Download Maldives National Macroeconomic Forecasting And Policy Simulation Models

Maldives
National Macroeconomic Forecasting And Policy Simulation Models
Content
• Maldivian Economy
• Forecasting Models
•
•
•
•
Process
Tourist Arrivals
GDP
Revenues
Real Sector
Co-Movement of GDP and Tourism
25%
60%
20%
15%
40%
10%
20%
5%
0%
2015
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
0%
2003
• Real GDP per capita in 2015: US$ 6,088.68 (projected)
• High dependency on tourism ‐
contributes about 30% to GDP
• Highly susceptible to external shocks
• Share of fisheries sector is small, but still a key source of employment for many
• Unemployment rate (broad definition): 28% in 2010
‐5%
Financial crisis
‐10%
‐20%
Indian Ocean tsunami
‐40%
‐15%
Tourism Sector Growth (Right Axis)
Real GDP Growth (Left Axis)
Real Sector ‐ Prices
Inflation
20%
Introduction of GST
18%
16%
14%
Rufiyaa devaluation
12%
10%
8%
6%
4%
2%
0%
Q1‐2009
Q2‐2009
Q3‐2009
Q4‐2009
Q1‐2010
Q2‐2010
Q3‐2010
Q4‐2010
Q1‐2011
Q2‐2011
Q3‐2011
Q4‐2011
Q1‐2012
Q2‐2012
Q3‐2012
Q4‐2012
Q1‐2013
Q2‐2013
Q3‐2013
Q4‐2013
Q1‐2014
Q2‐2014
Q3‐2014
Q4‐2014
Q1‐2015
Q2‐2015
• Inflation peaked in early 2012, reflecting the introduction of the GST and the devaluation of rufiyaa
• Gradually back to lower levels as these one off effects dissipated
• Food has a weight of 23.8% in the CPI basket
• Exchange rate pass through is extremely high, and international commodity price changes are reflected to a high degree
Inflation
Inflation excluding fish
Fiscal Sector
Revenue, Expenditure and Financing (% of GDP)
• Government expenditure consistently exceeds revenue
• Major revenue source was the import duty until 2011
• Now, GST (especially T‐GST) and the business profit tax are the main sources, and import duty rates have been reduced
• Largest share of government expenditure goes to employee remuneration
• 70% of total expenditure is recurrent expenditure
20%
160%
18%
140%
16%
120%
14%
100%
12%
80%
10%
8%
60%
6%
40%
4%
20%
2%
0%
0%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Total revenue and grants (% of GDP, left axis)
Total expenditure and net lending (% of GDP, left axis)
Deficit (% of GDP, right axis)
• Total debt has been rising sharply since 2008, mainly fuelled by rising domestic debt
• Newly introduced fiscal responsibility act stipulates that the debt‐to‐GDP ratio should be at most 60% by the end of 2016, and maintained below the threshold thereafter
80%
35,000
70%
30,000
60%
25,000
50%
20,000
40%
15,000
30%
10,000
20%
5,000
10%
0%
‐
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Total debt (right axis)
External Debt to GDP (left axis)
Debt to GDP (left axis)
Millions of Rufiyaa
Fiscal Sector ‐ Debt
External Sector ‐ Trade
• Trade deficit is expected to be at 56% of GDP in 2015
• Services account surplus is expected to be at 72% of GDP in 2015
2500
2000
Millions of USD
• Have consistently been experiencing trade deficits
• Current account deficit is expected to be at 7% of GDP in 2015
1500
1000
500
0
2005
2006
2007
2008
Domestic exports f.o.b.
2009
2010
2011
2012
Imports f.o.b.
2013
2014
2015
Trade deficit
External Sector ‐ Reserves
• Gross international reserves consist of the reserve assets held by MMA
5.0
• Built up mainly through tourism sector activity (tax receipts) and the accumulation of foreign currency assets by banks
3.5
• Challenge to intervene in FOREX market due to low reserve levels
• Reserves improving in recent years
800
4.5
700
4.0
500
3.0
400
Months
2.5
2.0
300
1.5
200
1.0
100
0.5
‐
0
2009
2010
2011
2012
2013
2014
Gross international reserves (right axis)
Gross international reserves import coverage (left axis)
Jun‐15
Millions of USD
600
Monetary Sector – Credit Growth
• Global financial crisis • Deterioration in the asset quality of banks (increase in NPLs)
• Diverted to investment in T‐bills
18,000.0
16,000.0
14,000.0
Millions of Rufiyaa
• Rapid increase in credit to government after 2008 partly reflects the widening fiscal deficit
• Expansion of credit to private sector stalled after 2008
12,000.0
10,000.0
8,000.0
6,000.0
4,000.0
2,000.0
‐
2005
2006
2007
2008
2009
Credit to central government
2010
2011
2012
2013
2014 Jul‐15
Credit to private sector
Monetary Sector – Interest Rates
T-bill Interest Rates
• This did not reduce T‐bill investment, reflecting the lack of other investment opportunities in the domestic market
• T‐bill interest rates mechanically halved in November2015.
11
10
9
8
Percent
• Interest rates increased dramatically for T‐bills of all maturities reflecting the widening fiscal deficit
• Tap system introduced in 2014 to curb the increasing interest burden on the government
7
6
5
4
3
2007
2008
28 days T‐bill
2009
2010
90 daysT‐bill
2011
2012
2013
182 days T‐bill
2014
Nov‐15
364 days T‐bill
Forecasting
Process
Macro Modeling
• Budget Process
• Macroeconomic Policy Coordination Committee • Medium‐term budgets
• Policy approval
• Medium Term Economic Framework
•
•
•
•
Being built with the help of World Bank
Brings together isolate models
Behavioral linkages
Excel based
Forecasting Models
GDP
Procedural Change
 National Bureau of Statistics (NBS) to discontinue production of annual medium term GDP forecasts‐ from 2015
 NBS to produce Quarterly National Accounts(QNA) –first release 20th
October • NBS will produce actual GDP, quarterly and annually • Not the international best practice for statistical agencies to produce annual forecasts
 EDPD/MoFT to produce annual forecasts for GDP growth and Nominal GDP Forecasting model
Univariate time series model
• Using the Quartely Real GDP series (2003Q1 to 2015Q2)
• Model using information on its own past values
• AR (1 ,4 ) SARIMA(0,1,0)4 model
Advantages
• Simple to model • Often found to produce better forecasts than structural models
• Incorporates information of the economic growth up to Q2 2015
Disadvantages
• Retrospective • Cannot dissect the growth impact of specific items
• A‐theoretical
Forecast – Univariate time series model
GDP GROWTH:2015‐2018
8.00%
7.50%
7.00%
6.50%
6.00%
5.50%
5.00%
4.50%
4.00%
2015
2016
2017
Annual GDP ARIMA(1,1,0) model
Annual GDP External Demand Model
QRGDP Trend + AR(1) + Seasonal dummies
QRGDP AR(1,4) + SARIMA(0,1,0,4)
2018
NGDP Forecasting method
GDP (EXP)
2014 GDP
2014 GDP
Expenditure breakdowns
C
Forecast growth for indicators 2015‐2018
Imports of Household Consumptio
n Goods
Indicator Weights 1 G
I
X (goods)
X (services)
M
Wage bill budget and forecast
Intermediat
e consumptio
n costs budget and forecast
Imports of capital goods
Imports of building material Govt Capital Expenditure budget and forecast
Exports of goods
Tourism Bednights
Imports of goods
Wage bill budget and forecast
Govt Capital Expenditure budget and forecast
0.75
0.25 0.6
0.3
0.1
1
1
0.9
0.05
0.05
Nominal GDP Series‐ October 2014 update 70000
35%
30%
60000
25%
50000
40000
15%
10%
30000
5%
20000
0%
10000
‐5%
‐10%
0
2001
2002
2003
2004
Source : National Bureau of Statistics (October 2014)
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Growth
MVR millions
20%
Issues with GDP forecasting
• Disconnected methods to forecast RGD and NGDP
• Not a critically scrutinized model
Forecasting Models
Tourist Arrivals
Medium term forecasts
• A global tourism elasticity‐based model for medium term forecasts
• Maldives is hyper‐sensitive to global tourism, with an estimated elasticity of 184 percent
Medium term model
• LOG(ARR_MDV) = C(1) + C(2)*LOG(ARR_GLOBAL) + C(3)*SHOCK + AR(1)
• Global arrivals forecasts from UNWTO is included. Within year forecast
• Temporal (or seasonal) based model for within year forecasts
• Seasonal patterns apparent and consistent
• High dependency on European tourists, peak season in Maldives corresponds to European winter months of December‐February. • However, since of late China has emerged as the single largest source of tourists and also the one which has shown the strongest growth. Interestingly, Chinese arrivals also reflect strong seasonal features but with effects opposite to that of the European segment. Within year (short term) model
• Calculate weights for main markets, namely, China, Europe and RoW.
• Run AR models for each market
• Multiply forecasts from AR models with weights to obtain short term monthly forecasts
Issues and updates
• Both models have relatively large standard errors.
• Shock variable (a binary variable) is exogenous to the model and is captured for only 1 period. This is not very robust and a better approach would be to have some indication of duration of a shock through an appropriate formulation.
• Models performed poorly in recent years.
• New research to update the model on‐going.
Forecasting Models
Revenues
Revenue Forecasting
• Historically done by collecting agencies.
• New Medium Term Economic Framework (MTEF) being built. Same models are incorporated to the MTEF.
• Over 90% of 2014 MIRA revenue comprise of 7 revenue codes:
•
•
•
•
•
•
•
GST (Tourism and Non‐Tourism Sectors)
BPT
Tourism Land Rent
Tourism Tax / Green Tax
Lease Period Extension Fee
Bank Profit Tax
Airport Service Charge
MVR billions
MIRA Revenue
14
11.50
12
8.98
10
7.15
8
6
4
4.56
2.41
2
0
2010
2011
2012
2013
2014
Others
Airport Service Charge
Bank Profit Tax
Lease Period Extension Fee
Tourism Tax
Tourism Land Rent
BPT
GST
Total
Tourism Sector GST
• Bednights information and forecast – from the Ministry of Tourism
• Spending Rate
• Calculate the ‘consumption’ of the tourism sector
• Equals to TGST collection / TGST rate
• Calculate Spending per tourist per bednight (spending rate)
• Tourism sector consumption / bednights
• Estimate the spending rate for the projection period
• Take a moving average of the past years’ spending rate
• Increase it by a rate indicative to the growth in past year’s spending rate and inflation
• TGST revenue = Spending Rate x Bednights x TGST Rate
Spending Rate
500
450
400
350
300
250
200
150
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2011
2012
2013
2015
2016
2017
2014
Non‐Tourism Sector GST
• Calculate the consumption subject to GST in the non‐tourism sector
• Equals to GGST collection / GGST Rate
• Analyse the consumption trends
• Analyse GDP and inflation
• GDP forecast not available at the time of projection
• Estimate a crude growth rate for the real GDP based on past growth rates
• Estimate a crude inflation figure for the projection period based on past figures
• Estimate nominal GDP growth based on inflation and real GDP growth rate
• Estimate the consumption for the projection period based on moving average of past consumption levels, increased by nominal GDP growth rate.
• Additional weight may be necessary to reflect growing compliance • GGST revenue = Consumption x GGST Rate
Airport Service Charge
• Tourist arrivals estimates are obtained from the Ministry of Tourism
• Calculate the share of foreign departures based on information obtained from Airport operators
• Share of foreigners in the total departures: 88.4%
• Share of tourists among foreign departures: 92.1%
• Estimate foreign departures
• Tourist arrivals divided by ‘share of tourists among foreign departures’
• Estimate total departures
• Foreign departures divided by ‘share of foriegners in the total departures’
• Estimate Maldivian departures to abroad
• Total departures minus foreign departures
• Multiply with respective ASC rates • USD 25 for foreigners and USD 12 for Maldivians
• Airport Service Charge = ASC from Foreigners + ASC from Maldivians
Economic Forecasting in Maldives
• Various models, various institutions and various software.
• The teams involved in forecasting have close dialogue and work together on some models (eg: new tourist arrivals model)