Re-examining the Sources of Real Exchange Rate Fluctuations: A Rational Expectations Structural VAR Approach

為了探討理性預期下小型開放經濟實質匯率波動的來源，本文提出「理性預期結構性VAR模型」的分析方式。一個外生干擾除了對實質匯率有著理論上的直接影響，它也可能會對預測自己和其他的外生變數有幫助，因此產生各種間接的影響，而「理性預期結構性VAR模型」的分析方式可以將這些直接與間接效果一起限制於VAR模型的係數之上。經過最大概似法的估計與檢定，加拿大、法國、義大利與日本均無法拒絕本文之理論模型。實證結果顯示：國內私人部門的自發性需求是法國與義大利最重要的實質匯率波動來源，貨幣衝擊是加拿大實質匯率波動的主因，而來自國外的衝擊則是日本長期實質匯率波動的最重要原因。同時政府支出衝擊並不具有顯著之說明能力；生產力的衝擊只有在法國存在微小但顯著的解釋力。To explicitly take into account market expectations for real exchange rate fluctuations, this paper proposes a rational expectations structural VAR (RE-SVAR) method to decompose structural shocks. An exogenous shock not only has effects on real exchange rates directly, but also has indirect effects through the change in predictions for other fundamental variables. The RE-SVAR method imposes all these direct and indirect channels to decompose the sources of real exchange rate fluctuations. We find that the over-identifying restrictions implied by our model cannot be rejected for Canada, France, Italy, and Japan. Our results also indicate that private expenditure shocks are the most important source of real exchange rate variations for France and Italy, monetary shocks are most important for Canada, and foreign price shocks are most important in the long run for Japan. Moreover, supply shocks have a small but significant explanatory power fot real exchange rate variations only in France