The%20BNB%20Quarterly%20Projection%20Model%20%20Emilia%20Penkova%20and%20Svilen%20Pachedzhiev

1
The BNB Quarterly Projection
ModelEmilia Penkova and Svilen Pachedzhiev
2
The BNB Quarterly Projection Model

• Twinning Project Adjustment of the Bulgarian
  National Bank to operate as a full-fledged member
  of the European System of Central Banks and the
  Euro-system
• Component 2 Research and preparation for
  monetary policy operations in line with ECB best
  practices

3
Introduction

• The first version of the Bulgarian National Bank
  Quarterly Projection Model (BNBQM) which belongs
  to the group of traditional structural
  macroeconomic models.
• The model is similar to the European System of
  Central Banks multi-country model country blocks.
  The guiding principle in designing the country
  blocks is that of close compatibility with the
  ECB Area-Wide Model.
• The development of a model over the period
  1998-2007 poses formidable challenges, given the
  short and volatile time series we calibrate
  coefficients.
• The model is continuously tested and simulated in
  order to improve it. It should be viewed as work
  in progress and an area for future empirical
  research within the BNB, rather than as a
  finished product.

4

Introduction

• The theoretical background Neo-Classical
  Synthesis. The long-run equilibrium is determined
  by supply side factors (Neo-Classical theory) and
  short-run fluctuations are demand driven
  (Keynesian theory).
• Backward looking -expectations are reflected via
  lagged variables, which is considered adequate
  for the purpose of generating short- to
  medium-term forecasts.
• Behavioral equations error correction form
  (Engle Granger two step procedure has been
  employed).

5
The purpose of the BNBQM is twofold

• First, to produce macroeconomic forecasts for the
  Bulgarian economy.
• Second, to assess the effects of economic
  shocks on the Bulgarian economy in simulated
  scenario analyses.

6
Outline of Presentation

• Theoretical background
• Structure of the model and estimated equations
• Simulations
• Concluding remarks and extensions of research

7
Supply side of the Economy

• The standard theory of monopolistic competition
  is applied.
• Profits of an individual firm are determined by
  returns from sale with costs of labour and
  capital subtracted.
• The production process is represented by a Cobb
  Douglas function.

8
Supply side of the Economy

9
Supply side of the Economy

• where ?(Yi) are profits of the firm, Li is the
  labour force used by the firm, Ki is the capital
  stock of the firm, as are the income shares ?
  is the exogenous growth rate of technological
  progress, s is the elasticity of the demand for
  goods produced by the firm i to their relative
  price w is the nominal wage level, c is the
  nominal cost of capital with
• where r is the real rate of interest, d is
  the physical depreciation rate of capital, PI is
  price of investment goods, Pi is the price of
  goods produced by the firm, P is the price of
  generic goods, Yi is the output of the firm i, Y
  is the aggregate supply of generic goods.

10
Supply side of the Economy
11
Supply side of the Economy

• First order conditions are

12
Supply side of the Economy

• Using the assumption of symmetric equilibrium
  (PiP, YiY, LiL, KiK), we receive

13
Supply side of the Economy

• The aggregate output and the long-run demand for
  capital and labour are given by

14
Structure of the model and estimated equations

• The simulation and projection features of the BNB
  Quarterly Projection Model are driven by twenty
  behavioural equations and additional forty three
  identities. Around one hundred and sixty
  variables enter the model.
• The model is in Eviews 5.1 (a program file which
  imports the data, estimates equations, solves the
  model and produces forecasts output).
• The model is structured into five blocks
  production function and factor demand equations,
  aggregate demand, prices and wages, monetary, and
  fiscal sector.

15
Model linkages
16
Error correction form

• Most of the dynamic equations take the following
  general form
• where log(yt-1)-log(yt-1) is the error
  correction term
• ?(.) and s(.) are polynomials
• l - the lag operator

17
Potential output

• R_YP_R Potential output, prices of 2005
• R_TFP_TD Total factor productivity, trend
  (2005100)
• L_EMPL_TD Employment, trend
• R_K_R Capital stock, prices of 2005

18
Potential output

• Total factor productivity is estimated as a
  residual from the production function for the
  estimated period then using Hodrick-Prescott
  filter we receive the trend.
• The potential employment is received from a
  labour force forecast and estimated NAIRU.
• NAIRU is assumed to be at around 7.7 level
  (slightly decreasing over the forecasting period)
  and is estimated using Elmeskov (1993) approach.

19
Employment

• Log(L_EMPL_STAR) 3.278 0.600log(R_Y_R) -
• (5.607)
  (-)
• - 0.400 log(L_W/I_HICP_P)
• (-)
• Dlog(L_EMPL) -0.089 - 0.296(log(L_EMPL (-1))
• (18.365) (5.138)
• - log(L_EMPL_STAR(-1))) 0.001 Dlog(L_EMPL(-1))
• (0.002)

L_EMPL Employment (employees self
employed) R_Y_R GDP, prices of 2005 L_W
Nominal wage bill per worker I_HICP_P
Harmonised index of consumer prices, 2005 100
20
Employment
21
Gross fixed capital formation

• Log(R_KF_R_STAR) 1.942 0.700log(R_Y_R)-
• (4.338)
  (-)
• 0.108M_LTIR_N - 0.007I_I_P
• (9.451)
  (2.516)
• Dlog(R_KF_R) 0.242 -0.003(log(R_KF_R(-1))-
• (6.285)
  (-)
• - log(R_KF_R_STAR(-1))) - 0.363Dlog(R_KF_R(-1))
• 
  (2.135)
  
• R_KF_R Gross fixed capital formation, 2005
  prices
• R_Y_R GDP, 2005 prices
• I_I_P Inflation, in percentages
• M_LTIR_N Nominal long-term interest rate, in
  percentages

22
Gross fixed capital formation
23
Private consumption

• Log(R_C_R_STAR) 2.365 0.800log(R_DI_N/I_HICP_
  P)
• 
  (21.979) (-)
• 0.300log(R_K_R)
• (-)
• Dlog(R_C_R) 0.077 - 0.483(log(R_C_R(-1))
• (3.743)
  (2.200)
• - log(R_C_R_STAR(-1))) -0.343Dlog(R_C_R(-1))
• 
  (2.074)
• R_C_R Private consumption, prices of 2005
• R_DI_N Disposable income (wages and salaries
  pensions and social benefits imputed rent
  compensation of employees(BOP) current
  transfers (BOP))
• I_HICP_P Harmonised index of consumer prices,
  2005 100
• R_K_R Capital stock, prices of 2005

24
Private consumption
25
Exports

• Log(E_EX_R_STAR) 6.108 log(A_WTV_R)-
• 
  
  
  
  (2.024) (-)
• - 0.451log(E_EX_P/(A_MEPAE_PE_ER_PI))
• (4.758)
• Dlog(E_EX_R) -0.265 -0.206(log(E_EX_R(-1))-
• (8.268)
  (1.448)
• -log(E_EX_R_STAR(-1))) 0.020Dlog(E_EX_R(-1))
• 
  
  
  (0.151)
• E_EX_R Exports, prices of 2005
• A_WTV_R Volume of world trade (weighted
  average), 2005 100
• E_EX_P Export deflator, 2005 100
• A_MEPAE_P Manufacturing export price for
  advanced economies, 2005100
• E_ER_PI Exchange rate BGUSD, 2005100

26
Exports
27
Imports

• Log(E_MP_R_STAR) -2.473 log(R_DD_R)
• 
  (13.023) (-)
• - 0.305log(E_MP_P/R_GDP_P)
• (3.821)
• Dlog(E_MP_R) 0.072 0.052(log(E_MP_R(-1))
• 
  (4.811) (0.645)
• -log((E_MP_R_STAR(-1)))
• E_MP_R Imports, prices of 2005
• R_DD_R Real domestic demand, prices of 2005
  (private consumption expenditure government
  consumption expenditure gross fixed capital
  formation)
• E_MP_P Import deflator, 2005 100
• R_GDP_P GDP deflator, 2005 100

28
Imports
29
GDP deflator

• Log (R_GDP_P_STAR) -0.011 log(L_ULC)
• (0.482)
  (-)
• Dlog(R_GDP_P)0.017-0.030(log(R_GDP_P(-1))-
• (2.202) (1.002)
• log(R_GDP_P_STAR(-1)))-0.071dlog(R_GDP_P(-
• (0.397)
• -1))0.100R_YG_R
• (-)
• R_GDP_P GDP deflator, 2005100
• L_ULC Unit labour cost, 2005100
• R_YG_R Output gap ( potential GDP)

30
GDP deflator
31
HICP without administered prices

• Log(I_HICPEXA_P_STAR) 0.007
  0.488log(R_GDP_P)
• 
  (0. 588) (4.155)
• 0.512log(E_MP_P)
• (-)
• Dlog(I_HICPEXA_P) 0.029 - 0.381(log(I_HICPEXA_P
  (-1))
  (5.701) (3.891)
• log(I_HICPEXA_P_STAR(-1)))
• 0.220Dlog(I_HICPEXA_P(-1))
• (1.526)
• I_HICPEXA_P Harmonized Index of Consumer Prices
  without administered prices, 2005 100
• L_ULC Unit labour cost, 2005100
• E_MP_P Import deflator, 2005 100

32
HICP without administered prices
33
Export deflator

• Log(E_EX_P_STAR) -2.313 0.592log(R_GDP_P)
• 
  
  (5.052) (3.376)
• 0.408log(A_MEPAE_PE_ER_PI)
• (-)
• Dlog(E_EX_P) 0.010 0.853(log(E_EX_P(-1))-
• 
  (2.110) (4.548)
• log(E_EX_P_STAR(-1))) -0.071Dlog(E_EX_P(-1))-
• 
  
  
  (0.697)
• 0.058Dlog(E_EX_P(-2)) 0.381Dlog(E_EX_P(-3))-
• (0.580)
  (3.798)
• 0.264Dlog(E_EX_P(-4)))
• (2.590)
• E_EX_P Export deflator, 2005 100
• A_MEPAE_P Manufacturing export price for
  advanced economies, 2005100
• R_GDP_P GDP deflator, 2005 100
• E_ER_PI Exchange rate BGUSD, 2005100

34
Export deflator
35
Import deflator

• Log(E_MP_P_STAR)-2.284 0.360log(A_EU15MP_PE_E
  R_PI)
• (2.810) (4.710)
• 0.640log(R_GDP_P)
• (-)
• Dlog(E_MP_P) 0.025 -0.116(log(E_MP_P(-1))
• (2.400) (0.827)
• -log(E_MP_P_STAR(-1)))- 0.381Dlog(E_MP_P(-1))
• 
  (2.530)
• E_MP_P Import deflator, 2005 100
• A_EU15MP_P EU 15 Import deflator, 2005 100
• E_ER_PI Exchange rate BGUSD, 2005100
• R_GDP_P GDP deflator, 2005 100

36
Import deflator
37
Wages

• Log(L_W_STAR) -5.822 log(L_LPR) 0.001L_UR
  
• 
  (3.087) (-)
  (0.712)
• log(I_HICP_P)
• (-)
  
• Dlog(L_W) 0.131 -0.207(log(L_W(-1))
• (10.598) (1.138)
• - log(L_W_STAR(-1))) - 0.352Dlog(L_W(-1))
• 
  (1.739)
• L_W Nominal wage bill per worker
• L_LPR Labour productivity GDP in prices of
  2005/Number of employees
• L_UR Unemployment, in percentages
• I_HICP_P Harmonised index of consumer prices

38
Wages
39
Fiscal sector

• Government expenditures and government revenues
  are modelled separately
• The government expenditures are disaggregated
  into five parts government consumption,
  government investment, government transfers,
  government interest payments and other
  expenditure.
• The government revenues consist of five
  components revenues from personal income tax,
  social security contribution, revenues from
  corporate income tax, revenues from indirect
  taxes and other revenue items.

40
Personal income taxes

• G_PIT G_PIT_TRR_CE_N
• G_PIT Personal income taxes (million leva)
• G_PIT_TR Personal income effective tax rate
• R_CE_N Compensation of employees (million leva)

41
Social security contribution

• G_SSCG_SSC_TRR_CE_N
• G_SSC Social security contribution (incl.
  employers and employees contribution in
  million leva)
• G_SSC_TR Social security effective tax rate
• R_CE_N Compensation of employees (million leva)

42
Indirect taxes

• G_INDG_IND_TRR_C_N
• G_IND Indirect taxes (incl. VAT, customs
  revenue, excise duties)
• G_IND_TR Indirect effective tax rate
• R_C_N Private consumption, in current prices

43
Corporate income tax

• G_CIT G_CIT_TRR_GOS_N
• G_CIT Corporate income tax
• G_CIT_TR Corporate income effective tax rate
• R_GOS_N Gross operating surplus and mixed
  income

44
Simulations

• To illustrate the simulation properties of the
  BNBQM, we assess the response of the models main
  variables to the following standard shocks
• an increase in government consumption by 1 of
  GDP
• an increase in volume of world trade by 1
• a depreciation of the lev against the US dollar
  by 1
• an increase in the price of oil by 10
• an increase in EURIBOR by 100 basis points

45
Simulation of an increase in government
consumption by 1 of GDP (Q12002
Q42008)Levels, percentage deviations from
baseline

• Year 1 Year 2 Year
  3 Year 4 Year 5 Year 7
• HICP 0.031 0.162
  0.319 0.476 0.607 0.666
• ULC -0.506 0.040
  0.316 0.538 0.670 0.709
• Comp. per employee 0.385 0.794 0.917
  0.933 0.840 0.450
• Productivity 0.896 0.753
  0.599 0.393 0.168 -0.256
• GDP 1.012 1.058
  0.973 1.053 0.498 -0.119
• Private consumption 0.060 0.333 0.517
  0.575 0.534 0.257
• Investment 0.003 0.009
  0.014 0.017 0.020 0.020
• Exports -0.009 -0.086
  -0.198 -0.300 -0.387 -0.427
• Imports 0.042 0.196
  0.386 0.595 0.781 1.099

46
Government consumption shock

• An increase in the government consumption boosts
  domestic demand and raises GDP by 1.01 as a
  primary effect. This effect remains for 5 years,
  after which it gradually dies out due to
  secondary effects.
• The fiscal expansion stimulates production and
  investment.
• Nominal wages and prices go up. Higher employment
  and wages lead to higher personal incomes and to
  an increase in consumption.
• Larger investment and stronger private and
  government consumption leads to increasing
  imports.
• Expanding economic activity translates into
  widening of the output gap that pushes up the
  price level.

47
Simulation of an increase in world demand by 1
(Q12002 Q42008)Levels, percentage deviations
from baseline

Year 1 Year 2 Year 3
Year 4 Year 5 Year7 HICP
0.005 0.071 0.204
0.379 0.564 0.807 ULC
-0.255 -0.205 0.077
0.366 0.551 0.788 Comp. per employee
0.101 0.554 0.884 1.076
1.144 1.010 Productivity
0.357 0.760 0.807 0.708
0.589 0.220 GDP
0.395 0.971 1.175 1.173
1.100 0.545 Private consumption 0.007
0.158 0.397 0.576 0.662
0.571 Investment 0.001
0.005 0.011 0.016 0.021
0.027 Exports 0.725
1.546 1.583 1.497 1.426
1.320 Imports 0.001
0.029 0.106 0.235 0.390
0.728
48
An increase in world demand

• This simulation is particularly important because
  of the openness of the Bulgarian economy. The
  external demand shock leads to a stronger
  domestic demand. The external shock directly
  drives up the volume of exports by 1.55 (second
  year), which in turn also increases imports.
• Employment and nominal wages increase which leads
  to higher private consumption (0.57-seventh
  year).
• Higher aggregate demand widens the output gap
  that pushes up the aggregate price level.

49
Simulation of a depreciation of the lev against
the US dollar by 1(Q12002 Q42008)Levels,
percentage deviations from baseline

• Year 1
  Year 2 Year 3 Year 4 Year 5 Year 7
• HICP 0.008 0.065
  0.129 0.190 0.250 0.310
• Import deflator 0.067 0.190
  0.279 0.337 0.390 0.005
• Export deflator 0.336 0.565
  0.493 0.511 0.572 0.006
• ULC -0.067 -0.009
  0.056 0.137 0.203 0.003
• Comp. per employee 0.032 0.145 0.236
  0.328 0.387 0.004
• Productivity 0.099 0.154
  0.180 0.191 0.184 0.001
• GDP 0.110 0.205
  0.266 0.308 0.325 0.288
• Private consumption 0.003 0.044 0.090
  0.118 0.133 0.011
• Investment 0.000 0.001
  0.002 0.003 0.004 0.001
• Exports 0.198 0.289
  0.294 0.308 0.295 0.002
• Imports -0.003 -0.028
  -0.066 -0.098 -0.110 -0.098

50
An exchange rate shock

• The decrease in the value of the lev against the
  US dollar has an immediate impact on both the
  import and export deflators they both increase.
  
• HICP increases by 0.25 in the fifth year.
  Compensation per employee adjusts and income
  increases which drives the consumption up.
• Because of the relative increase in foreign
  prices, imports decrease and exports increase
  slightly. The reaction of real GDP to an exchange
  rate shock achieves its maximum in the fifth year
  (0.32).

51
Simulation of an increase in the price of oil by
10(Q12002 Q42008)Levels, percentage
deviations from baseline

• Year 1 Year 2 Year 3 Year 4
  Year 5 Year7
• HICP 0.016 0.110
  0.177 0.227 0.296 0.413
• Import deflator 0.160 0.468
  0.656 0.851 1.052 1.299
• Export deflator 0.481 0.814
  0.618 0.886 1.031 1.103
• ULC 0.032 0.110
  0.084 0.117 0.185 0.204
• Comp. per employee -0.005 -0.055 -0.048
  0.028 0.088 0.352
• Productivity -0.037 -0.164
  -0.132 -0.089 -0.097 0.148
• GDP -0.041 -0.194
  -0.186 -0.140 -0.146 0.220
• Priv. consumption 0.000 -0.012 -0.067
  -0.125 -0.171 -0.231
• Investment 0.000 -0.001
  -0.003 -0.004 -0.005 -0.006
• Exports -0.081 -0.394
  -0.413 -0.432 -0.593 -0.621
• Imports -0.005 -0.057
  -0.154 -0.280 -0.424 -0.774

52
An oil price shock

• An oil price shock leads to increasing domestic
  prices through direct channels - rising import
  prices. The response grows gradually and achieves
  its maximum - 0.41 for HICP in the seventh year.
• The rise in prices causes a reduction in the
  demand for domestic and foreign goods and the
  consequent fall in household consumption, imports
  and exports. As a result, GDP and real disposable
  income are below baseline till year six.
• Higher oil price has a negative impact on output
  and domestic demand (private consumption declines
  by 0.17 and GDP falls by 0.15 in the fifth
  year).
• Reduced economic activity leads to a lower demand
  for imports, real exports are also decreasing due
  to an increase in domestic prices.

53
Simulation of an increase in EURIBOR by 100 basis
points (Q12002 Q42003) Levels, percentage
deviations from baseline

• 
  
  
  
  
  
  
  
  
  
  
  
  
  Year 1 Year
  2 Year 3 Year 4 Year 5 Year7
• HICP -0.001 -0.007
  -0.017 -0.021 -0.015 0.015
• ULC 0.026 0.011
  -0.049 -0.054 -0.031 -0.010
• Comp. per employee -0.010 - 0.051 -0.054
  -0.001 0.041 0.082
• Productivity -0.086 -0.195
  -0.156 -0.052 0.037 0.220
• GDP -0.096 -0.269
  -0.291 -0.182 -0.059 0.239
• Private consumption -0.066 -0.182 -0.243
  -0.289 -0.342 -0.359
• Investment -0.283 -0.933
  -1.256 -1.118 -1.038 -0.890
• Exports 0.000 0.000
  0.001 0.001 0.000 0.001
• Imports -0.006 -0.071
  -0.214 -0.396 -0.571 -0.865

54
Interest rate shock

• Because of the fixed exchange rate regime the
  exchange rate does not react to domestic interest
  rate changes. The main effect is through
  investment- raising costs of capital and reducing
  output. Investment decreases by 1.12 and GDP is
  reduced by 0.18 in the fourth year.
• The effect on domestic prices is negative due to
  a weaker domestic demand.
• Due to lower internal aggregate demand,
  employment and wages fall.

55
Concluding remarks and extensions

• A first step towards building a structural
  macroeconomic model.
• This practical work gives valuable information
  for the future development of the model which
  needs to be continuously developed and could be
  improved in a number of respects
• Availability of new data will require
  re-estimation and re-calibration of the model

56
Concluding remarks and extensions

• Developing a long-run baseline that reflects a
  fully theory-consistent long-run steady state
• To consider policy rules in the simulations
• Developing a more detailed representation of the
  trade block by including services on the one hand
  and different regions on the other hand
• An extension of forward-looking behaviour.
  Expectations should be incorporated, particularly
  to allow for a specific role in price and wage
  formation.

57
Baseline (work in progress)

• Real exports to GDP

58
Baseline (work in progress)

• Real imports to GDP

59
Baseline (work in progress)
Real private consumption to GDP
60
Baseline (work in progress)

• Real capital formation to GDP