Investigating Economic Uncertainty Using Stochastic Volatility in Mean VARs: The Importance of Model Size, Order-Invariance and Classification

Sharada Davidson, Chenghan Hou and Gary Koop
Additional contact information
Sharada Davidson: Department of Economics, University of Strathclyde
Chenghan Hou: Hunan University, Center for Economics, Finance and Management Studies, China

No 2306, Working Papers from University of Strathclyde Business School, Department of Economics

Abstract: Stochastic Volatility in Mean Vector Autoregressions (SVMVARs) are popularly used to jointly estimate macroeconomic and financial uncertainty and their economic effects. However, SVMVARs are computationally demanding. To ease the computational burden, existing approaches limit the number of variables included, adopt a speci cation which is not invariant to the way the variables are ordered and require the researcher to classify each variable as macroeconomic or financial before estimation. To overcome these limitations, we develop an efficient Markov Chain Monte Carlo (MCMC) algorithm for SVMVARs which are large, order-invariant and have unclassified ed variables. For each unclassified ed variable, the algorithm determines the appropriate classi cation at each point in time. We demonstrate the importance of these extensions using a large SVMVAR with over 40 U.S. variables, 16 of which are treated as unclassifi ed. We show that smaller SVMVARs overestimate the economic effects of macroeconomic uncertainty, failing to reveal that fi nancial uncertainty plays a larger role. When using large SVMVARs, however, different orderings yield conflicting results and it becomes critical to use an order-invariant speci cation. We also fi nd that most unclassifi ed variables change classi cation over time with changes often occurring during crisis periods.

Keywords: Large VAR; Uncertainty; Stochastic Volatility; Order Invariance (search for similar items in EconPapers)
JEL-codes: C11 C52 C55 E32 E44 (search for similar items in EconPapers)
Pages: pages
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:str:wpaper:2306

Access Statistics for this paper

More papers in Working Papers from University of Strathclyde Business School, Department of Economics Contact information at EDIRC.
Bibliographic data for series maintained by Kirsty Hall ().