چکیده:
هدف این مقاله گسترش مدلهای جدید سری زمانی چند متغیره پویای غیر ساختاری و ارزیابی عملکرد انواع این مدلهای جایگزین در پیش بینی تورم می باشد. روش شبه بیزی با اطلاعات Literman prior در چارچوب یک مدل خود رگرسیونی برداری برای اقتصاد ایران در دوره زمانی 2006- 1981جهت ارزیابی عملکرد مدلهای مختلف در طول افق زمانی متفاوت بکار گرفته شده است. همچنین به منظور محدود نمودن میانگین تغییر تورم به مقدار صفر، تبدیل Bewley جهت تصریح مجدد مدل خود رگرسیونی برداری استفاده شده است. نتایج نشان می دهد که کاربرد تبدیل Bewley در مدل خود رگرسیونی برداری مورد استفاده، باعث شده که دقت پیش بینی این مدلها نسبت به مدلهای خود رگرسیونی برداری بیزی کلاسیک بهبود یابد.
This paper focuses on the development of modern non-structural dynamic multivariate time series models and evaluating performance of various alternative specifications of these models for forecasting Iranian inflation. The Quasi-Bayesian method، with Literman prior، is applied to Vector autoregressive (VAR) model of the Iranian economy from 1981:Q2 to 2006:Q1 to assess the forecasting performance of different models over different forecasting horizons. The Bewley transformation is also employed for the re-parameterization of the VAR models to impose the mean of the change of inflation to zero. Applying the Bewley (1979) transformation to force the drift parameter of change of inflation to zero in the VAR model improves forecast accuracy in comparison to the traditional BVAR.
خلاصه ماشینی:
"Applying the Bewley(1979) transformation to force the drift parameter of change ofinflation to zero in the VAR model improves forecast accuracy incomparison to the traditional BVAR.
Introduction This paper investigates different Vector Autoregressive (VAR) specifications to improve the Iranian inflation forecasting by non- structural dynamic multivariate time series models.
Our results show that applying the Bewley (1979) transformation to impose a zero mean to the change of inflation, provides more accurate forecasts of inflation for the Iranian economy in comparison to the Traditional BVAR model.
1 The VAR model described in this article includes the logarithm of real GDP as a measure of real output, y, the first difference of the logarithm of the GDP deflator as a measure of inflation2, Pr, the logarithm of liquidity as a monetary variable, M2, and the first difference of the logarithm of the black market exchange rate, Exc. As the longer lags may raise the chance of over-fitting and thus lead to poor out-of-sample forecasting, the lag length specification in a UVAR model is another important step in constructing a UVAR model.
In summary, our results show that, using Bewley (1979) transformation to force the mean of the change of inflation rate to zero in a mixed drift VAR model accretes forecasts of Iranian inflation in comparison to the BVAR model with Litterman's prior.
The results show that using Bewley (1979) transformation to impose a zero mean to the change of inflation provides more accurate forecasts of inflation for the Iranian economy in comparison to the BVAR models with Litterman's priors."