MODELLING IN THE NATIONAL BANK OF TAJIKISTAN · 2015. 12. 17. · REPUBLIC OF TAJIKISTAN...
Transcript of MODELLING IN THE NATIONAL BANK OF TAJIKISTAN · 2015. 12. 17. · REPUBLIC OF TAJIKISTAN...
MODELLING IN THENATIONAL BANK OF TAJIKISTAN
Dec 09, 2015 Bangkok
Khurshed Ismatulloev
10.12.2015 1
REPUBLIC OF TAJIKISTAN
•Population: 8,3 million (UN, 2014) •Capital: Dushanbe•Area: 143,100 sq km (55,251 sq.
miles) •Languages: Tajik-State language,
Russian-language of the interethniccommunication
•Major religion: Islam•Life expectancy: 66 years (men), 75
years (women) (UN) •Monetary unit: 1 Tajik somoni = 100
dirams•Main exports: Aluminium, electricity,
cotton, fruit, textiles, gold•GDP per capita: US $1036.0 Internetdomain: .tj•International dialling code: +992
10.12.2015 2
Phase I
2013•Econometric training
Phase II
2013
•Broadening understanding of Monetary Economics
Phase III
2014
•Medium‐term forecasting using Mantis Scenario Builder
Development of modelling and forecasting in the National bank of Tajikistan
10.12.2015 3
All for short term inflation forecast:
Development of mod…
1. •ARIMA model
2. •VAR model
3. •VEC model10.12.2015 4
Mantis as a first step to get acquainted with structural models
• Mantis prepares “baseline” forecast on a quarterly basis and offers an online Scenario Builder tool to create alternative scenarios
• Pros– Baseline forecast:
• comprehensive review of economic developments over the quarter• most likely scenario forecast based on the developments• overview of risks to the baseline
– Scenario Builder:• possibility of adjusting the baseline scenario online• save, report, and compare scenarios online• download databases with different scenarios
• Cons– Scenario Builder:
• limited number of pre‐defined scenarios and shocks• limitations on NTF assumptions (i.e. NTFs can only be changed within a certain
range)• no possibility of changing the model structure or model calibration• limited flexibility in changing a specific single shock assumptions
10.12.2015 5
Current subproject
Phase I
2015
• Developing internal FPAS capacity (agreed)
Phase II
2016
•Strengthening internal and developing external FPAS capacity (possible)
Phase III
2016
• Strengthening external FPAS capacity (possible)
10.12.2015 6
QPMmodelA canonical gap model ‐simple model sufficient to describe basic workings of monetary transmissions Aggregate demand –IS Curve Aggregate supply –Phillips Curve Uncovered interest rate parity (simple/extended) Policy rule –Taylor Rule
Work horse at many central banks Developed countries: Bank of Canada, Reserve Bank of New Zealand, Switzerland
Emerging economies: Czech Republic, Serbia, Ukraine, Georgia, Romania, Colombia, Peru, Guatemala, Botswana
10.12.2015 7
Why we chose QPM model?
Quantity of transition variables
Quality of model (forecast)
ARIMA
VAR
VEC
QPM
DSGE
10.12.2015 8
Our model… Distinguishes between the observed value and a trend (≈ steady‐state value of the system)
It’s very easy to understand and explain to people Forward‐looking Model consistent expectations of …
• Inflation • Exchange rate
Policy analysis Central bank is part of the model
Open structure Expert views may be easily incorporated
10.12.2015 9
Ideal transmission mechanism
Shockshitting theeconomy
SR interestrate
LR interest rates
Ex rate
Financial shocks:Foreign ratesPortfolio changes
Demand shocks:Foreign demand
Fiscal policy
Inflation shocks:Indirect taxesEnergy prices
InflationRMCI Output gap
Inflation expectations
10.12.2015 10
Building TJ model
• Distinguishing features:– Monetary policy operates through FX– Remittances
• real remittances gap in IS curve• real remittances gap in MP rule• real remittances trend growth in REER trend depreciation
• effective foreign output gap affects real remittances gap– Nominal IRs are residual in the UIP condition– External sector approximated by RU (representing all CIS region) and US (representing rest of the world)
10.12.2015 11
Model extensions
• Decomposition of aggregate demand– C, I, G, X, M
• Decomposition of aggregate supply– Food, non‐food, and services Phillips curves– Relative prices
• Money demand• Fiscal side
– structural/cyclical deficit decomposition• Using different model calibrations for history and forecast
• Bayesian estimation of some model parameters
10.12.2015 12
Deceleration of Russian economic growth;
Decreasing of remittances and aggregate demand.
Expectations:
Inflation level in 2015 ‐2016 will maintain in scope of projection (6.0 percent);
Exchange rate pressures’ risk and its influence to inflation will remain.
Forecast
10.12.2015 13
Strengths
‐ Reform‐oriented, young, and ambitious management
‐ Recognition of the need for changes‐ Now we have good model
‐…
Weaknesses‐ Capacity constraints (lack of qualified staffs)‐ Backward‐looking monetary policy‐ Poorly working transmission mechanism
‐Weak communication (internal and external)‐ Low credibility, low transparency‐ Underdeveloped financial sector‐…
Opportunities‐ Increase of NBT’s independence‐ TAs in different areas‐NBT is not the first one to incorporate FPAS –build on the shoulders of others
‐ Starting from a low base (every step in right direction will have a big positive impact)‐…
Threats‐ Loosing momentum‐Getting sunk in bureaucracy ‐ Continuation of old practices
‐Weak economy (non‐diversified exports, strong dependence on remittances, low production base)‐…
External factors
NegativePositive
Internal factors
SWOT analysis
10.12.2015 14
You are welcome…
THANKS for your attention!!!10.12.2015 15