Forecasting cointegrated series with BVAR models

Amisano, Gianni; Serati, Massimiliano

doi:10.1002/(sici)1099-131x(199912)18:7<463::aid-for732>3.0.co;2-d

Cited by 10 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We follow the work of Amisano and Serati (1999) and give some informative prior to the Γ and β (factor loadings coefficients) matrices. Restrictions are imposed to the estimated adjustments terms using the results of the Johansen system cointegration test [4] .…”

Section: The Bayesian Vecm Frameworkmentioning

confidence: 99%

“…To overcome the presence of unit roots in our time series, first differences are estimated for each country such as X t ¼ x t −x t−1 with x t is the CDS (or the bond) spread at time t. The overtime CDS and bond first-difference plots [5] clearly show that changes are stationary. CDS and bond spreads changes exhibit a relatively similar time-varying evolution dynamic.…”

mentioning

confidence: 99%

See 1 more Smart Citation

International risk spillover in sovereign credit markets: an empirical analysis

Sabkha

Peretti

Hmaied

2019

View full text Add to dashboard Cite

Purpose The purpose of this paper is to study the volatility spillover among 33 worldwide sovereign Credit Default Swap (CDS) markets and their underlying bond markets. Design/methodology/approach In contrast to prior studies, the authors incorporate heteroscedasticity, asymmetric leverage effects and long-memory features of sovereign credit spreads simultaneously through a bivariate FIEGARCH model and a Bayesian cointegrated vector autoregressive model. Findings Similar to the literature, the findings confirm that strong evidence of credit risk spillover between credit markets is accentuated during two recent crisis periods. However, the country-by-country analysis indicates that countries exhibit different sensitivity levels and divergent reactions to financial shocks. Further, the authors show that the bidirectional interrelationship evolves over time and across countries emphasizing the necessity of time-varying national regulatory policies and trading positions. Originality/value Based on a large data set that covers the recent two financial crises and using complex methods, the work focuses on sovereign tensions that have repercussions on banks’ solvency and refinancing conditions. Yet, the study is a hot topic since that during crisis periods in the financial markets, direct and indirect interconnections increase between sovereign risk and banking risk. Using new econometric approaches, the results show that each country exhibits a different behavior toward the credit risk which is relevant to both portfolio managers and policy makers. The time-varying spillover effects detected between markets are an accurate indicator of financial stability, allowing policy makers to put in place personalized economic policies. On the other hand, markets’ participants could take advantages of the results by adjusting their trading and hedging positions on the dynamic co-movements. The findings reveal, as well, that the sovereign crisis has more weakened the global financial and banking system than the subprime crisis. The authors previously tackled the cross-country contagion phenomenon in the CDS markets, and this manuscript builds on the prior study to enhance the obtained results.

show abstract

Section: The Bayesian Vecm Frameworkmentioning

confidence: 99%

mentioning

confidence: 99%

International risk spillover in sovereign credit markets: an empirical analysis

Sabkha

Peretti

Hmaied

2019

View full text Add to dashboard Cite

show abstract

“…Numerous studies have provided evidence that the Bayesian variant of the VAR model yields forecasting improvements relative to the standard VAR (LeSage, 1990; Shoesmith, 1992; Joutz et al , 1995; Amisano and Serati, 1999). Consequently, the BVAR provides a useful benchmark for assessing the forecasting qualities of the combined Okun and Phillips model.…”

Section: Evaluation Of the Forecastsmentioning

confidence: 99%

A latent variable approach to forecasting the unemployment rate

Chua

Lim

Tsiaplias

2011

Journal of Forecasting

View full text Add to dashboard Cite

A forecasting model for unemployment is constructed that exploits the time series properties of unemployment while satisfying the economic relationships specifi ed by Okun's law and the Phillips curve. In deriving the model, we jointly consider the problem of obtaining estimates of the unobserved potential rate of unemployment consistent with Okun's law and the Phillips curve, and associating the potential rate of unemployment with actual unemployment. The empirical example shows that the model clearly outperforms alternative forecasting procedures typically used to forecast unemployment. 1 The ARMA(2, 1) is chosen since it accommodates the possibility of cyclical behavior in u c t . Alogoskoufi s and Stengos (1991) fi nd that the US unemployment rate is appropriately modelled as an AR(2) process with ARCH errors.

show abstract

“…Two different sets of priors are considered: a tight prior and a loose prior. The tight prior tends to emphasize the model's long-run 3Shoesmith (1995) fi nds that the short-and long-term forecasts from the BECM outperform those from BVAR and VAR models, whileChow and Choy (2006) fi nd that the BVAR performs better than the BECM in forecasting the global electronics cycle Amisano and Serati (1999). advocate the use of the Bayesian variant of the ECM on the basis that it allows for the super-consistent estimation of the model's long-run parameters.…”

mentioning

confidence: 99%

Can consumer sentiment and its components forecast Australian GDP and consumption?

Chua

Tsiaplias

2009

Journal of Forecasting

View full text Add to dashboard Cite

This paper examines whether the disaggregation of consumer sentiment data into its sub-components improves the real-time capacity to forecast GDP and consumption. A Bayesian error correction approach augmented with the consumer sentiment index and permutations of the consumer sentiment sub-indices is used to evaluate forecasting power. The forecasts are benchmarked against both composite forecasts and forecasts from standard error correction models. Using Australian data, we find that consumer sentiment data increase the accuracy of GDP and consumption forecasts, with certain components of consumer sentiment consistently providing better forecasts than aggregate consumer sentiment data. Copyright © 2009 John Wiley & Sons, Ltd.

show abstract

Forecasting cointegrated series with BVAR models

Cited by 10 publications

References 20 publications

International risk spillover in sovereign credit markets: an empirical analysis

International risk spillover in sovereign credit markets: an empirical analysis

A latent variable approach to forecasting the unemployment rate

Can consumer sentiment and its components forecast Australian GDP and consumption?

Contact Info

Product

Resources

About