Forecasting the South African economy

A DSGE-VAR approach

G. Liu, R. Gupta, E. Schaling

    Research output: Contribution to journalArticleScientificpeer-review

    Abstract

    Purpose
    – This paper aims to develops an estimable hybrid model that combines the micro‐founded DSGE model with the flexibility of the atheoretical VAR model.

    Design/methodology/approach
    – The model is estimated via the maximum likelihood technique based on quarterly data on real gross national product (GNP), consumption, investment and hours worked, for the South African economy, over the period of 1970:1 to 2000:4. Based on a recursive estimation using the Kalman filter algorithm, the out‐of‐sample forecasts from the hybrid model are then compared with the forecasts generated from the Classical and Bayesian variants of the VAR for the period 2001:1‐2005:4.

    Findings
    – The results indicate that, in general, the estimated hybrid‐DSGE model outperforms the classical VAR, but not the Bayesian VARs in terms of out‐of‐sample forecasting performances.

    Research limitations/implications
    – The model lacks nominal shocks and needs to be extended into a small open economy framework.

    Practical implications
    – The paper was able to show that, even though the DSGE model is outperformed by the BVAR, a microfounded theoretical DSGE model has a future in forecasting the South African economy.

    Originality/value
    – To the best of the authors' knowledge, this is the first attempt to use an estimable DSGE model to forecast the South African economy.
    Original languageEnglish
    Pages (from-to)181-195
    JournalJournal of Economic Studies
    Volume37
    Issue number2
    Publication statusPublished - 2010

    Fingerprint

    Africa
    DSGE models
    Dynamic stochastic general equilibrium
    Hybrid model
    VAR model
    Maximum likelihood
    Bayesian vector autoregression
    Forecasting performance
    Kalman filter
    Bayesian VAR
    Gross national product
    Small open economy
    Hours worked
    Recursive estimation
    Design methodology

    Cite this

    Liu, G., Gupta, R., & Schaling, E. (2010). Forecasting the South African economy: A DSGE-VAR approach. Journal of Economic Studies, 37(2), 181-195.
    Liu, G. ; Gupta, R. ; Schaling, E. / Forecasting the South African economy : A DSGE-VAR approach. In: Journal of Economic Studies. 2010 ; Vol. 37, No. 2. pp. 181-195.
    @article{33c9cab2ee7446dda1998cebb77a6b92,
    title = "Forecasting the South African economy: A DSGE-VAR approach",
    abstract = "Purpose– This paper aims to develops an estimable hybrid model that combines the micro‐founded DSGE model with the flexibility of the atheoretical VAR model.Design/methodology/approach– The model is estimated via the maximum likelihood technique based on quarterly data on real gross national product (GNP), consumption, investment and hours worked, for the South African economy, over the period of 1970:1 to 2000:4. Based on a recursive estimation using the Kalman filter algorithm, the out‐of‐sample forecasts from the hybrid model are then compared with the forecasts generated from the Classical and Bayesian variants of the VAR for the period 2001:1‐2005:4.Findings– The results indicate that, in general, the estimated hybrid‐DSGE model outperforms the classical VAR, but not the Bayesian VARs in terms of out‐of‐sample forecasting performances.Research limitations/implications– The model lacks nominal shocks and needs to be extended into a small open economy framework.Practical implications– The paper was able to show that, even though the DSGE model is outperformed by the BVAR, a microfounded theoretical DSGE model has a future in forecasting the South African economy.Originality/value– To the best of the authors' knowledge, this is the first attempt to use an estimable DSGE model to forecast the South African economy.",
    author = "G. Liu and R. Gupta and E. Schaling",
    note = "Appeared earlier as CentER DP 2008-32",
    year = "2010",
    language = "English",
    volume = "37",
    pages = "181--195",
    journal = "Journal of Economic Studies",
    issn = "0144-3585",
    publisher = "Emerald Group Publishing Ltd.",
    number = "2",

    }

    Liu, G, Gupta, R & Schaling, E 2010, 'Forecasting the South African economy: A DSGE-VAR approach', Journal of Economic Studies, vol. 37, no. 2, pp. 181-195.

    Forecasting the South African economy : A DSGE-VAR approach. / Liu, G.; Gupta, R.; Schaling, E.

    In: Journal of Economic Studies, Vol. 37, No. 2, 2010, p. 181-195.

    Research output: Contribution to journalArticleScientificpeer-review

    TY - JOUR

    T1 - Forecasting the South African economy

    T2 - A DSGE-VAR approach

    AU - Liu, G.

    AU - Gupta, R.

    AU - Schaling, E.

    N1 - Appeared earlier as CentER DP 2008-32

    PY - 2010

    Y1 - 2010

    N2 - Purpose– This paper aims to develops an estimable hybrid model that combines the micro‐founded DSGE model with the flexibility of the atheoretical VAR model.Design/methodology/approach– The model is estimated via the maximum likelihood technique based on quarterly data on real gross national product (GNP), consumption, investment and hours worked, for the South African economy, over the period of 1970:1 to 2000:4. Based on a recursive estimation using the Kalman filter algorithm, the out‐of‐sample forecasts from the hybrid model are then compared with the forecasts generated from the Classical and Bayesian variants of the VAR for the period 2001:1‐2005:4.Findings– The results indicate that, in general, the estimated hybrid‐DSGE model outperforms the classical VAR, but not the Bayesian VARs in terms of out‐of‐sample forecasting performances.Research limitations/implications– The model lacks nominal shocks and needs to be extended into a small open economy framework.Practical implications– The paper was able to show that, even though the DSGE model is outperformed by the BVAR, a microfounded theoretical DSGE model has a future in forecasting the South African economy.Originality/value– To the best of the authors' knowledge, this is the first attempt to use an estimable DSGE model to forecast the South African economy.

    AB - Purpose– This paper aims to develops an estimable hybrid model that combines the micro‐founded DSGE model with the flexibility of the atheoretical VAR model.Design/methodology/approach– The model is estimated via the maximum likelihood technique based on quarterly data on real gross national product (GNP), consumption, investment and hours worked, for the South African economy, over the period of 1970:1 to 2000:4. Based on a recursive estimation using the Kalman filter algorithm, the out‐of‐sample forecasts from the hybrid model are then compared with the forecasts generated from the Classical and Bayesian variants of the VAR for the period 2001:1‐2005:4.Findings– The results indicate that, in general, the estimated hybrid‐DSGE model outperforms the classical VAR, but not the Bayesian VARs in terms of out‐of‐sample forecasting performances.Research limitations/implications– The model lacks nominal shocks and needs to be extended into a small open economy framework.Practical implications– The paper was able to show that, even though the DSGE model is outperformed by the BVAR, a microfounded theoretical DSGE model has a future in forecasting the South African economy.Originality/value– To the best of the authors' knowledge, this is the first attempt to use an estimable DSGE model to forecast the South African economy.

    M3 - Article

    VL - 37

    SP - 181

    EP - 195

    JO - Journal of Economic Studies

    JF - Journal of Economic Studies

    SN - 0144-3585

    IS - 2

    ER -