新凯恩斯宏观模型的符号限制：“准不可知”识别过程的结果（英）

Sign Restrictions with aNew-Keynesian MacroModel Results From a “Quasi-Agnostic”Identification Procedure Gregorio Impavido WP/25/162 IMF Working Papersdescribe research inprogress by the author(s) and are published toelicit comments and to encourage debate.The views expressed in IMF Working Papers arethose of the author(s) and do not necessarilyrepresent the views of the IMF, its Executive Board,or IMF management. 2025AUG IMF Working Paper Middle East and Central Asia Department Sign Restrictions with a New-Keynesian Macro Model:Results From a “Quasi-Agnostic” Identification ProcedurePrepared by Gregorio Impavido* Authorized for distribution by Ali Al-EydAugust 2025 IMF Working Papersdescribe research in progress by the author(s) and are published to elicitcomments and to encourage debate.The views expressed in IMF Working Papers are those of theauthor(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management. ABSTRACT:This paper proposes a “quasi-agnostic” sign restriction procedure to identify structural shocks infrequentist structural vector autoregression (SVAR) models. It argues that low acceptance rates, inherent toagnostic sign restriction procedures, are not necessarily an indication of model misspecification. They can below because agnostic procedures fail to exploit theex antepriors on the sign of responses of macro variablesto structural shocks. RECOMMENDED CITATION:Impavido, G. (2025) “Sign Restrictions with a New-Keynesian Macro Model:Results From a “Quasi-Agnostic” Identification Procedure”.IMF Working PaperNoWP/2025/162 Sign Restrictions with a New-Keynesian Macro Model Results from a “quasi-agnostic” identificationprocedure Prepared by Gregorio Impavido1 Table of Contents Page 1Introduction12The baseline model23Imposing agnostic sign restrictions54From agnostic to “quasi-agnostic” sign restrictions65Conclusions9 References10 List of Figures 1Stability condition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122Structural impulse responses - recursive model . . . . . . . . . . . . . . . . .133Contemporaneous sign restrictions on IRFs using the SRC procedure 1/ . . .144Ex ante restrictions on structural parameters . . . . . . . . . . . . . . . . . .145Distribution of restricted parameters using alternative identification procedures156Results from “quasi agnostic” procedure /1 . . . . . . . . . . . . . . . . . . .15 List of Tables 1Lag Selection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122Granger causality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123Sign restrictions for positive structural shocks. . . . . . . . . . . . . . . . .13 1Introduction In the early literature on structural vector autoregression (SVAR) models, policy and macroe-conomic shocks were identified through parametric restrictions, predominantly zero restric-tions, on the immediate responses of macroeconomic aggregates to these shocks. Contempo-raneous zero restrictions continue to be justified often on the grounds that some variables aresluggish to adjust to new information or that new information is observed with a delay. Forinstance, a tightening of monetary policy could lead to a temporary decline in real economicactivity, albeit with some delay. More recent studies have sought to relax these identificationassumptions by employing sign restrictions. Unlike parametric restrictions, which necessitatespecific coefficient values derived from prior knowledge of the structure of the economy, signrestrictions simply stipulate the direction of a shock’s impact on macroeconomic aggregatesbut are agnostic about the specific structure of the economy.Hence, procedures used toidentify shocks, are often referred to as “agnostic” identification procedures. Prominent applications of sign restrictions to study monetary policy effectiveness includepapers like Astveit et al. (2017), Canova and De Nicoló (2002), Elekdag and Han (2015), Faust(1998), and Uhlig (2005). Notably, Uhlig (ibid.) finds that contractionary monetary policyshocks do not exhibit a discernible effect on real GDP, suggesting that the neutrality of moneyis consistent with empirical data. Subsequent refinements to Uhlig’s identification approach(Arias et al. 2019) and the use of identification based on external instruments (Gertler andKaradi 2015) have tended to reaffirm the conventional view that monetary policy exertsa short-term influence on GDP, as, for instance, predicted by models incorporating pricerigidities à la (say) Calvo (1983). This paper proposes a “quasi-agnostic” identification procedure to overcome the low accep-tance rates typical of agnostic identification procedures. Agnostic identification proceduresbased on sign restrictions typically yield a low proportion of simulated structural shocks thatsatisfy the imposed sign restrictions (acceptance rates). For instance, Fisher and Huh (2020),find acceptance rates of less than 1 per