Transcript P6.2

Economic Monitoring in an
Uncertain Global Environment
Processes and Tools from the World Bank
Soong Sup Lee
Hans Timmer
Gauresh Rajadhyaksha
The World Bank Group
Washington, DC
Presentation at the International Seminar on Early Warning and Business Cycle
Indicators, Scheveningen, The Netherlands
Monitoring and High Frequency data
• Who we are
– Team of macro-economists
– Provide “data-driven, forward looking analyses”
• Produce Bank global forecasts
• Provide planning, investment, and policy-advice to Senior
and Executive Management
– Monitor real and financial indicators
• Trade, financial flows, commodity prices, remittances flows
– Maintain Datasystems for high-frequency data
• Service internal and external clients with a daily-updating
high frequency data-set
Outline of the presentation
• Examples from some of our analytical work
during the onset of the financial crisis
• Details of our data infrastructure
• Conclusions and “lessons for the future”
1. “Revealed Vulnerability Index”
Identifying risk during the financial crisis
Change in
Spreads (Daily)
Change in
Current Account
Balances, %GDP
(Quarterly)
Change in
Equity Mkts
(Daily)
Exchange Rate
Depreciation
(Daily)
Index
Change in
Capital Flows,
%GDP
(Monthly)
2. “Credit Risk Monitoring”
Inputs to briefing for the Bank’s Risk Committee
• Monthly briefing to Sr. Management on the
Credit Risks of client countries
• We provide “global overview” that
supplements the “country details”
• We “aggregate” country data to provide a
broader implication of the existing climate
“Credit Risk Monitoring”
Example of input provided in Dec ‘ 08
IP now in decline across all regions
industrial production, ch% (3m/3m saar)
25
East Asia
20
15
10
5
0
All Developing
Latin America
-5
Europe and Central Asia
-10
Jan-07
Apr-07
Jul-07
Oct-07
Source: Thomson/Datastream, DECPG
.
Jan-08
Apr-08
Jul-08
Oct-08
3. Industrial Production Forecasting”
3 month IP index forecast
• Forecast created using OECD leading
indicators and ARIMA models
Example of input provided in Dec ‘ 08
Brazil’s industrial production advances supported by
domestic demand
Industrial production growth
Forecast
40
30
Momentum
(3m/3m saar)
20
10
0
-10
Year-on-year
growth
-20
-30
-40
-50
-60
Jan-00
Jul-01
Source: DECPG.
Jan-03
Jul-04
Jan-06
Jul-07
Jan-09
Supporting our diverse data needs
“World Bank’s High Frequency Data system“
Sourcing
“Value
Added”
• DataStream/Thomson Reuters, Bloomberg
• World Bank, UN, IMF, OECD databases
• Country Statistical Office websites
• Organize and classify
• Process: gap-fill, extend, seasonally adjust, rebase
• Share on internal and external websites (“Global Economic Monitor”
http://www.worldbank.org/gem)
• Create web-portals, integrate with our “forecasting tools” and produce Excel files
Dissemination
for download
Data Coverage
“World Bank’s High Frequency Data system“
•
•
•
•
•
•
•
•
•
Exchange Rates (145 countries, Daily)
Spreads (44 countries, Daily)
Policy rates (55 countries, Daily)
Stock Market Indices (70 countries, Daily)
Bond Indices (20 countries, Daily)
Trade: Imports, Exports (160 countries Monthly)
Industrial Production (85 countries, Monthly)
Commodity Prices (40 Daily, 87 Monthly)
Effective Exchange Rates (140 countries, Monthly)
Regional Aggregation
“World Bank’s High Frequency Data system“
• “Aggregation” of country data to provide
global picture is core to our analysis
• We create regional estimates based on weight
of reporting countries for that time period (>
2/3rd reporters)
• New data releases change estimates but
weighting criteria prevent rapid changes on a
day-to-day basis
Visualization and Dissemination
“World Bank’s High Frequency Data system“
• Data feeds to numerous websites, web-portals
and publications
– World Bank’s Global Economic Monitor website
• http://www.worldbank.org/gem
– World Bank’s Prospects for Global Economy website
• http://www.worldbank.org/prospects
– iSimulate @ World Bank (forecast system)
• http://isimulate.worldbank.org
– Internal World Bank data portal
• Will soon be available “programmatically” (using
an API ) within the Bank
Interactive Data Portal
“World Bank’s High Frequency Data system“
Heat Maps
to visualize
global
trends
Interactive
Charts to
compare
“Aggregate”
with
“country”
data
Conclusions
• We rely heavily on high frequency data (and
estimates) for our analyses
• We have learnt that “aggregating” high-frequency
data provides a unique flavor to any economic
analyses
• Modern-day graphing and visualization tools
make it very easy to compare country data with
larger groups and aid in information
dissemination
• And yes, our appetite for high frequency data is
only increasing!
Some Road-blocks …
• Our biggest challenge to-date has been in
“harmonizing” diverse reporting practices across
the world
– Same indicator is often reported with different
assumptions across countries
– Any harmonization efforts would be very much
welcomed by us!
• We source our data from commercial providers
and there is a limit to our “sharing” of this data
with our external customers
Lessons Learnt
Keys to Success
• At the onset of the financial crisis, our analysts could
respond in large part due to our robust Datasystems
that update information on a Daily basis
– Includes retrieving data, processing and harmonizing, and
creating aggregates
• High Frequency data is seen to be playing an
increasingly important role in various economic
modeling exercises
• Successful dissemination must cater to the needs of
both the novice and advanced users
– For eg: we provide data in Excel format to our novice user
and will soon provide programmatic access (via an API) to
our advanced users