, accepted for publication in the Journal of Econometrics), we showed that existing methods for inference on the length of cycles are distorted when cycles are long. We developed a new inferential procedure for the periodicity of cycles specifically designed to account for the possibility of long, persistent stochastic cycles. Our procedure produces a confidence interval for the cycle length and has the correct asymptotic size uniformly. In other words, it is uniformly valid regardless of whether the cycle is long or short. A significant limitation of that paper is that it was restricted to univariate time series only. This proposal seeks to extend the methodology to multivariate time series and presents a long-term research program that will result in novel statistical methods for analyzing cyclical processes. The generalization is far from a mechanical extension of the previous paper, as multivariate time series processes can have more complex dynamics. Most importantly, cyclical behavior may arise due to interactions between variables and thus shed light on deeper data-generating mechanisms behind data. The proposal will result in several academic papers that will be submitted for publication to leading journals in statistics and econometrics. I will start by developing a uniformly valid inference procedure for the cycle length in multivariate processes. The next part of the program is concerned with detecting common long cycles in multivariate processes, their estimation, and inference. Identifying such cycles will enable examining long-run relationships between variables, particularly their impacts on each other in the steady state. A longer-term objective of the proposal is to study the implications of the proposed long-cycle framework for leading indicators that play an essential role in forecasting and planning. The proposed approach presents a novel probabilistic framework for analyzing the data-generating process behind leading, coincident, and lagging indicators and developing novel superior indicators. The long-term goal of this research program is to merge the long-cycle framework with structural approaches to improve our understanding of the nature and causes of business cycles.