Back Matter
Author:
Nils Mæhle 0000000404811396 https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Nils Mæhle in
Current site
Google Scholar
PubMedClose
,
Tibor Hlédik 0000000404811396 https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Tibor Hlédik in
Current site
Google Scholar
PubMedClose
,
Mikhail Pranovich 0000000404811396 https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Mikhail Pranovich in
Current site
Google Scholar
PubMedClose
,
Carina Selander
Search for other papers by Carina Selander in
Current site
Google Scholar
PubMedClose
, and
Mikhail Pranovich 0000000404811396 https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Mikhail Pranovich in
Current site
Google Scholar
PubMedClose
Publication Date:
10 Dec 2021
ISBN:
9781557753274
Language:
English
Download PDF (1.3 MB)

Annex 1. The Forecast Calendar

The organization of the forecasting process during every forecasting round should be based on a forecasting calendar (FC) that is available to all involved staff and approved by the central bank management. The FC frames the forecasting process and ensures that the forecasting round is carefully planned and can easily be monitored by the head of the FT and management. In particular, the FC should include the:

  • ■ Release dates for key data. This is important for sequencing the analytic work at departments’ level.

  • ■ Dates for national holidays, potentially limiting the available time for producing intermediate forecast inputs.

  • ■ Timing of all meetings at the level of the department as well as departmental meetings with the policymakers.

  • ■ Deadlines (specified in terms of day and hour) for updates of the centralized database and submissions of all key intermediate forecast inputs. Ideally, the FC also includes personal responsibilities for the delivery with deadlines.

An example of such a FC is provided in Annex Table 1.1. The example assumes that:

  • ■ The quarterly national accounts data for the first quarter of the calendar year are released in the afternoon on Friday, June 28. The work—analyzing the data and preparing the forecasts—starts immediately after that (weeks 1–6 of the FC).

  • ■ The CPI data (June in this case) are released in the morning 12 working days after the end of the month. In the example, that is just after the meeting with the MPC on the initial conditions but almost a week before the meeting with the MPC on the first version of the forecasts, which provides staff sufficient time for in-depth analysis of the data. With the final MPC meeting and announcement of the policy decision and release of the forecasts taking place five working days after the end of the month (August in this case), the release of the next CPI data would take place around seven working days after that again, which also provides sufficient distance between them.

  • ■ No major new data are released between the time the final version of the forecasts and policy recommendation are finalized and the time of the MPC monetary policy decision that could require staff having to make last minute changes to the forecasts.77

  • ■ The central bank has one large Monetary Policy Department comprising of (1) an Economic Policy Analysis Division with a Monetary Policy Analysis Unit and a Fiscal Analysis Unit; (2) a Macroeconomic Forecasting Division with a Near-term Forecasting Unit, a Medium-term Forecasting/Core Model Unit, and a Model Development Unit; and (3) an International Economic Analysis Division responsible for preparing the forecasts and monetary policy recommendations. This department, together with the communication department, prepares all documents for the meetings with the MPC as well as the externally published monetary policy documents (press releases, MPR, and presentations and speeches by the governor and other members of the senior management).

Annex Table 1.1.

Prototype Forecast Calendar: 2019 24 June – 9 August

article image
article image
Source: Authors’ calculations. CPI = consumer price index; MPC = monetary policy committee; MPR = monetary policy report; NTF = near-term forecasting.

Week 0

The period running up to the start of the forecasting/policy analysis round (referred to as week 0 in Annex Table 1.1) should be devoted to preparatory work and introductory meetings. Staff should continuously be involved in analyzing/transforming new data as they are released. They may also be working on improving the forecasting techniques (both for nowcast/NTF and the QPM-based medium-term forecasts). The meeting to discuss and decide on changes to the forecasting techniques should take place ahead of the release of the new quarterly national accounts (QNA) data because the background materials for this meeting are not dependent on new data releases and thus can be produced well in advance of the start of the process. The announcement of the detailed time schedule for the upcoming forecasting cycle and any other internal briefings on the exercise should take place before the release of the QNA data so that staff can focus fully on working on those data once they are released.78

Week 1

Week 1 is dedicated to analyzing the new QNA data, identifying any issues that might affect the forecast preparation, and preparing the first NTF estimates. The purpose of the departmental-level issues meeting is to identify and discuss any issues that might affect the forecast round, including important domestic and international developments. Special emphasis should be put on those events that potentially affect the NTFs and the medium-term forecasts/policy analysis and that are not explicitly captured by the NTF models and/or the QPM. This meeting is set a few days after the release of national accounts for staff to have made some preliminary analysis of the data and, if time allows, a first version of the NTFs. The departmental-level meeting on the NTFs, BOP, and fiscal developments and the related proposed trajectory of exogenous model variables (foreign output gap, foreign inflation, foreign interest rate, commodity prices, fiscal, etc.) is scheduled for the last day of the week.

The timing of this departmental-level meeting provides the sectoral experts with approximately four or five days to analyze the national accounts data, produce the first estimates, and prepare their presentations.79 During this period, the real sector, external sector, fiscal sector, and financial sector teams are expected to process and analyze the newly available data and reach an internal consensus on the NTFs and proposed trajectory for the exogenous variables. This might involve several internal team meetings, during which the members of the team are presenting, with the participation of the department’s senior management, their intermediate results. The discussions and the resulting feedback from team members or department management frequently require follow-up meetings. The forecast should be finalized via iterations, aiming at polishing the argumentation and the overall story. Although this is often time-consuming, only these iterative forecast rounds, distilling and synthetizing the knowledge of team members, can yield results that reflect the collective know-how.

Week 2

Week 2 is reserved for determining the initial conditions for the medium-term forecasts and preparing the background materials for the first preparatory meeting with the MPC on the initial conditions. The NTFs as well as the sectoral experts’ assessment and detailed sectoral knowledge of the economy are important inputs into developing a consensual view on the current cyclical state of the economy. Setting the initial conditions in the core medium-term forecasting and policy analysis model thus involves all members of the full FT. It requires in particular a close interaction between the sectoral experts (both those preparing the NTFs and other sectoral experts) and the core modeling group. Group level as well as full FT and departmental-level discussions are needed to develop a consensual view on how to interpret the new data, agree on how to understand the factors behind the earlier rounds forecast errors and finding economic interpretation for them, cross-check the assessment of the data with information from other sources, and identify uncertainties. The focus should be on the overall story that the data tells and whether the new information and new NTFs are sufficiently robust. Only cross-checked, thoroughly analyzed information should warrant fundamental changes to the assessment of the cyclical position and the medium-term forecasts compared to the earlier round or not. These discussions are crucial for focusing the material for the MPC meeting on analytically backed information that are essential for determining the cyclical position of the economy and on the likely key drivers of the medium-term forecasts and associated policy recommendations, without overwhelming them with details.

It is important to provide adequate time for these tasks. The assessment of the cyclical state of the economy is one of the main factors affecting the medium-term forecasts. It is also one of the main factors determining whether a policy adjustment could be needed. Approximately one week is typically needed for the technical-level discussions and the preparation of the material for the preparatory meeting with the MPC. The meeting with the MPC is scheduled for the Monday of week 3 to provide, among others, sufficient time for preparing the material and to allow the MPC members some time ahead of the meeting to review it.

Week 3

Week 3 is dedicated to preparing the first version of the medium-term forecasts and associated monetary policy recommendations. This work can start immediately after the meeting with the MPC on the initial conditions. At that point there should be a reasonable common understanding of how to interpret the new data and what the current cyclical position of the economy is. The staff should then have received feedback from the policymakers whether they share the staff’s view on the fundamental interpretation of the data or whether parts of the analysis should be complemented or reassessed. It is important that policymakers and staff have the same information and understanding of the initial conditions at this stage to reduce the risk of fundamental disagreements later in the process.

The departmental-level meeting on the first version of the forecasts and policy recommendations is scheduled for the middle of week 3. Because this meeting takes place only two days after the initial conditions meeting with the MPC, the FT has a relatively short time to produce the first version of the forecast. Therefore, members of the core modeling unit should start working on the first version of the forecast before the MPC initial conditions meeting. Besides helping the team to identify any potential bottlenecks and finding technical solutions to address them, it also helps with the preparation for the initial conditions meeting.

The FT may organize several intermediate meetings before the formal departmental meeting on the first version of the forecasts. Several rounds of iterations are typically needed to produce the first version of the forecasts, to be discussed at the departmental level. It is not uncommon to have several team-level forecasting meetings during a day to discuss the results from one iteration, how to best incorporate the feedback from the MPC meeting, and whether the results suggest that the estimates of the initial conditions may need to be revised. These team-level meetings may be organized on a short notice and therefore are often not included in the formal FC.

The draft description of the first version of the forecasts and policy recommendations will typically have to be revised following the departmental-level meeting. The calendar provides staff with only two days for doing this and preparing the material for the second meeting with the MPC. Again, the team may in practice have prepared a first cut of the material for the MPC meeting ahead of the departmental-level meeting and based its presentation to the department on that draft version. The meeting with the MPC is again scheduled for the Monday of subsequent week to allow the MPC members time to look at the material ahead of the meeting. The meeting material and presentation should focus on the business cycle characteristics of the economy, the main factors affecting the monetary policy decision, risks to the baseline forecast, and any new information explaining the change of the forecast (compared with the previous quarter).

Week 4

Week 4 is devoted to the meeting with the MPC on the first version of the forecast and policy recommendation, and to preparing final version reflecting the outcome of that meeting. This meeting is the most important preparatory meeting with the MPC during the process. Obtaining feedback from the policymakers and incorporating it into the final version of the forecast and policy recommendation is essential for ensuring that the process properly supports policymaking and for achieving the needed consistency between the forecasts and the policy decision. Besides feedback on the baseline scenario, the meeting should also lead to agreement on alternative scenarios that the staff should produce. These can either be based on policymakers’ request or proposed by the staff. It is important that the timing of the meeting provide the staff sufficient time for incorporating the feedback from the policymakers into the final baseline forecast and the alternative scenarios.

The departmental meeting on the final version of the forecast is scheduled to take place two days after the meeting with the MPC. At this meeting, the FT presents for discussion a draft final version of the baseline and alternative scenarios that incorporate the policymakers’ feedback on the earlier version. This meeting may either approve the draft presented as the final version or provide a last round of fine-tuning observations that the team would incorporate into the final version shortly after the meeting.80 A second departmental-level meeting to agree on this revised version may, or may not, be needed. The head of the FT should then inform the rest of the department and other staff involved in preparing the material for the MPC meeting, the MPR, and the press release (to be published shortly after the MPC meeting) that the final version of the forecast has been completed and that the database has been updated and saved on the departmental forecast directory.

Week 5

The last week before the MPC policy meeting is dedicated to preparing the background material for this meeting and the documents to be published. Completing the MPR, based on the material prepared earlier on in the process, may take two or three days. It typically requires holding several drafting sessions—one for each major part of the report and one final session for the executive summary—to ensure that the different sections of the report are internally consistent as well as consistent with the overall story emerging from the team’s work. All authors of the individual sections of the report are assumed to be present at the sessions covering their sections as well as for the final session. Editors and the core drafting team responsible for shaping the overall story and the internal consistency of the reports should be present at all sessions as well. Authors of other sections may sometimes also be required to be present. The sessions should focus on the report’s main storyline and overall consistency, its drafting style and the consistency of the sections drafting style and with the central bank guidelines, and on removing long descriptions and eliminating sections with too much focus on up-and-down movements in data (“elevator economics”).

The FC should include a clear plan with deadlines for submission of the drafts of the individual sections that provides the authors sufficient time to prepare their drafts and the members of the drafting sessions sufficient time to review the text. Some parts of the report are not dependent on the final version of the forecasts and policy recommendations.81 These sections should be submitted first (referred to as first-round, second-round, etc., inputs to the report in Annex Table 1.1). The drafting sessions for these parts of the report can, therefore, also be held early in the process. Some of these inputs may have been produced even before the formal start of the forecasting process. Consequently, they may have gone through several rounds of edits before getting to the drafting session stage. The material prepared for the MPC meetings on the initial conditions and the first version of the forecasts would typically constitute early drafts, or PowerPoint presentations, of the other main parts of the report. Thus, while these parts cannot be finalized before the final version of the forecasts is completed, preparing the close to final draft for the drafting session may not take that long.

The MPR should ideally be submitted to the MPC at least 1.5 to 2 days before the MPC policy meeting. At the same time, drafts of all communication material (especially press releases) that must be entirely consistent with the MPR should be completed before the MPC’s monetary policy decision meeting so that they can be discussed, revised, and approved during that meeting. Staff may also be required to prepare for internal use by the MPC a separate document, or section to the MPR, that elaborates on the policy recommendation, risks and tradeoffs involved, alternative scenarios, and particular communication-related considerations or challenges that the MPC members should be aware of before the meeting. The example FC includes a separate high-level/department management meeting on the last day of week 5 to discuss and finalize this document before it is submitted to the MPC.

Week 6

The final week is dedicated to the MPC policy decision meeting. The release of the policy decisions should ideally take place in the middle of the week so that the markets, press, and public at large have some time before the weekend to absorb the decision, which should help reduce the risk of major market distortions. This timing also allows central banks that use longer maturity instruments for their main open market operations to schedule those operations for after the release of the MPC decision and still have them taking place mid-week, which also helps reduce the risk of major market distortions. Some central banks schedule this meeting to take place over two days to allow sufficient time for the staff presentations and discussions before a closed-door MPC member-only meeting, and to allow for last-minute changes to the MPR following those discussions before the final decisions.

The final set of documents should be submitted to the MPC at least one day before the meeting. This material typically consists of the draft press release and summary presentation of the main message of the MPR and policy decision for the press and public, the internal document for the MPC elaborating on the policy recommendation, and the staff’s prepared presentation for the MPC meeting.

On the day of the monetary policy decision, the staff incorporates all requests regarding the final wording of the communication document from MPC. Drafting suggestions from policymakers or their advisors should be collected as soon as possible and incorporated into the final version of the MPR. To allow for the whole communication package, including both the press release and the MPR to be published together shortly after the MPC meeting while allowing staff time to make the necessary edits to the material, some central banks schedule the release on the morning after the meeting. This requires tight control of the process and access to the documents to minimize the risk of any leaks.

The process ends with the postmortem meeting. This technical-level meeting is aimed at drawing lessons from systematic or ad hoc errors that emerged during the forecasting round and adopting procedures that would prevent repeating them in the future.

Annex 2. Examples of Typical Forecasting and Policy Analysis System Capacity Development Log Frames

Annex Table 2.1.

FPAS Log Frame

article image
article image
article image
article image
article image
article image
article image
article image
Source: Authors’ calculations. Note: CB = central bank; DSGE = dynamic stochastic general equilibrium; FPAS = forecasting and policy analysis system; FT = forecasting team; H=high risk; HQ = IMF Headquarters (Washington DC); ICD = Institute for Capacity Development; L=low risk; MFU = modeling and forecasting unit; MP = monetary policy; MPC = monetary policy committee; MT = modeling team; NTF = near-term forecasting; PC = projection coordinator; QPM = quarterly projection model; STX = short-term expert.

References

  • Adolfson, Malin, Stefan Laséen, Jesper Lindé, and Mattias Villani. 2007. “RAMSES – a New General Equilibrium Model for Monetary Policy Analysis.” Sveriges Riksbank Economic Review 2: 5–40.

    • Search Google Scholar
    • Export Citation

  • Alichi, Ali, Douglas Laxton, Marshall Mills, and H. Weisfeld. 2008. “Inflation Forecast Targeting in a Low-Income Country: The Case of Ghana.” Ghana: Selected Issues,” IMF Country Report No. 08/332, International Monetary Fund, Washington, DC: 2–18.

    • Search Google Scholar
    • Export Citation

  • Alichi, Ali, Huigang Chen, Kevin Clinton, Charles Freedman, Marianne Johnson, Ondra Kameník, Turgut Kısjnbay, and Douglas Laxton. 2009. “Inflation Targeting Under Imperfect Policy Credibility.” IMF Working Paper 09/94, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Amarasekara, Chandranath, Rahul Anand, Kithsiri Ehelepola, Hemantha Ekanayake, Vishuddhi Jayawickrema, Sujeetha Jegajeevan, Csaba Kober, Tharindi Nugawela, Sergey Plotnikov, Adam Remo, Poongothai Venuganan, and Rasika Yatigammana. 2018. “An Open Economy Quarterly Projection Model for Sri Lanka.” IMF Working Paper 18/149, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Andrle, Michal, Tibor Hlédik, Ondra Kameník, and Jan Vlček. 2009. “Implementing the New Structural Model of the Czech National Bank.” CNB Working Paper 2/2009, Czech National Bank, Prague.

    • Search Google Scholar
    • Export Citation

  • Andrle, Michal, Andrew Berg, R. Armando Morales, Rafael Portillo, and Jan Vlček. 2013a. “Forecasting and Monetary Policy Analysis in Low-Income Countries: Food and Non-Food Inflation in Kenya.” IMF Working Paper 13/61, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Andrle, Michal, Andrew Berg, R. Armando Morales, Rafael Portillo, and Jan Vlček. 2013b. “Money Targeting in a Modern Forecasting and Policy Analysis System; an Application to Kenya.” IMF Working Paper 13/239, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Argov, Eyal, Natan Epstein, Philippe Karam, Douglas Laxton, and David Rose. 2007. “Endogenous Monetary Policy Credibility in a Small Macro Model of Israel,” IMF Working Paper 07/207, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Baksa, Dániel, Ales Bulfr, and Dyna Heng. 2020. “A Simple Macro-Fiscal Model for Policy Analysis: An Application to Cambodia.” IMF Working Paper 20/202, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Beneš, Jaromir, Jaromir Hurník, and David Vávra. 2008. “Exchange Rate Management and Inflation Targeting: Modeling the Exchange Rate in Reduced-Form New Keynesian Models.” Czech Journal of Economics and Finance 58 (03–04): 166–94.

    • Search Google Scholar
    • Export Citation

  • Beneš, Jaromir, David Vávra, and Jan Vlček. 2002. “Medium-Term Macroeconomic Modeling and Its Role in the Czech National Bank Policy.” Czech Journal of Economics and Finance 52 (4): 197–231.

    • Search Google Scholar
    • Export Citation

  • Beneš, Jaromir, Andrew Berg, Rafael A. Portillo, and David Vávra. 2015. “Modeling Sterilized Interventions and Balance Sheet Effects of Monetary Policy in a New-Keynesian Framework.” Open Economies Review 26: 81–108.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Beneš, Jaromir, Kevin Clinton, Asish George, Pranav Gupta, Joice John, Ondra Kamenik, Douglas Laxton, Pratik Mitra, G. V. Nadhanael, Hou Wang, and Fan Zhang. 2017. “Inflation Forecast Targeting for India: An Outline of the Analytical Framework.” Working Paper 17/32, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Berg, Andrew, and Rafael Portillo. 2018. Monetary Policy in Sub-Saharan Africa. Oxford: Oxford University Press.

  • Berg, Andrew, Philippe Karam, and Douglas Laxton. 2006. “A Practical Model-Based Approach to Monetary Policy Analysis – Overview.” IMF Working Papers 06/80, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Berg, Andrew, Rafael Portillo, and D. Filiz Unsal. 2010. “On the Optimal Adherence to Money Targets in a New-Keynesian Framework: An Application to Low-Income Countries.” IMF Working Paper 10/134, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Benlamine, Mokhtar, Ales Bulir, Meryem Farouki, Ágnes Horváth, Faical Hossaini, Hasnae El Idrissi, Zineb Iraoui, Mihály Kovács, Douglas Laxton, Anass Maaroufi, Katalin Szilágyi, Mohamed Taamouti, and David Vávra. 2018. “Morocco: A Practical Approach to Monetary Policy Analysis in a Country with Capital Controls.” IMF Working Paper 18/27, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Black, Richard, Douglas Laxton, David Rose, and Robert Tetlow. 1994. “The Bank of Canada’s New Quarterly Projection Model, Part 1. The Steady-State Model: SSQPM.” Technical Report 72, Bank of Canada, Ottawa.

    • Search Google Scholar
    • Export Citation

  • Black, Richard, Vincenzo Cassino, Aaron Drew, Eric Hansen, Benjamin Hunt, David Rose, and Alasdair Scott. 1997. “The Forecasting and Policy System: The Core Model.” Research Paper 43, Reserve Bank of New Zealand, Wellington.

    • Search Google Scholar
    • Export Citation

  • Brubakk, Leif, Tore Anders Husebø, Junior Maih, Kjetil Olsen and Magne Østnor. 2006. “Finding NEMO: Documentation of the Norwegian Economy Model.” Staff Memo 2006–6, Norges Bank, Oslo.

    • Search Google Scholar
    • Export Citation

  • Chiumia, Austin, Yaroslav Hul, Wytone Jombo, Grant P. Kabango, Shalva Mkhatrishvili, Marietta Mpingasa, Arnold Palamuleni, and Kisukyabo Simwaka. Forthcoming. “An Open Economy Quarterly Projection Model for Malawi.” IMF Working Paper, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Clinton, Kevin, Jihad Dagher, Ondrej Kamenik, Douglas Laxton, Ali Alichi, and Marshall Mills. 2010. “A Model for Full-Fledged Inflation Targeting and Application to Ghana.” IMF Working Paper 10/25, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Clinton, Kevin, Tibor Hlédik, Tomás Holub, Douglas Laxton, and Hou Wang. 2017. “Czech Magic: Implementing Inflation-Forecast Targeting at the CNB.” IMF Working Paper 17/21, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Coats, Warren, Douglas Laxton, and David Rose (Eds.). 2003. The Czech National Bank’s Forecasting and Policy Analysis System. Prague: Czech National Bank.

    • Search Google Scholar
    • Export Citation

  • Czech National Bank. 2004. “Practical Experience with Inflation Targeting.” Proceedings from an International Conference held at the Czech National Bank, May 13–14.

    • Search Google Scholar
    • Export Citation

  • Dizioli, Allan, and Jochen M. Schmittmann. 2015. “A Macro-Model Approach to Monetary Policy Analysis and Forecasting for Vietnam.” IMF Working Paper 15/273, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Ericsson, Neil R. 1998. “Empirical Modeling of Money Demand.” Empirical Economics 23: 295–315.

  • Ericsson, Neil R., David F. Hendry, and Kevin M. Prestwich. 1998. “The Demand for Broad Money in the United Kingdom, 1878–1993.” Scandinavian Journal of Economics 100 (1): 289–324.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Fukač, Martin, and Adrian Pagan. 2006. “Issues in Adopting DSGE Models for Use in the Policy Process.” CNB Working Paper 6/2006, Czech National Bank, Prague.

    • Search Google Scholar
    • Export Citation

  • González, Andrés, Diego Rodríguez, Aleš Bulfr, Dániel Baksa, Juan S. Corrales, and Dyna Heng. 2021. “Using Macroeconomic Frameworks to Analyze the Impact of COVID-19: An Application to Colombia and Cambodia.” IMF Technical Notes and Manuals 2021/001, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Holub, Tomáŝ. 2011. “ The Policy Formulation Process: Czech Experience”, CNB’s presentation, “Improving Monetary Frameworks- Inflation Targeting Policy and Implementation.” High Level Workshop organized by the IMF, the Central Bank of Armenia and the EBRD, Yerevan and Tshakhkadzor, Armenia, April 27-May 6, 2011.

    • Search Google Scholar
    • Export Citation

  • Holub, Tomáŝ. 2017. “Developing Internal Capacity for Policy Analysis and Forecasting. CNB’s Experience.” Presentation at the September 2017 IMF-Joint Vienna Institute Seminar on Evolving Monetary Policy Frameworks, September. Czech National Bank.

    • Search Google Scholar
    • Export Citation

  • Hlédik, Tibor. 2011. “Forecasting and Policy Analysis System at the Czech National Bank.” Presentation at “Improving Monetary Frameworks– Inflation Targeting Policy and Implementation,” High Level Workshop organized by the IMF, the Central Bank of Armenia and the European Bank for Reconstruction and Development, Yerevan and Tshakhkadzor, Armenia, April 27–May 6.

    • Search Google Scholar
    • Export Citation

  • Hlédik, Tibor, Karel Musil, Jakub Ryšánek, and Jaromir Tonner. 2018. “A Macroeconomic Forecasting Model of the Fixed Exchange Rate Regime for the Oil-Rich Kazakh Economy.” CNB Working Papers 2018/11, Czech National Bank, Prague.

    • Search Google Scholar
    • Export Citation

  • Hunt, Benjamin, David Rose, and Alasdair Scott. 2000. “The Core Model of the Reserve Bank of New Zealand’s Forecasting and Policy System.” Economic Modelling 17 (2): 247–74.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund (IMF). 2015. “Evolving Monetary Policy Frameworks in Low-Income and Other Developing Countries.” IMF Policy Paper, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Laxton, Douglas, and Alasdair Scott. 2000. “On Developing a Structured Forecasting and Policy Analysis System Designed to Support Inflation-Forecast Targeting (IFT).” In Inflation Targeting Experiences: England, Finland, Poland, Mexico, Brazil, Chile, 6–63. Ankara: The Central Bank of the Republic of Turkey.

    • Search Google Scholar
    • Export Citation

  • Li, Bin Grace, Christopher Adam, Andrew Berg, Peter Montiel, and Stephen O’Connell. 2019. “Structural VARs and the Monetary Transmission Mechanism in Low-Income African Countries.” Journal of African Economies 28 (4): 455–78.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Mæhle, Nils. 2020. “Monetary Policy Implementation: Operational Issues for Countries with Evolving Monetary Policy Frameworks.” IMF Working Paper 20/26, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Mishkin, Frederic S., and Adam S. Posen. 1997. “Inflation Targeting: Lessons from Four Countries.” Federal Reserve Bank of New York Economic Policy Review 3: 9–110.

    • Search Google Scholar
    • Export Citation

  • Murchison, Stephen, and Andrew Rennison. 2006. “ToTEM: The Bank of Canada’s New Quarterly Projection Model.” Bank of Canada Technical Report 97, Bank of Canada, Ottawa.

    • Search Google Scholar
    • Export Citation

  • Musil, Karel, Mikhail Pranovich, and Jan Vlček. 2018. “Structural Quarterly Projection Model for Belarus.” IMF Working Paper 18/254, International Monetary Fund, Washington, DC.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Nalban, Valeriu, Abubakar Addy, James Attuquaye, Simon Harvey, Nana Kwame Akosah, Shalva Mkhatrishvili, and Zakari Mumuni. Forthcoming. “Quarterly Projection Model for the Bank of Ghana.” IMF Working Paper, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Reserve Bank of New Zealand. 1997. “The Forecasting and Policy System: An Introduction.” Reserve Bank of New Zealand Bulletin 60 (3): 225–35.

    • Search Google Scholar
    • Export Citation

  • Vlček, Jan, Mikhail Pranovich, Patrick Hitayezu, Bruno Mwenese, and Christian Nyalihama. 2020. “Quarterly Projection Model for the National Bank of Rwanda.” IMF Working Paper 20/295, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Waggoner, Daniel F., and Tao Zha. 1999. “Conditional Forecasts in Dynamic Multivariate Models.” The Review of Economics and Statistics 81 (4): 639–51.

    • Crossref
    • Search Google Scholar
    • Export Citation

  • Woodford. Michael. 2003. “Optimal Interest-Rate Smoothing.” The Review of Economic Studies 70 (4): 861–86.

2

While some larger central banks with substantial resources and extensive modeling expertise rely on a suite of models, typically only one serves as the core model for preparing the regular baseline forecasts and policy recommendations.

3

FPAS CD has been provided to Albania, Angola, Armenia, Belarus, Botswana, Chile, Ghana, Honduras, Jamaica, Kazakhstan, Kenya, Kyrgyz Republic, Malawi, Mauritius, Moldova, Morocco, Mozambique, Russia, Rwanda, Seychelles, Sri Lanka, Tanzania, Uganda, Ukraine, Vietnam, and, Zambia as well as to the East African Community.

4

Specifically, FPAS projects were often implemented in parallel with other important CD projects of the IMF, such as (1) reforming the monetary policy framework, (2) enhancing monetary operations and liquidity management, (3) improving foreign exchange management practices, (4) enhancing financial stability, (5) financial markets development, and (6) designing effective central bank communication. Simultaneous capacity building in the abovementioned areas generates positive externalities and results in more robust policy frameworks.

5

Examples of such working papers include Amarasekara and others (2018) for Sri Lanka; Andrle and others (2013a, 2013b) for Kenya; Benlamine and others (2018) for Morocco; Chiumia and others (forthcoming) for Malawi; Dizioli and Schmittman (2015) for Vietnam; Musil, Pranovich, and Vlček (2018) for Belarus; Alichi and others (2008), Clinton and others (2010), and Nalban and others (forthcoming) for Ghana; and Vlček and others (2020) for Rwanda.

6

To achieve this and bring structural issues and trends properly into the spotlight, the forecast horizon should be reasonably long. Just looking four quarters ahead, as is often done, is not enough as inflation is measured (and targeted) as an annual change—a twelfth of that outcome is already determined with a one-year horizon. Furthermore, the transmission mechanism will not have played out over that horizon. Eight or more quarters are generally needed in order to not lose sight of the medium-term policy horizon.

7

See IMF (2015) for further discussion on enhancing the formulation of monetary policy in the context of modernizing a monetary policy framework.

8

The point is more general. In “financial programming,” IMF country teams use macroeconomic forecasts—baselines and alternative scenarios—to structure policy discussions and guide the stance of a wide range of policies. Along these lines, the Institute for Capacity Development has developed a CD practice in macroframeworks, in the spirit of financial programming, to help a much broader range of institutions, including ministries of finance or economy, use macroeconomic frameworks for forecasting and policy analysis in order to inform policymaking. For early examples, see Baksa, Bulír, and Heng (2020); and González and others (2021).

9

If the forecasts do that, and the implied policy changes are implemented and communicated, the inflation forecasts should also on average be more accurate than those prepared by others. This is because they would also have to forecast the policymaker’s decisions.

10

An older approach, where the forecast was based on an unchanged monetary policy stance, was a product of strictly backward-looking, medium- to large-scale econometric models. Once central banks acknowledged the crucial role of inflation expectations, forecasts based on active monetary policy became more common both for decision-making and for external communication. Forecasts that hypothetically assume an unchanged policy stance are internally inconsistent, which could be problematic both in terms of policymaking and external communication. They are inconsistent because the policy stance should change if the forecast suggests that inflation will deviate from the central bank’s (formal or informal) target, and this fact should influence expectations and thus the inflation forecast. Forecasts based on unchanged policy are problematic for internal policymaking because they do not inform policymakers about the required policy setting for achieving the monetary policy objective. In terms of external communication, the main problem is that they do not communicate what the central bank aims at achieving nor how the economy, and in particular inflation, would likely evolve after the central bank has adjusted the policy stance. They can thus also not help anchor expectations.

11

Recognizing the importance of NTF and nowcasting for the decision-making process (including in the context of FPAS), Africa Training Institute in collaboration with African Department, the Institute for Capacity Development, Statistics Department, and Information Technology Department have been working on launching a comprehensive CD program on mainstreaming NTF and nowcasting capacity in sub-Saharan Africa. While this CD program focuses on sub-Saharan Africa, following a successful pilot, it is intended that this area of CD will become part of the Institute for Capacity Development’s broader practice on macroeconomic frameworks.

12

To avoid ad hoc or politically driven forecast adjustments, these tunes need to be well documented and presented to management for adequate accountability of the forecasting process.

13

See also IMF (2015), Principle IV, on this.

14

Ideally, central banks would have a well-structured data warehouse, but many do not. Because establishing such a data warehouse can be a huge undertaking that requires years to build, often by external consultants, FTs typically build less sophisticated databases, satisfying their immediate needs.

16

In the more data-rich countries, it often also includes other variables covering labor market, financial markets, and fiscal policy.

17

However, see footnote 11 on the ongoing preparations to launch a comprehensive CD program pilot on mainstreaming NTF and nowcasting capacity in sub-Saharan Africa. The objective of the CD program is to provide relevant organizations in sub-Saharan Africa with the cutting-edge tools and the types of data in use or developed at the IMF on nowcasting, and to familiarize the region’s officials with the concepts and methods to incorporate high-frequency economic indicators into the modeling process (including by the use of dynamic factor models, mixed frequency data sampling, and machine learning techniques). Following a successful pilot in sub-Saharan Africa, this area of CD is intended to become part of the Institute for Capacity Development’s broader practice on macroeconomic frameworks.

18

For example, times series models for inflation would unlikely include variables that capture direct taxes such as value-added tax, because variability and therefore explanatory power of such variables can be low in the historical data sample. However, if changes in value-added tax are envisaged in the near-term future with likely effect on inflation, such effect has to be incorporated by “adding” judgment to the model-based NTF.

19

The headline 12-month rate is a (asymmetric geometric) moving average of the one-month rate of change in the underlying series and thus is by construction autocorrelated with a long memory (X_t / X_(t - 12) – 1) · 100 = (X_t / X_(t - 1) · X_(t – 1) / X_(t -2). ...X_(t – 11) / X_(t - 12) – 1) · 100). It does not show the current underlying development in the series but the average development over the last year, and as a moving average is by construction a primitive and asymmetric trend filter. Relatively large errors in a forecast of the one-month rate for the next month X_(t + 1) / X_t would have a relatively minor impact on the forecast of the headline 12-month rate as it represents only one-twelfth of the moving average, with the other eleven-twelfths (that is, X_t / X_(t -1)·X_(t – 1) / X_(t – 2). ...X_(t – 10) / X_(t - 11)) and the one-month rate that is dropping off from the moving average (the base effect, Xt-11/Xt-12 being known. Base effects can cause large changes in 12-month growth rates that can easily be accounted for when instead analyzing and modeling the level or period-to-period change in seasonally adjusted time series.

20

Examples include the Czech National Bank’s “g3” introduced in 2008. See Andrle and others (2009). See also Clinton and others (2017) for the stages of model development at the Czech National Bank), the Norges Bank’s “NEMO” (Brubakk and others 2006), the Riksbank’s “RAMSES” (introduced in 2005; see Adolfson and others 2007), the Bank of Canada’s “TotEM” (introduced in 2005; see Murchison and Rennison 2006).

21

Typically, only one of these serves as the core model for preparing the regular baseline forecasts and policy recommendations with the other models serving as satellite models to help analyze and quantify the effects on the forecast of factors that are not explicitly captured by the core model.

22

However, some projects have also helped with building supplementary fully micro-founded DSGE models.

23

Examples include the difference in the medium-term fiscal impact of a widening deficit caused by increased infrastructure investment versus government consumption spending, or the fact that the real exchange rate or longer-term real interest rate response to increased borrowing and debt depends on the underlying driver of the borrowing dissaving (demand for credit versus supply of credit). The lack of an explicit and detailed micro structure, while making the model less data intensive, may possibly also make it harder to calibrate the model or make the calibration less stable when, for example, a significant share of households or producers may not have access to credit compared to benchmark countries, or there are significant and changing differences in labor/capital intensity between export- or import-competing producers and those naturally sheltered from external competition. Thus, it would require reassessing the calibration of the model regularly (but not too frequently) to ensure that it is still reflective of the perceived structure of the economy.

24

The time required for a full-fledged implementation of a well-functioning FPAS depends on many factors, including the frequency of TA missions, size of the central bank’s FT, level of technical skills of the central bank’s staff, intensity of the follow-up work in between missions at the institution receiving the TA, as well as the support of the medium- and top-level management.

25

Judgment/tunes can be applied outside the core model infrastructure, but incorporating it into the model infrastructure makes it easier to apply and ensure consistency (without having to rerun the model conditioning on all variables).

26

The literature struggles with matching the behavior of full-fledged micro-founded DSGE models to observed basic business cycle correlations in the data. Specifically, demand shocks, which often are an important driver of cyclical developments, are difficult to model as the DSGE framework stresses optimization of agents. In addition, permanent shifts observed in macro data are often difficult to be explained by micro-foundations without setting parameters of the production functions to values not supported by micro-evidence or introducing artificial model technologies. Often, the calibration of the deep structural parameters of DSGE models is also ad hoc, similarly as in the case of semistructural models. Consequently, the uncertainties are the same for both classes of models. See, among others, Fukač and Pagan (2006) on issues in adopting DSGE models for use in the policy process.

27

All variables are in 100*natural logs, except for interest rates and rates of change, that are in annualized percentage points.

28

The output gap (y^t=y¯t−yt) is defined as a deviation of the log of real output (that is, GDP volume (yt)) from its potential (y¯t), with potential defined as the level of output that can be produced without generating pressures for inflation to increase or decrease—the nonaccelerating inflation product.

29

zt=st+pt*−pt is the nominal exchange rate expressed as local currency per unit of where st foreign currency, and pt* and pt are, respectively, the foreign and local price level (all in logs).

30

One country model also included the deviation of the country risk premium from its long-term trend mcit=aa1(r^t+prem^t)+(1−aa1)(−z^t+1) to capture a steepening of the yield curve and increase in the credit risk premium when country risk was elevated. This was a special version of a more general model with the interest rate component of monetary condition in equation (2) including a credit risk premium that could be a function of additional factors (mcit=aa1(r^t+prem^t)+(1−aa1)(−z^t+1)).

31

Inflation is measured as the annualized quarter-to-quarter rate of change in the seasonally adjusted consumer price index: πt = (pt – pt-1)’ 4 = (pt/pt-1 – 1) · 400 unless otherwise stated.

32

Many versions also added the direct impact of imported inflation (foreign inflation (πt*) plus exchange rate depreciation, and adjusted for the change in real exchange rate trend in order to align its steady state value with the domestic inflation target, πtim=πt*+Δst−Δz¯t) in equation (3) as follows: πt=b1πt−1+(1−b1−b2)Etπt+1+b2πtim+b3rmct+εtπ.

33

Some versions also added the direct impact of changes in oil import prices (πtimoil) directly in equation (3) and equation (4) as follows: πt=b1Etπt−1+(1−b1−b2−b3)Etπt+1+b2πtim+b3πtimoil+b4rmct−1+εtπ, rmct=(bb1+bb2)z^t+bb2rp¯toil+(1−bb1−bb2)y^t, where the oil import price inflation component is specified as the lagged change in international oil prices (πt−1oil*) plus exchange rate depreciation and adjusted for the change in real exchange rate trend and the lagged real international oil price trend (πtimoil=πt−1oil*+Δst−Δz¯t−Δrp¯toil*).rp^toil is the real oil price gap and with the real oil price equal to (the log of) international oil prices in US dollars relative to the US consumer price index (rptoil*=ptoil*−ptUSCPI).

34

The equilibrium nominal interest rate (itn) is the sum of the equilibrium real interest rate (r¯t) and some measure of expected annual inflation four quarters ahead Etπt+44=Et(πt+1+πt+2+πt+3+πt+4)/4.

35

The inflation target (πtT) is typically specified as a time-varying variable to facilitate policy experiments and the setting of a target path for disinflating countries. At the same time, in order to anchor expectations, the best practice is choosing and communicating a constant medium-term inflation target.

36

This version features smoothing of the policy rate to reflect the fact that in practice central banks do not typically change the policy rate in large increments. This is to keep the policy signal clear (Woodford 2003) and safeguard against the risk of having to frequently reverse course, which could harm central bank efficiency and credibility and cause policy rate changes to be interpreted by the public as a random noise and thus disregarded.

37

Fiscal measures that are budget neutral (for example, changes in tax structure) typically require more elaborated analysis that is beyond simplified structure of the model and needs to be applied judgmentally on the projection.

38

The fiscal deficit is measured as a share of GDP (both variables expressed in current prices). It may be defined as before grants, especially in aid-dependent countries.

39

The cyclical component of the fiscal deficit is excluded because it is directly derived from the position of the output gap and thus effectively already captured by the IS curve. Including it in the fiscal impulse measure would, therefore, double count the impact of the cyclical component of fiscal deficit.

40

For example, a negative supply shock that causes domestic food prices to rise may lead to an estimate of output falling below potential (a negative output gap), which in the textbook version of the model would cause inflation to fall, not rise.

41

The main reason being that part of the surplus of domestic food products are either exported or because of a coinciding bumper harvest in a major export crop.

42

Andrle and others (2013b) also specified the relative price of domestic food prices relative to international food prices and international food prices relative to the international price level. The country models developed during the CD projects have not done so.

43

The TOT gap (tot^t) may or may not be included in such cases—it was included in one model as a proxy for the export crop part of total agricultural production. Where the crop is mainly exported, a bumper harvest increases income and thus demand for nonagriculture as in equation (1d), but it does not represent an increase in the supply of domestic food products and thus should not be included in the food inflation Phillips curve, as in equation (3c). For the purpose of equation (3c), a measure of the domestic agriculture food production would have been more appropriate than total agriculture. The term (y^tagr−α3tot^t) in equation (3c) is intended to serve as a proxy for the part of the agriculture output that is supplied to the domestic market.

44

The money aggregate used also has differed. Some country models have been based on base/reserve money, some on broad money (M2, M3), and some have included both base and broad money with a modeling of the evolution of the money multiplier as well.

46

It is common to refer to this equation as a money demand equation, but as such, it is wrongly specified (except as a cash demand equation). The liquidity preference theory implies that the demand for money is not a function of the level of short-term interest rates but a function of the difference between the return on holding money (“own interest rate on money”) and the interest rate on the alternative (but less liquid) assets (that is, effectively the slope of the yield curve). A dominant part of broad money is deposits, which now is interest earning. Ericsson and colleagues (Ericsson 1998; Ericsson, Hendry, and Prestwich 1998) found that assuming that the return on holding money was zero, and thus modeling the demand for money as a function of the level of interest rates, was the main reason for estimated money demand equations becoming unstable following the financial liberalization in the 1990s. Once this misspecification of the opportunity cost of holding money was corrected, the equations became well behaved. In line with endogenous money theory, it is more correct to think about the money-interest rate relationship in equations (23) and (24) as a reduced form (commercial bank) money supply relationship where banks’ lending, and thus their deposit creation, depends on output growth and the level of interest rates.

47

Such projections can, among others, be used to cross-check the assumptions and forecasts underpinning the financial programming exercise, including the inflation forecasts or targets used. Deviations between the latter and the model-derived inflation forecast would suggest a need to revisit the assumptions used in both exercises. Similarly, deviations between the model forecast for short-term interest rates and the observed rates would suggest a need for reassessing the assumptions and possibly revising the money targets. In addition, this will assist with interpreting how contractionary/expansionary the set money target is and thus with how to communicate monetary policy.

48

It may in other countries be endogenous or based on exogenous central bank forecasts for GDP. While the financial programming practice is to set the money target based on projected GDP growth, it might be better to set it based on potential GDP growth, similarly to the practice in Germany in the 1990s (Mishkin and Posen 1997).

49

Although this is in the context of money targeting, the money growth targets may, and should, be based on an explicit inflation objective. Money growth targets based on projected inflation, which again are a function of monetary policy (for example, past money growth), would render the model (and the economy) unanchored.

50

This approximates how the targets often are set under the financial programming approach, although they typically are reset only every six months. Some also set the money growth target based on a forecast instead of a target for inflation. See Andrle and others (2013b) for a similar version. Note that this approach differs from both the Friedman rule for money growth targets and how they were set by for example Germany in the 1990s (the money growth target was set equal to the growth in potential GDP plus the target for inflation). Note that it may also render inflation indeterminate.

51

See Berg and Portillo (2018, Chapter 1) on the nature of money targeting regimes in practice and implications for the FPAS.

52

As proposed by Beneš, Vávra, and Vlček (2002). See also Beneš, Hurník, and Vávra (2008) for alternative specifications of the naïve forecast that, instead of being based on long-term, or targeted, domestic and foreign inflation and long-term real exchange rate, is based on recently observed values of some or all of them as in: stNF=st−1+2[Δz¯t+(πt−πt*)]/4 or StNF=st−1+2[Δzt+(πt−πt*)]/4. Inflation targets may also be made time variant to capture cases where the central publish one-year-ahead inflation forecasts that also serve as their targets instead of having a medium-term inflation target that is kept fixed over the policy horizon.

53

The term in brackets in equation (27) is multiplied by two because the difference between the lagged and the future exchange rate (from t - 1 to t + 1) is two quarters. It is then divided by four because all elements are expressed in annualized terms.

54

See Hlédik and others (2018) for further details on this model version.

55

Or financial account frictions, according to the terminology used in the BOP5 and 6 manuals.

56

That is, the degree that sterilized interventions can influence the exchange rate beyond the very short term (that is, a few days or weeks).

57

See Mæhle (2020) for the alternative frameworks for the daily liquidity management operations that a central bank can use to steer short-term interest rates.

58

This is even true in a situation with full capital mobility.

59

Erroneous assumptions about the effectiveness of sterilized interventions would also result in policy inconsistencies and missed policy objective(s)/target(s).

60

This implies a reverse causality misunderstanding.

61

This irrespectively of whether that is to achieve a numerical inflation target, a certain money growth rate, the exchange rate (depreciation rate), or a combination as implied by equations (6c) and (6d).

62

And with market operations staff instructed to undertake the liquidity management operations needed to keep short-term market rates to that path.

63

Because the QPM is a quarterly model, these three short-term interest rates (the policy rate, the overnight interbank rate, and the 91-day treasury rate) should in principle be broadly the same.

64

And having a properly designed operating, or liquidity management, framework.

65

Central bank market operations that are not aimed at aligning market rates with the policy rate can result in opposite changes in short-term market rates from the announced change in the policy rate.

66

Such measures can be inflation deviations from target or some measure of the alignment of market rates with the policy rate.

67

For a more general, but also more complex and nonlinear, version of the model with endogenous credibility, see Argov and others (2007), Alichi and others (2009), and Beneš and others (2017).

68

See Li and others (2019) for an analysis of the limitations of structural econometric methods to detect the monetary transmission mechanism in low-income African countries, for example.

69

Econometric analyses of various sorts can inform the calibration, and Bayesian estimation of the QPM model can be a useful complement to calibration. See, for example, Amarasekara and others (2018).

70

Note that several parameter combinations can imply the same dynamics.

71

Fewer than five weeks would not allow sufficient time for the analytical work nor for the internal discussions among the different parts of the central bank and between staff and the policymakers for reaching (a reasonable degree of) consensus on the forecast and policy recommendations. Greater than seven weeks would interfere with the preparations for the recommended interim policy meetings between the full forecasting rounds.

72

To solidify political support and commitment to the FPAS CD projects, the TA teams also found it useful to hold dedicated seminars for the central bank management and for the monetary policy decision-making bodies. Such higher-level events help to highlight the benefits of FPAS and the organizational and processes transformations that may need to be undertaken in central banks for the FPAS projects and broader framework reforms to succeed.

73

Switching teams between forecasting and model development agenda usually makes model development and research plans better aligned with practical problems that emerge during the forecasting rounds and that time constraints did not allow for resolving.

74

To ensure that the staff of the modeling team was able to cope with the technicalities of forecasting and modelling, some central banks recruited people with good quantitative skills but without a formal education in economics and trained them to become applied mathematical economists.

75

IT-Lite countries float their exchange rate and announce an inflation target, but are not able to maintain the inflation target as the foremost policy objective.

76

The IMF has established standardized logframes for FPAS CD projects (see Annex 2 for examples of how they have been adapted to some country projects with different starting positions).

77

This may not always be possible but is something toward which the staff preparing the calendar should strive.

78

The FC should be prepared, approved, and known to staff well ahead of the start of the process, and preferably for the next 12 months. A reasonably detailed version of it would ideally be prepared as part of the process of determining the timing of the MPC monetary policy decisions meetings—the timing of the MPC meetings for the next 12 months should also ideally be available to the public. The internal calendar may also usefully include key activities involving the forecasting staff planned for the period in between the interim MPC meetings.

79

If it had not been for the national holiday, the NTF meeting could have been scheduled for the Monday or Tuesday of week 2. In that case, the sectoral experts should preferably provide the meeting participants, and importantly the core modeling group, with their NTF estimates before the end of the working day on Friday week 1.

80

There is no room for fundamental, far-reaching discussions in this meeting. Those discussions must be completed before this meeting.

81

Such as sections or boxes on specific economic/research topics, sections focusing on historic development, and to a large extent the discussion of initial conditions.

  • Collapse
  • Expand
Cover Departmental Papers
Taking Stock of IMF Capacity Development on Monetary Policy Forecasting and Policy Analysis Systems
Author:
Nils Mæhle
,
Tibor Hlédik
,
Mikhail Pranovich
,
Carina Selander
, and
Mikhail Pranovich
Volume/Issue:
Volume 2021: Issue 026
Publisher:
International Monetary Fund
ISBN:
9781557753274
ISSN:
2616-5333
Pages:
67
DOI:
https://doi.org/10.5089/9781557753274.087