Transcript Benchmarking - United Nations Statistics Division

Workshop on the Methodological Review of
Benchmarking, Rebasing and Chain-linking of
Economic Indicators
24-26 August 2011, Vientiane, Lao People’s
Democratic Republic
Vu Quang Viet
Why benchmarking?
 Economic statistics and national accounts are estimates
that are extrapolated from a base year using short-term
indicators.
 Data from a base year is considered the most reliable as
they are based on economic census that covers the
complete population.
 Short-tem indicators are based on survey (or
administrative report) of a limited number of units in an
activity that deems to signify the full activity.
 When a new base year is compiled, the estimates must
be benchmarked to the data of the new base year.
What are normally benchmarked?
 Quarterly values are benchmarked so that the sum
of quarterly values is the same of the annual value.
 Annual estimates should be benchmarked so that
the last annual estimate should match the
benchmark value.
 These benchmarking applies to an individual
statistics or a composite statistics such as quarterly
GDP and annual GDP.
 Preferred approach: benchmark each individual
statistic series.
Indicators
 Economic performance indicators are mostly drawn
from annual or quarterly accounts and therefore are
fully consistent with one another and provide useful
overview of the economy, its strength as well as
weakness.
 However, they need to be supplemented by other
important indicators that are prepared from
specialized statistics such as monetary and
government budget statistics.
 All indicators would be more meaningful in the
context of changes over time, therefore, time series of
statistics are required.
National accounts aggregates as
indicators
 Indicators based solely on national account aggregates.
 Familiar indicators are: rate of growth in GDP, final
consumption, investment in fixed assets, saving rate
(saving/GDP), investment rate, effective individual and
corporate tax rates, etc.
 These indicators can be derived directly from national
account data. They allow to compare not only the
performance of the economy over time but also to other
countries of the same level of development.
 Consumer prices indexes.
 Producer price indexes (or wholesale price indexes).
Indicators that relate national accounts
with other indicators
 Government budget balance / GDP.
 Current external account balance / GDP.
 Foreign debt payment / export (which includes both
interest payment and payment principal).
 Etc.
Indicators that are used to
national accounts aggregates
 Industrial production indexes.
 Crop yield indexes.
 Employment indexes, based on:
 Establishment survey that captures only employment in
formal activities that are covered by updated census
frame
 Household survey that captures employment in
informal activities.
 Retail sale indexes.
 Investment (GCF) indexes.
Use of indicators for extrapolation
 Base year: 2000.
 It-1,t : Volume index indicating growth from t-1 to t.
 Q: Value in constant prices.
Q2000,t = Q2000,t-1*It-1,t
Other specialized indicators
 Non-performing loan ratio.
 Foreign exchange reserves (reserve over average
monthly imports).
 Short-term liability denominated in foreign
currencies.
 Leading, coincident and lagging indicators to track
the economy:
 Inventory over sale (leading indicator)
 GDP (coincident indicator)
 Employment (lagging indicator)
Comments on indicators
 To be useful, indicators must be timely.
 Quarterly accounts are extremely useful.
 For quarterly accounts, other indicators must be
available monthly, particularly important:
 CPI, PPI
 Industrial production indexes
 Employment
 Retail sales
UN recommenations on shortterm indicators
 To be distributed.
Topics to be discussed
 Benchmarking a series of values of annual estimates
to match the new annual value of the benchmark
year.
 Benchmarking growth rate approach
 Benchmarking the sum of the quarterly GDP to the
annual value of GDP.
 Pro rate distribution of discrepancy method
 The Bassie method
 The Denton method
 Linking seasonally unadjusted quarterly data
Scheme for growth and value
benchmarking of annual data
Benchmark
year 2 =13
Benchmark
year 1 =10
Annual
Estimates
Conditions:
 New rates of growth are close to the old rates of growth
 The new rates of growth should permit the obtaining of
the new benchmark value at the new benchmark period.
Example for benchmarking annual
values
Time period
Preliminary GDP
Actual benchmark value
GDP after
benchmarking
1
10
2
11
3
11.5
4
12
13
10
11.3
12.1
13
Method for benchmarking
 Find the percentage growth difference between the
estimate and the new benchmark for the same
benchmark year.
= 13/12=1.083

 Distribute that percentage difference to the n-years
in the old series.

= (1.083)^(1/3) = 1.027
 New rate of growth = old rate * ig
A benchmarking is linking
 This will be discussed by Benson Sim.
Benchmark quarterly data to
annual data – pro rata method
Quarterly
Values
Total annual
Value
New Quarterly
Values
Q1
970.0
977.1
Q2
995.7
1003.0
Q3
1009.5
1016.9
Q4
995.7
Sum
3970.9
1003.0
4000.0
Percentage
Difference
4000.0
4000/3970.9
=1.07
Problems of pro-rata method
Quarter
values
Rates of
change
Annual
value
Step
problem
New Quarterly Adjusted rates
values
of change
1991
Q1
970.0
Q2
995.7
2.6%
1003.0
2.65%
Q3
1009.5
1.4%
1016.9
1.39%
Q4
995.7
-1.4%
1003.0
-1.37%
3970.9
1.007
Q1
977.9
-1.8%
1018.5
1.5%
Q2
1003.6
2.6%
1045.2
2.6%
Q3
1014.4
1.1%
1056.5
1.1%
Q4
1002.6
-1.2%
1044.2
-1.2%
Sum
3998.5
1.041
4164.4
4%
Sum
977.1
4000
4000
1992
4164.4
Other methods
 Bassie method aims at allocating the values so as to
reduce the reduction in the first year and increase the
value in the second year to reduce the step problem.
 Denton method aims at minimizing the difference in the
rates of growth of the old and the new series, while
maintaining the annual value equal the sum of the
quarters. This requires software. STATA IS THE
SOFTWARE THAT CAN DO THE JOB.
 Observations: All methods are mechanical. It is thus
important to review the indicators and see if they are
good indicators or if they should be replaced.
Index linking seasonally unadjusted
data: method
1. Convert quarterly data of the two consecutive
years to constant values
2. Calculate seasonal index of T (of the current year)
as compared to T-4 (of the previous year), SIT .
SIT = VT / VT-4
3. Link by seasonal indexes to the previous index
series by multiplying the seasonal index to the
index of the same quarter.
IT = SIT * IT-4
4. (See Exel-file).