Preparations and Prerequisites for the Introduction of Inflation Targeting in Romania Presentation by Deputy Governor Cristian Popa National Bank of Romania NBR-BoE BoE Conference on Inflation Targeting: International Experience and Romania s s Prospects
Outline 1. Recap: Context and Initial Specifications 2. Internal Preparations 2.1 Organizational issues 2.2 The Policy-Making Process 2.3 Forecasting 2.4 Communication issues 3. Prerequisites for Inflation Targeting 3.1 Absence of Fiscal Dominance 3.2 International Comparisons 4. Steps Ahead
1. Recap: Context and Initial Specifications Inflation targeting mentioned in Romania s s Medium-Term Strategy for Economic Development (2000), confirmed in PEPs (2002 and beyond) Initial specifications: (i) target specified in terms of CPI; (ii) mid-point with band (some tolerance for undershooting target); (iii) annual targets w. multi-period perspective (initially two years); (iv) flexible interpretation of inflation targeting (esp. coexistence w. managed float regime); (v) existence of few ex ante defined exceptions conditional to attaining the target; (vi) joint announcement of target with government
2.1 Organizational Issues Task force set up to implement the IT framework, led by coordinating DG Monetary Policy Department Research and Publications Department Macroeconomic Modeling and Forecasting Department Development of a detailed quarterly macroeconomic forecasting framework Centralized database Data monitoring & reporting Expectations survey Near-term forecasting of key variables Medium-term forecasting model, consistent with achieving multi-period inflation targets
DATA FOR THE CORE MODEL NATIONAL BANK OF ROMANIA Preparing the Introduction of Inflation Targeting Data management Satellite models NTFs for domestic exogenous variables (RPD, MPD and MMFD) Satellite models (MMFD) Expert judgment DETAILED MACROECONOMIC DATA raw historical data for domestic and external variables updated on a regular basis near-term forecasts for external variables NEAR-TERM FORECASTS (NTF) domestic and external variables exogenous for the core model CURRENT POSITION OF THE ECONOMY (initial conditions for the core model) equilibrium values for relevant variables and their medium term profile Central NBR database External sources Consensus Forecasts Cut- off date The core model (MAMTF) 3 rd party forecasts (surveys) QUARTERLY MODEL PROJECTIONS inflation and relevant variables Conference on Inflation Targeting
2.2 The Policy-Making Process Role of the NBR Board Exceptional initial involvement in model specification, starting conditions (in order to foster ownership/understanding: successful dialogue with the model ) Final decision-making body Focuses on how to best employ forecasts in policy making In practical terms: identifies alternative scenarios to consider or special issues to be examined from the viewpoint of strategy ( big( decisions ); these are then run through model again for consistency judges how the path for the policy interest rate should be adjusted to account for uncertainties and risks
Role of the Monetary Policy Committee (MPC) Institutionalized analytical forum at staff/executive management level Critical focus on ensuring that all relevant information is brought to bear on the forecasts In practical terms: responsible for reviewing assumptions on initial conditions for the forecasts, including the outlook for exogenous variables reviewing the projections and formulating policy recommendations for the Board, including analyses of their technical implications
Time T-57 T-51 T-34 T-27 T-23 T-18 T-15 T-10 T T+3 T+6 Quarterly Forecasting & Decision Schedule Event Task Force First issues meeting Task Force Second issues meeting MPC regular monthly meeting Board regular monthly meeting Task Force meeting Task Force meeting MPC meeting Task Force meeting MPC monthly meeting Board meeting Inflation Report MPC meeting Board meeting Model meeting Start of Forecasting Cycle Other projection details Policy review and recommendation Policy review and decision Near-Term Forecasts (NTF) Final projection meeting Forecasts and risk scenarios Monetary Policy Decision Details Initial conditions, equilibrium conditions and exogenous variables Initial conditions, equilibrium conditions, exogenous variables and NTF Forecasts and Inflation Report sent to MPC Discussion of forecasting cycle and Inflation Report Discussion & Eventual Adoption of Inflation Report
2.3 Forecasting Based on formalized models and expert judgment Two types of modeling approaches: Estimation approach at the short-run run horizon Calibration approach at the medium-term horizon Final forecast integrates information from short-term term models, medium-term model and expert judgment
2.3.1 Near-Term Forecasting Framework Two-quarter horizon at present A set of methods (one model approach is risky) Economic theory as a basis of analysis, but applied to statistical data (estimation vs. calibration) Tool for simulating the impact of changes in exogenous variables Used for analysis and for establishing the initial conditions for the QFM
Near-term Inflation Forecasting Two methods: Expert judgement forecast for CPI components ARMAX model for the main CPI inflation components (yoy( data, monthly frequency) Food Non-food excluding regulated and fuel prices Services excluding regulated prices Exogenous variables: EUR/ROL and EUR/USD exchange rates, net wages, regulated prices, fuel prices The forecast presented to the MPC = weighted average of the results provided by the two methods (with 0.65 weight given to the ARMAX model)
Near-term GDP Forecasting Small semi-structural structural model (qoq( qoq) Behavioural equations for household consumption, gross fixed capital formation, exports, imports and core inflation Exogenous variables: EUR/ROL and EUR/USD exchange rates, public consumption, net wages, regulated prices, fuel prices, 3M Euribor,, euro area GDP, HICP OLS and 2SLS estimation with error correction mechanism
History 2.3.2 Medium-Term Forecasting Assembling the Model for Analysis and Medium-Term Forecasting (MAMTF) started in mid-2004 Significant progress achieved, with technical assistance support from two IMF missions and bilateral exchanges/expert visits with the Czech National Bank (MAMTF conceived in similar fashion to Czech QPM) Two forecasting dry runs to test MAMTF adequacy and the functional integration of short-term term and expert information; currently in second dry run First-pass calibration of MAMTF based on priors discussed by the Bank s s Board
General characteristics of MAMTF Model open to continuous improvement, while maintaining the core structure Small semi-structural structural calibrated model with a New-Keynesian core (ST and MT non-neutrality) neutrality) Economy assumed to converge to well-defined and stable long-run equilibrium Deviations from trends the gaps - reflect cyclical behavior of the economy, while trends are analogous to steady-state state values (primarily assessed on inflation neutrality)
Core structure of MAMTF Core inflation determined by its structural persistence, agents expectations, output gap, and import price inflation Output gap or excess demand factor determined by its own persistence and real monetary conditions, incl. gaps of real (bank) interest rates and real exchange rate Exchange rate determined according to uncovered interest parity relationship including a risk premium Monetary policy behavior (as implied by flexible inflation targeting) ting) described by a forward-looking policy interest rate rule that reacts to deviations of inflation i from target, the output gap and excessive interest rate volatility Agents expectations modeled as hybrids of backward-looking (inertial) and forward-looking ( model( model-consistent ) ) expectations
Specifics of the MAMTF transmission mechanism Low sacrifice ratio (but may vary over time, according to progress of disinflation) Expectations channel quite significant for monetary policy direct influence on inflation Exchange rate channel also important, especially for speed of disinflation (two channels: direct, via pass-through; indirect, through influences on real sector and asset portfolios) Aggregate activity (output gap) channel relatively less effective e at present, but crucial for ensuring continuity of disinflation process Monetary policy impulses intermediated by relatively sluggish bank deposit and lending rates
Increasing role of policy rate in influencing inflation Response of core inflation to monetary policy induced shocks Sample: June 1997 - December 2000 Response of core inflation to monetary policy induced shocks Sample: September 2000 - November 2004 Response to a real interest rate gap shock Response to a real interest rate gap shock.008.002.001.004.000 -.001.000 -.002 -.003 -.004 -.004 -.005 -.008 Months 2 4 6 8 10 12 14 16 18 20 22 24 -.006 Months 2 4 6 8 10 12 14 16 18 20 22 24 90 80 Variance decomposition of core inflation Sample: June 1997 - December 2000 Contributions of shocks (% of total variance) 70 60 Variance decomposition of core inflation Sample: September 2000 - November 2004 Contribution of shocks (% of total variance) 70 60 50 50 40 40 30 30 20 10 0 5 10 15 20 Months 20 10 0 5 10 15 20 Months Output gap Real interest rate gap Real exchange rate gap Core inflation Output gap Real interest rate gap Real exchange rate gap Core inflation
2.3.3 The forecasting process Role of near-term forecasting Covers short end of forecast horizon Shapes the initial conditions of the forecast (In the future) Serves as cross-check check for final forecast Role of expert judgment Flexibility of MAMTF allows direct incorporation of expert input (incl. survey-based information) Forecasts of effects of anticipated exogenous events (e.g. capital al account liberalization) Forecasts of variables not explicitly modeled (e.g. fiscal stimulus) Model forecasts can be tuned if unrealistic Role of medium-term model Integrates all information in a consistent way Generates an interest rate path which can serve as policy guideline ine Can be used to generate risk scenarios, analyze consistency of alternative a scenarios
Flow of information in the forecasting process at the National Bank of Romania Near-term models and expert forecast Assessment of initial conditions and medium-term trends Near-term forecast NTF Inflation, GDP, EUR/ROL etc. Trends & Gaps Medium-term (core) model Final medium-term forecast Exogenous variables forecasts Anticipated shocks, fiscal impulse, etc. Tuning Uncertainty Expert judgment
2.4 Communication issues Restructuring of the existing inflation report Policy made more transparent Supporting external communication
2.4.1 The New Inflation Report Plays the central role in the communication strategy under IT To be published quarterly, reflecting the forecasting cycle; first quarterly report (2005Q1) expected July, also serving to emphasize multi-period preoccupation for attaining the inflation targets Based on the same information set used in the forecasting process Compared to existing 6-month 6 Inflation Reports: (i) more timely publishing; (ii) transparent inclusion of NBR inflation forecast,, policy decision, detailed risk assessment; (iii) balance favors forward-looking analysis over backward-looking reporting; (iv) will replace a substantial portion of analysis currently included in NBR Annual Reports, some streamlining of monthly bulletins also a result Unpublished draft inflation reports for the two dry-run run forecasting rounds, for purposes of internal analysis
2.4.2 Increased Policy Transparency Increased frequency and quality of communication with public on policy issues Prior announcement of calendar for policy-related Board meetings Press releases on contents of between-forecast meetings and on the monetary policy decisions Prior announcement of confirmation of fully-fledged fledged implementation of IT regime, with details on bandwidth, formalized exceptions, etc.
2.4.3 Supporting external communication The NBR website will be re-designed to allow easier access to policy related information and to include background information on inflation targeting Dedicated briefings on benefits of IT and the related policy process to be organized Brochure series targeted for broader audience Background papers to be published Conferences, interviews, public speeches
3. Prerequisites for Inflation Targeting 8 4 0-4 -8 3.1 Absence of Fiscal Dominance (dep var. IINTER = (BUBID3M + BUBOR3M)/2) 12 8 4 0-4 -8-12 -12 1999 2000 2001 2002 2003 2004-16 1999 2000 2001 2002 2003 2004 Recursive log NBR steriliz. flow estimates ±2S.E. Recursive log MoF debt issuance <=1Y estim. ± 2 S.E.
120 Variance Decomposition of IINTER 100 80 60 40 20 0 1 2 3 4 5 6 7 8 9 10 11 12 II N T E R LFLUX LVOLMOF
120 100 percent of GDP 3.2 International Comparisons Credit to Non-Government* 94.4 112.0 113.8 80 60 40 20 61.5 23.0 31.7 49.3 37.6 16.4 58.2 76.0 71.7 8.3 23.9 17.5 0 CZ (1997) PL (1998) HU (2001) CL (1990) BR (1999) Source: EUROSTAT, NBR, NIS, IMF MX (1999) IS (1992) TH (2000) NZ (1990) UK (1992) SE (1993) ZA (2000) AL (2004)** UA (2004)** *) at the time of shifting to inflation targeting **) intends to adopt inflation targeting ***) to shift to inflation targeting in 2005 RO (2004)***
Bank Assets* 250 percent of GDP 229.0 200 150 136.0 138.0 127.2 100 104.4 92.5 50 56.0 69.1 53.9 49.8 39.4 38.3 0 CZ (1997) PL (1998) HU (2001) BR (1999) IS (1992) TH (2000) Source: Websites of central banks and national statistics institutions NZ (1990) UK (1992) ZA (2000) AL (2004)** UA (2004)** *) at the time of shifting to inflation targeting **) intends to adopt inflation targeting ***) to shift to inflation targeting in 2005 RO (2004)***
Stock Market Capitalisation* 180 160 140 120 percent of GDP 160.0 100 80 60 45.0 42.9 45.0 32.1 40 24.1 12.9 20.1 20 0 CZ (1997) PL (1998) HU (2001) CL (1990) BR (1999) MX (1999) IS (1992) Source: EUROSTAT, NIS, BSE, RASDAQ 24.4 TH (2000) 20.0 NZ (1990) 86.5 UK (1992) 52.6 SE (1993) ZA (2000) 0.9 UA (2004)** *) at the time of shifting to inflation targeting **) intends to adopt inflation targeting ***) to shift to inflation targeting in 2005 17.6 RO (2004)***
Insurance Premia* 18 16 14 12 10 percent of GDP 16.9 8 6 4 2 6.1 2.9 2.9 2.8 3.5**** 2.0 2.5 1.7 1.3 0 CZ (1997) PL (1998) HU (2001) BR (1999) Source: EUROSTAT, Swiss Re, ISC, NIS MX (1999) TH (2000) SE (1993) ZA (2000) *) at the time of shifting to inflation targeting **) intends to adopt inflation targeting ***) to shift to inflation targeting in 2005 UA (2004)** RO (2004)*** ****) 2003
Fiscal Surplus/Deficit* 2 0 percent of GDP 0.7 2 4 6 8-1.3-2.2-4.4-1.1-2.2-4.4-6.4-1.8-5.4-3.0-1.1 10 12-9.9-9.2-11.4 14 CZ (1997) PL (1998) HU (2001) Source: EIU, MPF, NIS CL (1990) BR (1999) MX (1999) IS (1992) TH (2000) NZ (1990) UK (1992) SE (1993) ZA (2000) AL (2004)** UA (2004)** *) at the time of shifting to inflation targeting **) intends to adopt inflation targeting ***) to shift to inflation targeting in 2005 RO (2004)***
4. Steps Ahead Additional scope in data reporting, surveys: integrating reliable information & expectations regarding housing & financial assets Improvements in near-term forecasting and integration into decision-making: ARMAX, expert judgment Improvements in MAMTF: (i) refining calibration; (ii) adding extra wrinkles ;; (iii) interaction with additional blocks (satellite sub-models): external accounts, fiscal; (iv) extending range to 6 and 8 quarters ahead Transparency & communication: improvements in content of QIR; briefing sessions on regular basis; possible publication of minutes/voting records