VTI rapport
No. 413A - 1996
nese Heavy Duty A Forecast for Ch
Truck Demand
Ma Yun
Swedish National Road and ansport Research Institute Ir
VTI rapport
No. 413A 0 1996
A Forecast for Chinese Heavy Duty
Truck Demand
Ma Yun
Swedish National Roadand
Publisher: Publication:
VTI rapport 413A
Published: Project code:
Swedish National Road and 1996 50015
' Transport Research Institute
8 581 95 Linkoping Sweden Project:
A Forecast for Chinese Heavy Duty Truck Demand
Author: Sponsors:
Ma Yun Volvo Truck Corp.
Swedish Transport and Communications Research Board (KFB)
Title:
A Forecast for Chinese Heavy Duty Truck Demand
Abstract (background, aims, methods, results) max 200 words:
China is probably the automotive industry s last fast growing market. In this report, we have given a
forecast for the demand for Chinese heavy duty trucks (payload _>_ 8 ton) on the level of national and re
gional groups. The report also contains a forecast for the total population of trucks and for heavy duty trucks in use. The forecast years are 2000, 2005 and 2010. There are mainly three steps in the approach of this forecast. The first step of six different methods gives a forecast for the total number of trucks in use
at the end of the forecast years. The second step focuses on the forecast for the population of heavy duty
trucks. Two different methods were used in this step. The third step gives a forecast for the demand for
heavy duty trucks for the forecast years. The main results of this report have been obtained with several different methods. The demand for heavy duty trucks per year will be 95,000 vehicles in the year 2000,
150,000 in 2005 and 210,000 in 2010.
ISSN: Language: No. of pages:
Foreword
Acknowledgment is given to Ulf Norman, Sven Olof Nilsson and Rolf Smedberg (Volvo Truck Corp.) for introducing me to this project and for supporting me in
my research.
The work has benefited greatly from the advice of Bertil Agren and Carl Magnus Berglund.
The author is indebted to research leader Dr. Jan Eriksson for his support in the research.
Thanks are also given to Research Director Borje Thunberg and to my col-leagues in the Transport Economics Group.
This research was sponsored by Volvo Truck Corp. and the Swedish Transport
and Communications Research Board (KFB).
The author is a guest researcher in the Transport Economics Group, Swedish
Road and Transport Research Institute (VTI, S 581 95 Linkoping, Sweden) and lecturer at the College of Economics, Northern Jiaotong University (Beijing,
100044, China).
Table of contents 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.3.1 2.3.2 2.3.3 3.1 3.2 3.3 4.1 4.1.1 4.1.2 4.2 4.2.1
4.2.2
4.2.3
4.2.4
4.3
4.3.1
4.3.2
4.4
4.4.1
4.4.2
Summary Introduction Project backgroundThe chinese economy Highways
The truck industry in China Forecast results
A Forecast for Chinese heavy duty truck demand
A Forecast for Chinese heavy duty truck demand by region
Other forecast and data analysis results Regional truck disaggregations
HDT percentages of total trucks Number of HDT s by province
Data sources
Data from the China Statistical Yearbook
Data from the Chinese National Heavy Duty Truck Corp. (CNHDTC) Data from the Motor World Vehicles Data
Forecast methods
Forecast for the total number of trucks Direct truck forecast
Forecast based on road freight transport demand
Forecasting the number of HDT s
Relationship between the number of trucks and total truck capacities
Forecast at total country level
Forecast based on a regional level study HDT forecast result
Forecast for yearly HDT demand
Forecast based on the number of HDT s Forecast based on the growth of HDT sales
General comment on the forecast method and results The present Stage of China s transportation development
Discussion on validity and reliability of data and forecast method
in this study
Chart list References
Appendix
VTI rapport 413A
13 13 13 13 13 15 15 15 16 16 17 17 18 18 18 18 19 19 19 21 Ar 23 23 24 25 26 26 27 28 29 29
3o
32
44
A Forecast for Chinese Heavy Duty Truck Demand
by Ma Yun
Swedish National Road and Transport Research Institute (VTI)
8 581 95 Linkdping Sweden
Summary
The People s Republic of China (PR. China) is currently one of the fastest growing economies in the world. China s huge population of 1.2 billion people and an anticipated high GDP (gross domestic product) growth rate in the next few
decades promise surging vehicle demand in what is probably the automotive in
dustry s last fast growth market.
In this report, we present a forecast for Chinese heavy duty truck (HDT, pay-load 2 8 ton) demand over the entire country and in regional groups. We also pre
sent a forecast for the total numbers of trucks and heavy duty trucks. The forecast years are 2000, 2005 and 2010.
The modeling approach for this forecast is restricted by the limited data available.
There are three main steps in the forecast approach (see Chart on next page). The first step gives a forecast for the total population of trucks in China at the end of each forecast year. In this step, six different forecast methods are used. The forecast methods can be divided into two groups: the first method is to forecast the number of trucks directly and the second is first to forecast road freight transport
demand, then estimate truck demand.
Based on the result of the first step, the second focuses on the forecast of the
number of heavy duty trucks. Two different methods are used in this step and the results are very close. The first method is to study the percentage of heavy duty trucks compared to total trucks at national level directly. The second is based on a regional level study.
The third step gives a forecast for the demand for heavy duty trucks in the fore cast years. Here we also use two forecast methods: the first is based on the popu
lation forecast for heavy duty trucks and the average life expectancy of heavy duty trucks. The second is based on the expected volume of sales and increase in the number of heavy duty trucks.
II
FC
ro
ss
-s
ec
ti
io
sn
st
ud
y,
v
3
re
gi
on
s
i
fo
re
ca
st
19 93 tr uc k us e da ta Re gi on al ' aéf
or
ec
as
ts
.
:
H D T fo re ca st g
Ti
me
-s
er
ie
s
st
ud
y
[p
er
ce
nt
ag
es
5
19 65 -1 99 33 tr uc k us e da ta H D T us e at H D T th e en d of : de ma nd:e
ad
l
pe
r
ye
ar
fo re ca st ye arLo ng it ud in al da ta re gr es si on st ud y
of
Ti me -s er ie s st ud y Se co nd me th od se t . Re gr es si on da tafo
re
ca
st
:P GT avera ge
tr
uc
k
Ec on om ic fa ct or s af fe ct in g H D T sa le s -- - ---_-_~__-____-_
In
di
re
ct
fo
re
ca
st
Ot he r mo de l Fi rs t St ep Se co nd St ep 5 Th ir d St epVTI rapport 413A
Fi gur e A Th ef or ec as t ap pr oa ch .
VTI rapport 413A 111 _ pe rc enta geof ' re gi onC 1992 : 14.8 % 2000 : 10 .6% 20 05 : 9.4% 20 10 : 8. 4% pe rcen ta geof regi onA 1992 :49 .0 % 20 00: 58.7 % I. 2005 :63 .5% K 2010 :66 .5% perc en tage of re gi onB 19 92: 35.7 % 2000 : 30.7 % 2005 : 27.0 % --'" . Re gion c 2010 : 25.1 % 1-Fi gur e B Popul at io nof H D T InUs e Sp lit in to Re gion Gr oup s. -_\ Re gi onA Re gi on B
The forecast results of this report are based on analysis of development and policy regarding economy, transportation and the truck industry. The main results of this report have been obtained with several different methods.
The main forecast results:
Table A Forecastfor Chinese heavy duty truck demand
(10 000 vehicles).
Forecast year 2000 2005 2010
Annual HDT market demand 9.5 15 21
HDT s
(population end of year) 75 115 165
Total truck use
(population end of year) 830 1 150 1 530
Table B Forecastfor Chinese heavy duty truck demand by region
(10 000 vehicles).
Forecastyears 2000 2005 2010
HDT use
(population at end of year)
Total 75: 115: 165:
Region A 44 73 110
Region B 23 31 41
Region C 8 11 14
Total truck use
(population at end of year)
Total 830: 1 150: 1 530:
Region A 430 660 950
Region B 280 350 420
Region C 120 140 160
On the basis of economic growth rate analysis and the GDP per person for each
province, the Chinese market is divided into three regions. Region A is a deve
loped region with a high economic growth rate, Region B is a region with a me-dial economic growth rate and Region C is an undeveloped region with a low economic growth rate.
Region A: Beijing, Tianjin, Shanghai, Liaoning, Jiangsu, Zhejiang,
Fujian, Shandong, Guangdong, Hainan
Region B: Hubei, Shanxi, Jilin, Anhui, Jiangxi, Henan, Hubei, Hunan, Guangxi, Sichuan
Region C: Inner Mongolia, Heilongjiang, Guizhou, Yunnan, Tibet,
Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang
1 Introduction
The People s Republic of China (P. R. China) is currently one of the fastest growing economies in the world. China s huge population of 1.2 billion people and an expected high GDP (gross domestic product) growth rate during the next few decades promise surging vehicle demand in what is probably the automotive industry s last fast growing market.
1.1 Project background
The aim of this study is to present a forecast for the expected truck population in China for the years 2000, 2005 and 2010. The main focus is put on heavy duty trucks, which are the subject of special forecast.
Based on the analysis, investigations in China and data collection, this forecast
project was finished in July 1995. It gives apicture of future demand for Chinese heavy duty trucks (HDT s, payload >= 8 tons) in the country and by region. It also gives a View of future total and heavy duty truck numbers. The forecast years are 2000, 2005 and 2010.
The forecast results of this project are based on analysis of development and economic policies, transportation and the truck industry. The main results of this
project have beenobtained with several different forecast methods.
1.2 The Chinese economy
Following China s reform in 1978, its economy has been transformed. Since then, the average economic growth rate has been 9% per year. The GNP (gross national
product) annual growth rate was 13.4% and 17.2% in 1993 and 1994 respectively. A rapid formation of collective enterprises has reduced the share of the old state enterprises in total employment from 81% to less than 50%. During the next ten years, the target for the average economic growth rate is 8 9% per annum.
China has a vast land mass and the state of economic development is un balanced among its provinces. Most of the recent economic development in China
has taken place in the eastern part of the country where nine coastal provinces and
three major coastal cities contain 40% of the population, 60% of GNP, 60% of the heavy duty truck eet, but only 13% of the land mass.
1.3 Highways
There were 1 080 000 km of highways in P. R. China in 1993. Of these, 4 633 km
were high grade. The total annual turnover volume of road freight transport was 440 billion ton km in 1994 and is expected to reach 750 billion ton km at the beginning of the next century. The development of China s highway system has become a high priority, far reaching goal. According to the Ministry of Commu
nications (MOC), which is responsible for the construction and maintenance of highways outside urban areas, a total of 12 national truck highway systems
covering between 30 000 35 000km of high grade highways will be built over
the next 30 years.
1.4 The truck industry in China
In 1992, the total population of trucks in P. R. China was 4.41 million (China Sta
tistical Yearbook 1994). Of these, the total number of heavy duty trucks was about 380 000. The national production capacity of heavy duty trucks was 18 000
vehicles in 1990 and 37 000 vehicles in 1993, with an average growth rate of 27%
per year, but internal production could still not meet the demand in those years. Therefore, from 1990 to 1993, 24 000 vehicles were imported. The total sales
volume of heavy duty trucks in China was about 16 000 vehicles in 1981 and 52 000 vehicles in 1993. The average growth rate was about 10% per year.
Along with the national economic growth andthe improvements of the high-way system, the highhigh-way transportation structure will change from middle ton-nage to heavy tonton-nage, this being a result of the development of logistics and the profession alisation of freight transport.
2 Forecast results
2.1 Forecast for Chinese heavy duty truck demand
Table 2.1 (10 000 vehicles).Forecast year 2000 2005 2010 __ _
Annual HDT market demand 9.5 15 21
HDT s
(population at end of year) 75 1 15 165
Total truck use
(population at end of year) 830 1150 1530
2.2 A Forecast for Chinese heavy duty truck demand by
region
Table 2.2 (10 000 vehicles).
Forecast years 2000 2005 2010
HDT s
(population at end of year)
Total 75: 115: 165:
Region A . 44 73 110
Region B 23 31 41
Region C 8 11 14
Total trucks (population at
end of year)
Total 830: 1150: 1530:
Region A 430 660 950
Region B 280 350 420
Region C 120 140 160
Based on Chinese provincial economic growth rates and the GDP per capita,
the market is divided into three regions (see map on next page), Region A is a
developed region with high economic performance, Region B is a region with medial economic performance and Region C is an undeveloped region with low
economic performance. '
Region A: Beijing, Tianjing, Shanghai, Liaoning, Jiangsu, Zhejiang, Fujian,
Shandong, Guangdong, Hainan
Region B: Hebei, Shanxi, Jilin, Anhui, Jiangxi, Henan, Hubei, Hunan,
Guangxi, Sichuan
Region C: Inner Mongolia, Heilongjiang, Guizhou, Yunnan, Tibet, Shananxi,
Gansu, Qinghai, Ningxia, Xinjiang
2.3 Other Forecast and Data Analysis Results
Table 2.3 Ton km offreight transport.Forecast Year
2000
2005
'
'
2010
Ton km (100 000 000) 7 500 10 500 14 000
Table 2.4 Total HDT demandfrom year 1994 (based on the assumption that life expectancy ofHDT 's is 12 years).
Forecast Year 1 994 - 2000 1994 2005 1994 201 0
Vehicles (10 000)
50
115
198
Table 2.5 Forecast Chinese annual GDP growth rates (%).
Time period 1994 2000 2000 2005 2005 201 0
National average 9 7 a 6
Region A 12 10 8
Region B H 7 _ 6 5
Region C 4 4 4
2.3.1 Regional truck disaggregations
The map on page III shows that the percentage of HDT s in region A, the region with a high economic growth rate, is expected to increase from 49 % in 1992 to 59 % in 2000 and 67 % in 2010.
However, the percentages in region B and region C are expected to decrease. Table 2.6 Relative regional distribution ofHDT 's (%.)
Years 1992 2000 2005 2010
Region A 49 59 64 67
Region B 36 31 27 25
Region C 15 11 9 8
Table 2. 7 Relative regional distribution of total trucks (%)
Years 1993 2000 2005 2010
Region A ~ 43 52 57 62
Region B 38 34 31 27
Region C 19 14 12 11
16 VTI rapport 413A
2.3.2 HDT percentages of total trucks
The data is from 1992 where the average percentage of HDT s to total trucks is nationally 8.6%. The regional percentages are: Region A 9.6%, Region B 7.6% and Region C 6.6%. In the future, these percentages are not expected to change greatly. This conclusion is based on the forecast relationships between HDT s and total trucks.
The provinces Where the HDT percentage is over 11% are Beijing, Henan and Shandong. See also Chart 4 on page 35.
2.3.3 Number of HDT s by province See Chart 3 on page 35 .
About 50% of the heavy duty truck population is located in the six provinces Guangdong, Shandong, Liaoning, Henan, Beijing and Hebei. Apart from Henan and Hebei, four of them are within region A.
3 Data sources
3.1 Data from China Statistical Yearbook
In China, the State Statistical Bureau is responsible for collection and publication of statistical data. The China Statistical Yearbook is an annual publication edited
by the State Statistical Bureau, which contains very comprehensive data on
na-tional and local levels by province, autonomous region, and city directly under central government and therefore re ects various aspects of China s social and economic development.
The major data sources in this publication are annual statistical reports. Some data have been taken from sampling surveys. There are special laws for statistical reports and surveys. The data in this publication are considered to be reliable.
Most of the data used in this report are from the China Statistical Yearbook (several yearly publications). It is the main source of data for this project and is always usedunless otherwise stated.
3.2 Data from Chinese National Heavy Duty Truck Corp.
(CNHDTC)
The HDT data is from CNHDTC, which is one of 55 experimental enterprises in China and is independently listed by the State Planning Committee, which has comprehensive functions such as manufacturing, marketing, research and deve-lopment, personnel training, finance and foreign trade. CNHDTC has a data base on heavy duty trucks. The author collected data in China at the end of April 1995.
Regional HDT data were obtained for one source year.
3.3 Data from Motor World Vehicles Data
World Motor Vehicles Data is published annually by the American Automobile Manufacturers Association. The chapter about China contains data on total re-gistrations, production, sales and exports/imports. Since it is often necessary to use different data sources from year to year, there are occasional differences in the yearly numbers reported by different publications.
4 Forecast methods
The modeling approach used for this forecast is restricted by the limited data
available. See Chart 1 (page 33) and Chart 2 (page 34), Which show the forecast
methods and results. ' '
There are three main steps in the forecast approach. The first step gives a fore-cast for the total number of trucks in China at the end of each forefore-cast year. Based on the result of the first step, the second focuses on the forecast for the numbers of heavy duty trucks. The third step gives a forecast for the demand for heavy duty trucks in the forecast years.
4.1 Forecast for the total number of trucks
The first step of the forecast is to forecast the total number of trucks. This is shown in the left part of Chart 1. In this step six different forecast methods are used. The forecast methods can be divided into two groups. The first group of
methods to forecasts the number of trucks directly and the second group forecasts
road freight transport demand and then estimates truck demand.
4.1.1 Direct truck forecast
The first group of methods forecasts the number of trucks directly. Three methods are used. They are cross section, time series and longitudinal data regression studies.
Cross section study -- A conclusion from the analysis of the relationship between truck demand and influential factors is that the key factor in uencing truck demand is economic growth as indicated by GDP. We use a cross section regression study to analyze the relationship between GDP and number of trucks at total national and regional group levels. Based on the GDP growth forecast by region, we then forecast the truck population of each region, aggregation of re gional forecast results gives the total national result. For a full view of estimation methods and results, we refer to Charts 1 and 2 (pages 33 34) and the Charts 16 to
19 (pages 41 42).
The forecast formulas are:
Region A Y1 = 1.07+0.01252GDPi (Chart 17 page 42, Appendix)
Region B Y1 = 3.61+0.013603GDPi (Chart 18 page 42, Appendix)
Region C Y1: 1.04+0.019441GDPi (Chart 19 page 42, Appendix)
where Yi : The population of trucks in province i (10 000 vehicles)
GDPi: GDP of province i (10 000 RMB yuan)
More detailed information about the regressions is given in the Appendix. The estimates of economic growth rates by province are shown in Section 2.3.3 on page 17. The forecast is based on economic growth rate analysis at total
na-tional and provincial levels. For growth rate values, we refer firstly to China s
economic and social development plan, both the medium term 5 year plan and the long term 15 year plan. In the (9th) 5 year plan for 1995 2000, the expected eco nomic growth rate is 8 9%. In the long term plan, the aim for economic growth
from 2000 to 2010 is to double GNP. Secondly, we refer to the results of a Chi-nese macro economic simulation model. The reference is (1) A composite model
of macro economic functioning of China . The purpose of the model is to show the quantitative relationships between economic growth, in ation and other basic
macro economic indicators. Thirdly, reference is also made to estimates of
China s economic growth rate as published by economists in Western countries.
The basic conclusion from the estimates is that the Chinese economy will continue to expands. More detailed information can be found in Reference (2) The future
of China s economic reforms , (3) China s business cycles , (4) China s automotive industry and market: Powerhouse of the 2lst century etc. Fourthly,
the estimated economic growth was also dependent on the time series data ana-lysis of both GDP index and annual growth rate. For GDP index trend anaana-lysis, see Chart 13 (page 40) and Appendix.
Based on the estimated economic growth rates and the forecast formulas shown above, we obtain forecast results at provincial level. The following are the forecast
result for Yi by region:
Region A
year 2000;
415
10 000 vehicles
- 2005: 661
2010: 969 _
Region B year 2000: 268 10 000 vehicles
2005;
347
2010: 443
Region C , year 2000: 117 10 000 vehicles
" 2005: p 140
2010: 160
National total r) year
2000:
800
10 000 vehicles
' 2005:
1 148
2010: 1 572
Time series study - The Chinese national truck data from 1965 to 1993 and
polynomial trend lines are used to forecast future total trucks. These are referenced in Charts 5 6 (page 36). The residual outputs of trend lines are shown in the
Ap-pendix. The fOrecast was based on the data (1978 1993) from the China
Statisti-cal Yearbook. ' ' a '
The forecast formula is: y = 1.099076 x2 4338.411802x + 4281360695
R2 = 0.995914
-where y: Chinese total truck population at the end of each year
x: years 1979, 1980
The forecast total truck population at the end of year:
year 2000: 841 10 000 vehicles
2005: 1 158
2010: 1 530
Longitudinal data regression study - The regression study between GDP and the total truck population was based on 1978 1993 data. See Chart 7 (page 37) and here it can be seen that GDP and the number of trucks have a very close
relationship. The forecast for truck demand is based on the economic growth rate estimates.
The forecast formula is: y = 1.4705x 46.963 R2 = 0.9958
where: y: Chinese total truck population at the end of each year
x: GDP index 1978:100
More detailed information on the regression is shown in the Appendix.
The estimates of economic growth rates are: 1994 2000 9%
2000 2005 7%
2005 2010 6%
See 2.3.3 on page 17.
The forecast total truck population at the end of year:
2000: 835 10 000 vehicles
2005: 1 144
2010: 1 504
4.1.2 Forecast based on road freight transport demand
The second group of methods forecasts road freight transport demand and then estimates truck demand. Three methods are used to forecast road freight turnover
(ton km). They are time series analysis, longitudinal data regression and elasticity
analysis. *
Time series study --- The first method is designed to find the trend line of road freight turnover volumes. See Chart 8 (page 37) and Chart 9 (page 38). The
data used to estimate the trend line are from 1949 to 1993. There are two lines in Chart 8: one is total ton km statistics and the other is ton km of transport agen
cies since 1979 (from 1979, total ton km statistics include non market driven
internal transport services. From 1984, the statistics also include quantities trans-ported by private freight trucks). The forecast is based on the total statistics for ton km using a polynomial trend line calculation.
The forecast formula is: y = 2.487915 x2 - 9613.026382 x+ 9280836609
R2 = 0.9617 (Chart 9, Residual output on page 38)
where y: Road turnover volume (100 000 000 ton km) for each year
x: Years 1979, 1980,...
The forecast result for Y for the years:
2000: 7 191 100 000 000 ton km
2005: 10 441
2010: 15 334
Longitudinal data regression study - using regression analysis to formulate
the relationship between GDP and road turnover volumes based on the data bet-ween 1980 1993 (Reference Chart 10). More detailed information on the regres-sion output is given on page 38.
The forecast formula is: y = 13.784x 669.9 R2 = 0.9339
where: y: Road turnover volume (100 000 000 ton km) for each year
x: GDP index 1978:100
The estimates of economic growth rates are: 1994 2000 9%
2000 2005 7%
2005 2010 6%
The forecast result for road freight turnover volumes, Y, for the years:
year 2000: 7 604 100 000 000 ton km
2005: 10 489
2010: 13 884
Other Models - There are various other methods which are used for transport
demand forecasts in China. Here we use an additional method, elasticity analysis,
to forecast the road freight transport turnover volume. Based on the data for 1983 1993
A V / V 11.4%
E = = = 1.036
A GDP / GDP 11.0%
Where V: Volume of road freight turnover
GDP: GDP index 1978:100
The elasticity, E, can be used to estimate the rate of increase in road freight turnover volume. The estimated forecast value of E is approximately equal to 1. Thus, it can be said that the annual rate of increase in road freight turnover volume
will be almost the same as the annual economic growth rate. The total volume of road freight turnover in 1993 was 4070.5 (100 000 000 ton km) (China Statistical
Yearbook 1994). By taking the estimate of national economic growth rate and
GDP index, see 2.3.3 page 17, the following forecast results are calculated:
2000: V: 4070.5(1+9%)7 = 7441 (100 000 000 ton km)
2005: V: 7441 (1+7%)5 = 10436 (100 000 000 ton km)
2000: V=10436 (1+6%)5 = 13966 (100 000 000 ton km)
where V: Volume of road freight turnover
Average forecasts can be estimated from the results of the three methods,
which yield approximately (see lower left, part of Chart 2):
Volume of road freight turnover: year 2000 7 500 (100 000 000 ton km)
2005 10 500
2010 14 000
The next step is to convert the road transport demand into truck demand.
A study of the average truck performance per year is undertaken. The annual
number of ton km per truck each year has not changed clearly through the years.
On average, this is about 88 000 ton km/truck for each year from 1984 to 1993. A forecast for average truck performance per year is given using the economic growth trend line analysis (see Chart 11, page 39).
Road transport is often at a primitive stage at the beginning of industrialisation, (this is discussed in more depth at the end of this report). Social trucks are ex pected to increase in the future. It is assumed that the average performance by the year 2000 will be a little lower than the trend line number given (90 000 compared to 91 000 ton km, see Chart 11). The estimated values for the years 2005 and 2010 are downward adjusted trend line forecast values.
The downward adjusted forecast values are:
2000 average performance 90 000 (ton km / truck each year)
2005 92 000
2010 94 500
It is not difficult to convert the road transport demand into truck demand using the annual average performance per truck. The converted truck population forecast results are:
year 2000: 8 330 000 vehicles
2005: 11 410 000
2010: 14 800 000
To validate the above result, a regression analysis on the population of trucks
and the volume of road freight turnover is undertaken (see Chart 12).
By using the methods described in method groups one and two, fourforecast results for the number of trucks are produced. The average results from the dif-ferent methods are approximately (see Chart 2):
year 2000: 8 300 000 vehicles
2005: 11 500 000
2010: 15 300 000
This forecast result is then disaggregated into regional groups.
Table 4.] Regional fo'recastsfor the truck population
(10 000 vehicles).
Forecast year
20.00
2005
2010
Region A 430 660 950 Region B 280 350 420 Region C 120 140 160 Total 830 1 150 1 5304.2 Forecasting the number of HDT s
This is the second step of the forecast. The purpose is to forecast the population of heavy duty trucks. Two different methods are used in this step and the results are
very similar (see the central part of Chart 1).
4.2.1 Relationship between the number of trucks and total truck ca-pacities
The total truck capacity is the sum of the maximum payloads of all trucks. Changes in total truck capacity in a region depend on two factors: one is the total
number of trucks, the other is the proportion of different truck duty types light,
medium and heavy.
Two conclusions about the change in average capacity per truck in China can be drawn. The conclusions are based on the change in the total truck numbers and their total capacities and also on the regression analysis (see Chart 14, page 40 and
Chart 15, page 41). The conclusions are:
. a. There is a close relation between the number and capacity of trucks in China. This means that the increase in the total truck capacity has been mainly due to the increasing number of trucks in recent years. This is an occurrence which will probably continue in the future.
b. The relationship between the number of trucks and their total capacities can
be formulated as follows (see Chart 20 22, page 43):.
Region A: N: 2.07 +0.2711C (Chart 20 page 43, Appendix)
Region B: N: 2.256+0.205 C (Chart 21 page 43, Appendix)
Region C: N: 0.548+0.211 C (Chart 22 page 43, Appendix)
where N - number of trucks in the region.
C -- truck capacities in the region.
More detailed regression information is shown in the Appendix. In this case, the intercepts have no economic meaning.
The study suggests that the proportion of heavy duty trucks in recent years could be used in the forecast of heavy duty truck demand as a basic proportion. According to the forecast results, a slight increase in this proportion may occur in the future.
4.2.2 Forecast at total national level
The first method of this step is to estimate directly the percentage of heavy duty trucks to total trucks at national level.
Data analysis shows that the proportion of heavy duty trucks to total trucks de-pends heavily on one indicator, GDP per truck. The higher this indicator, the larger the proportion of heavy duty trucks. This is indicated by the data for 1992:
Table 4.2 1992 GDP, trucks, GDP/truck and HDT's/all trucks.
GDP (yuan) Trucks GDP per truck HDT 's /all
(100 000 000) (10 000 vehicles) (10 000 yuan) trucks
(%)
Region A 11837 182.9948 64.68 9.6
Region B 8667 171.7824 50.46 7.6
Region C 3449 86.6722 39.79 6.6
Nation 24363 441.45 55.19 8.6
Based on economic growth rates in the forecast years and forecast values for the number of trucks, and also using the 1992 price level forcalculation of GDP index, values of GDP per truck can be estimated.
Table 4.3 F0recast values for variables in the previous table.
Year GDP (yuan) Trucks GDP per truck HDT/ all
(100 000 000)
(10 000 vehicles)
(10 000 yuan)
trucks
(W
2000 ' 50 000 , 830 60.24 9.3
2005 73 000 1150 63.48 9.9
2010 100 000 1530 65.35 10.3
The calculation of HDT / all trucks formula:
P1992 * Gforecast year
forecast year HDT / All trucks = (42 1)
G1992
where: P1992-- HDT / all trucks by data 1992
G1992 --- GDP per truck by data 1992
Gforecast year - - GDP per truck, forecast years
e. g.
year 2005: HDT / all trucks = (8.6%*63.48)/55.19=9.9%
Based on the above calculation method, the analysis of Chapter 4.2.1 and the forecast by experts at the Chinese National Heavy Duty Truck Corp., estimated percentages of heavy duty trucks in relation to total trucks are:
year 2000 9%, year 2005 10%, year 2010 11%
The forecast numbers of future heavy duty trucks are:
year 2000 750 000
2005 1 150 000
2010 1 680 000
4.2.3 Forecast based on a regional level study
The second forecast method is based on a regional level study. The proportion of heavy duty trucks to total trucks at regional group level using the same method as above is then estimated.
Based on the economic growth rate estimates at regional level (See Chapter 2.3.3, page 17) and the forecast results of the number of trucks (See Chapter 4.1.2, page 21), the formula 4.2.1 (page 23) to calculate the proportion of heavy duty trucks can be used. The results are shown in the following tables:
Table 4.4 Estimates ofGDP / truck and HDTproportions in Region A.
Year GDP per truck HDT/ all tracks
(10 000 yuan) (%)
1992 64.68 9.6
2000 69.77 10.4
2005 74.24 11.0
2010 77.37 11.5
Table 4.5 Estimates of GDP / truck and HDTproportions in Region B.
Year GDPper track HDT/ all tracks
(10 000yuan) (%)
1992 50.46 7.6
2000 53.57 8.1
2005 58.57 8.8
2010 63.10 9.5
Table 4.6 Estimates of GDP / truck and HDTproportions in Region C.
Year GDP per truck HDT /All trucks
(10 000 yuan) (%)
1992 39.79 6.6
2000 41.67 ' 6.9
2005 43.57 ' 7.3
2010 46.88 ' 7.8
According to the above calculations, the estimated forecast values for the per-centage of heavy duty trucks in relation to total trucks in each of the regional groups are:
Year 2000 2005 2010
Region A 10.4% 11% 11.5%
Region B 8.1% 8.8% 9.5%
Region C 6.9% 7.3% 7.8%
Based on these proportions of heavy duty trucks, forecast values for the number of heavy duty trucks are:
Table 4. 7 Forecast resultsfor the number ofHDT's (10 000 vehicles). Year 2000 2005 201 0 Region A 44.72 72.60 109.25 Region B ~ 22.68 30.80 39.90 Region C 8.28 10.22 12.48 National 75.68 113.62 161.63
This method gives us the following national forecast values for future HDT numbers adjusted upwards:
year 2000 757 000 vehicles
2005 1 140 000
2010 1 620 000
4.2.4 HDT forecast result
Based on the forecast from the above two methods, an average result can also be
presented (see upper right part of Chart 2 on page 34). Forecast numbers of heavy duty trucks:
year 2000 750 000 vehicles
2005 1 150 000
2010 1 650 000 f
4.3 Forecast foryearly HDT demand
This is the third step of the forecast. Here, a forecast of the demand for heavy duty
trucks per year is given. The results are supported by different forecast methods (see the right part of Chart 1 on page 33).
4.3.1 Forecast based on the number of HDT s
The first method is a forecast for heavy duty trucks based on the assumption that the average life expectancy of a heavy duty truck is 12 years. This assumption is based on the analysis of Chinese truck use data.
It is known that the total sales of heavy duty trucks were 16 000 vehicles in the year 1981 and 52 000 vehicles in 1993. Using an average rate of increase 10.3%, sales of heavy duty trucks per year during 1982 1992 can be estimated. According to the average life expectancy of heavy duty trucks, vehicles bought before 1988
will on average be scrapped before the year 2000 and the trucks sold during the years 1989 to 1993 will on average still be used in the year 2000 and will on aver-age be scrapped before the year 2005. In the year 2010, heavy duty trucks bought
before 1998 will on average be scrapped.
In step 2, an estimated result for heavy duty trucks at the end of each forecast year was presented:
year 2000 750 000 vehicles
2005 1 150 000 vehicles
2010 1 650 000 vehicles
From these figures, the total demand for heavy duty trucks from the year 1994 to the years 2000, 2005 and 2010 can thus be calculated:
Years 1994 2000, total demand for trucks:
= number of trucks at the end of 2000 truck sales from 1989 to 1993 = 750 000 250 000 = 500 000 (vehicles)
Based on the average growth rate, the forecast demand in each year will be:
year 1994 1995 1996 1997 1998 1999 2000 1994 2000
demand 56 61 65 71 76 82 89 500
(1 000 vehicles)
Years 1994 2005, total demand for trucks
= number of trucks in use at the end of 2005
= 1 150 000 (vehicles)
Years 1994 2010, total demand for trucks
2 number of trucks at the end of 2010 + truck sales from 1994 1998 = 165 000+33 000
= 198 000 (vehicles)
To sum up, years 1994 2000 500 000 vehicles
1994 2005 1 150 000 vehicles
1994 2010 1 980 000 vehicles
Based on the average truck growth rate, it is not difficult to give a forecast of the demand for heavy duty trucks per year. This is given by the average growth rates from the following equations: where x is the average growth rate of HDT
demand. year 1994 2000 5.2*(1+x)+5.2*(1+x)2 + 5.2*(1+x)3 + ...+ 5.2*(1+x)7 = 50 result: x = 0.079447 year 1994 2005 5.2*(1+x)+5.2>*<(1+x)2 + 5.2*(1+x)3 + ...+ 5.2*(1+x)12 = 115 result: x = 0.091067 year 1994 2010 5.2*(1+x)+5.2*(1+x)2 + 5.2*(1+x)3 + ...+ 5.2*(1+x)17 = 198 result: X = 0.084358
The above average growth rates are then used to estimate the demand in each of the forecast years.
The forecast result for heavy duty truck demand per year is:
year 2000 89 000 vehicles 2005 148 000 vehicles
2010 206 000 vehicles
4.3.2 Forecast based on the growth in HDT sales
The second method is based on the analysis of sales volume and population growth of heavy duty trucks. In 1981, the total number of HDT s sold was 16 000 vehicles and in 1993 52 000 vehicles. The average growth rate was 10.3%. During
the same time period, the GDP average annual growth rate was about 9.3%. Based on the data between 1981 1993 the elasticity, E, is:
A SH/SH 10.3%
E: __________________ -_ : --=
AGDP/GDP 9.3%
where SH: total sales of HDT
GDP: GDP index year 1978:100
Based on the analysis of economic growth rates, the growth rate of HDT de-mand using E is estimated.
Table 4.8 The estimated HDT demand average growth rates.
Time period 1994 2000 2000 2005 2005 201 0
Annual economic growth rate 9% 7% 6%
Annual HDT demand growth rate 10% 8% 7%
These annual growth rates are then used to forecast the demand in the forecast
years. The forecast result for HDT demand for each forecast year is:
year 2000 101 000 vehicles
2005 149 000 vehicles
2010 209 000 vehicles
The method used in Section 4.3.1 gave the forecast result:
year 2000 89 000 vehicles
2005 148 000 vehicles 2010 206 000 vehicles
Based on the analysis of economic development, transport structure and the
model calculation, we sum these two results and give a final, rounded forecast
result for HDT demand:
year 2000 95 000 vehicles / year
2005 150 000 vehicles / year
2010 210 000 vehicles / year
The design of this forecast approach depends on the available data. All forecast numbers are based on the economic analysis and model calculation. Chart 2 on page 34 shows an overview of the total forecast approach and the results of each method and step.
4.4 General comment on the forecast method and results
4.4.1 The stage of China s transportation development
As previously mentioned, a good agreement in the results of the different forecast methods used in this report was obtained. However, the total forecast process re-lies on previous data and depends on the assumption that the future situation does not drastically differ from earlier developments and that the relationship between truck demand and the key in uencing factors remains stable.
According to the theory of transportation (Reference 9), transportation is an important feature of industrialization. There is also an economic process taking place alongside industrialization. In the process of transportation, the scale of spa-tial movements of humans and cargo has greatly expanded as a result of the use of recent and modern means of transport. Transportation has become a very
impor-tant basic industry, infrastructure and environmental condition, on which the eco
nomy depends for entering into modern expansion. The characteristics of indus-trialization are such that people previously to put more emphasis on specialization,
mechanization, large scale production, electrification and research and de
velopment. However later, in relation to industrialization, it can be seen that trans portation has the same importance as several other characteristics. Transportation and industrialization accompany each other in their development. There can be no industrialization without transportation.
Based on one ask Bertil theory of transportation, economic development can
be divided into three stages: pre transportation, transportation and post transpor-tation; of these, the stage of transportation can be subdivided into two quasi stages of primary transportation and perfect transportization. See the following figure:
Table 4.9 Different stages in transportation.
Pre transportation Transportation Post transportation
/ \
Primary transportation Perfect transportation
In the different stages, economic development, the demand for road transport and the relationship between road transport demand and economic factors are dif ferent. Otherwise, in the same stage, they remain stable.
In the foreseeable few decades of economic development, sharp contradictions between supply and demand in both passenger and freight transport are likely to last for a long time. Transportation in China will remain in the primary stage during the first decade of the next century. A feature of this stage is rapid expan sion in transport capacity. The details of this development can be found in
Refe-rence (9). Based on the estimated results of this report, it can be said that the fu
ture will not differ greatly from earlier development of road transportation and truck demand.
The estimate of economic growth in China is a major factor used in this fore-cast. The economic reform that begun in 1978 has been successful so far. The
market economy will replace the plan economy. After 17 years of reform and con tinued growth, the Chinese economy will increase steadily and is expected to con-tinue to do so in the future. This is a common prediction of most economists both in China and other parts of the world.
4.4.2 Discussion on validity and reliability of data and forecast
method in this study
Most of the analyses used in this study are based on regression models. Estimating
regression models creates residuals between known observations and the fitted line. There are several reasons for these differences. They include:
1) Incorrect measures due to difficulties in accurately measuring the variable in question. Measurement errors in the data can be both systematic and/or random. A systematic measurement error causing no extra variation will not show up in the regression tests. Therefore these errors can be very difficult to detect. There are no good methods when one is restricted to secondary data. When it comes to random measurement errors in the data used in this study, the generally small variations
can be interpreted as small measurement errors an or indication that data has been
adjusted to fit a perfect line, which is not a very reasonable supposition. The author has found no reason to believe that the quality of data used in this study is
questionable.
2) Real variation , i.e. differences due to actual behavior of the economy and its agents.
3) Incorrect model specification, i.e. omitting a relevant variable, including an unnecessary variable or the mathematical specification of relations between
in-cluded variables is wrong. There are several econometric tests which can be undertaken to detect these types of mistake . In this study, only one explanatory
variable is used for each regression. The presented F values together with high R2 values, indicate that the chosen variables are relevant as explanatory variables and the inclusion of further variables would not improve the predictive power very greatly.
The inclusion of unnecessary variables is tested with the usual t test. In some regression equations, the intercepts are not significant. The reason for not ex cluding the intercepts in such cases is that the goodness of fit is improved when intercepts are included and that the intercepts do not affect the forecast values by a great magnitude. Here it is regarded that the goodness of fit will improve the pre
dictive power of the model, which is preferred to other regression model purposes,
for instance, to give a validation/rejection of a theoretical hypothesis from a par-ticular regression model.
The D W d test is a common way of testing for autocorrelation in the residuals
and drawing conclusions about the chosen functional form of the model. In some cases D W d tests on regressions performed in this report show, signs of auto-correlation and in others they are in the zone of indecision (Gujarati, 1988, p.
377). However, the D W d-tests are disregarded since the number of observations
is small for all regressions and if an outlying observation is removed, then the
chosen functional forms would fit the data very well. On the other hand, observa-tions have not been excluded in any case since the number is small.
Since the purpose of this study is purely to make a forecast, the main criteria has been its predictive power. A high R2 can in some sense be regarded as a good
predictive power. High R2 values givegood predictions of observed data. In this
report, forecasts are undertaken by trend extrapolation and there is always a risk
with that method, since if the trend is broken for some reason, the forecast will be
of little value. When deciding on trend lines in this study, high R2 values have
been the main decision rule. This may result in functional forms which are not relevant for an infinite time horizon, but hopefullywithin certain time limits. It is
the author s belief that those limits have not been exceeded in the work reported
here.
5 Chart list
List of charts: Chart Chart 1 1. Chart 12. Chart 13. Chart 14. Chart 15. Chart 16. Chart 17. Chart 18. Chart 19. Chart20. Chart21. Chart22. 32 1. Forecast approach2. Detail approach of HDT demand forecast 3. Population of HDT by region
4. Percentage of HDT (HDT / All trucks by regions) 5. Total truck population trend in China (1965 1991) Chart 6.
7 8 9 0
Truck population trend in China
. Relationship between the truck population and GDP . Chinese turnover volume of road freight traffic 1949 1993
. Trend of road freight turnover volume (1979 1993)
Chartl . Regression between population of trucks and total freight turnover
volume (ton km) 1984 1993
Ton km per truck per year (total trucks average)
Regression between GDP and road turnover volume
GDP trend
Change in truck number and total capacity
Change in average capacity per truck in China
Relationship between trucks and GDP by region (1993) Relationship between trucks and GDP in region A ( 1993) Relationship between trucks and GDP in region B ( 1993)
Relationship between trucks and GDP in region C (1993)
Relationship between the number and capacities of trucks in region A (1993)
Relationship between the number and capacities of trucks in region B (1993)
Relationship between the number and capacities of trucks in region C (1993)
Chart 1 _C ro ss -s ec ti on st ud y, 30 re gi on s, I 3 re gi on gr oup s,
fo
re
Ca
st
re
sp
eC
ti
ve
ly,
Fo re ca st on re gi on le ve lVTI rapport 413A
Forecast approach. Fi rs t 19 93 tr uc k us e da ta me th od H D T pe rc en -: ta ge of to tal group . . § tr uc ks , givi ng T lm e-se rles st ud y 5 rise to H D T Fo reca st 19 65 1 99 3 truc k us e da ta Th e end fo reca st Th e en d di re ct ly of fo reca st -va lue s offo re ca st H DT
Ve
rt
ic
al
da
ta
re
gr
es
si
On
.I
_
pe
r
ye
ar
'
st ud y T r uc k H D T G D P an d tr uck num be r use in in us etot
al.
HD
T
We
en
-to
ta
l
ta ge so f to ta l Ti me -s er ie s tr uc ks on st ud y R d : re gi on al le ve l, ' 3 50 3 " A : « _-' . Se co nd to nk m fo re ca st -' ._ _ . St ud y of ; ,g l- V-ln 'g II SC to . _ me th od fr el gh t to n-km / g H D T fo re ca st H D T St hY Of m u l R . tr an sp or t t g k -3 va lue s pr od uc tl on ,g
p
$
65
15
50
de
ma
nd
pe
r
we
5
im
po
rt
/e
xp
or
t,
ve mc a at a fo re ca st pe r ye ar vo lum e of sa le Fo re ca st to n-km fo re ca st to ta l tr ucksin
di
re
ct
]
an
d
ot
he
r
y ave ra ge in ue nc in g to n-km Ot he r mo de l 33 fa ct or s st ud y to n-km fo re ca st34 VTI rapport 413A
Chart 2 Detail approach ofHDT demandforecast.
Tr uck in us e Cr os s-se ctio n st ud y, Aa. 41 5 30 regi on s, b. 66 1 3re gi on gr oup, c. 969 fore ca st re sp ecti ve ly, -B a.26 8 19 93 tr uck us e da ta b.34 7 c. 44 3 C a. 11 7 b. 14 0 Tr uc ks in use a. 80 0 Ti me -s er ie s st ud y C. 16 0 b. 11 48 c. 15 72 Tr uc ks in us e a. 841 19 65-1 99 3 tr uc k us e da ta Lo ng it ud in alda ta re gr es si on b. 11 58 ,c .1 53 0 Fi na l re sul ts Tr uc ks in use Tr uc ks _i n use G D P an d truc k num be r To n-km Ti me-s er ie s a. 71 91 st ud y b. 10 44 1 to n-km fo re ca st c. 15 334 To nk m
\
a. 76 04 b. 10 48 9 Re gres si on lo ng itud in al da ta To n-km a. 75 00 b. 10 50 0 b. 14 00 0 a. 83 5 b. 11 44 c. 15 04 a. 83 0 .b .1 15 0 c. 15 30 Tr uc ks in us e to n-km fo re ca st c. 13 88 4 To n-km a. 74 47 b. 10 43 6 Ot he r mo de l stud y to n-km fore ca st 0. 13 96 6 a.83 3 b. 1141 c.14 80 St ud y of to n-km pe r tr uc k pe r year to ta l tr uc ks ave ra ge Fo re ca st ye ars a 2000 b 20 05 c --20 10 A Re gi on A B Re gi on B C -Re gion C H D T pe rc en -tr ucks , gi vi ng H D T in us e a. 75 b. 11 5 c. 168 H D T pe rc en -H D T in us e re gi on al le ve l, a. 74 gi vi ng ri se to H D T fo reca st c. 16 6 b. 11 4 H D Tst ud y of pr od ucti on , im port /e xp or t volum e of sa le an d ot he r in ue nc ing fact or s Un it of fo re ca stre sul ts tr uc k an d HD T: 10 00 0 vo lum e of ro ad 10 0 00 000 0t 0n -k m Re sul t of H D T in us e a. 75 b.11 5 c. 16 7 HD T ve hi cl es fr ei gh t de ma nd pe r ye ar a. 8. 9 H D T dema nd pe r ye ar b. 14.8 c. 20.9 a. 9. 5 b. 15 c. 21 H D T de ma nd pe r ye ar a. 10 .1 b. 14 .9 c. 20 .9VTI rapport 413A 35
Chart 4 Percentage ofHDT (HDT /All tracks by regions).
glon Be mn g' i He na n Sh an do ng Gua ng Ua on mg 'a nn g a Qi ng ha i 55;: Gua ng do ng Sh an gh ai Fua n Sh an an xi XM a ng *s He be i An hui Zh ea ng Si ch ua n In ne r Mo ng ol ia s'féi a ng su Sh an xi Ti be t Hub ei a ng Yun na n Gui zh ou Hub ei linJi Ha in an Ga ns u He il on gj ia ng Ni ng xi a ;_ :;
HDT/All Trucks
Percentage of HDT (HDTI All Trucks) 1992
Chart 3 Population ofHDT by regzon.
Gua ng do ng Sh an do ng Ua on mg He na n Be mn g He bm sm hua n a ng su Sh an a M m g Zh ea ng He o ng a ng : An hui Gua ng k a a n g Sh an gh ai Hub ei Hun an Sh an an Fuan Y un n a n In ne r Mo ng ol ia ,1 925; : Jil in ma ng Guuh ou Ga ns u Qi ng ha i Ha ma n n g xm Tibet F)LA $3LA F3LL F3 01 I - 0870.87 0.87. 0.87» 0.76 0.76 0.76 0.99 0.99 0.99 0.99 0.99 H D T (1 0 00 0 ve hi cl es )
3761176 2742174 Eai jT Population of HDT by region 1992
Trend in total trucks in use in China Data from Motor World Vehicles Data (1965-1991)
14000000 2 1 E : a" y = 7020.981879x2 - 27618822809570xg+ 27180710270000006 12000000 v R3v=0,992931 § if -i ,l 5,1 ; .4 10000000 --- ; if g I i f . t {4' 8000000 i
g xx 3 [ Total trucks in use
0 3 1" i EO iE f,I - - - -Po|y Trendline > g .1 6000000 If" 4000000 -2000000 . 0 i J; 1 1 e 1985 1970 1975 1980 1985 1990 1995 2000 2005 2010 Year
Chart 5 Total truck population trend in China (1965-1991).
Trend in the number of trucks in China 1800 y = 1,099075x.2 - 4338,411802); + 4281380695313 1400 W = 09959134 1200 2- 1,, . . . 1. 4. 131.7. a ? 8 1000 ",v S - r ' 7 x .g' E 800 . ' z 3 5 Trucks 0 ' , E ; ,." '0 a . . g 600 - - . -- g - - - . - -- Trendhne Z ' , ' 400 y; / 0 i 1 1975 1980 1985 1990 1995 2000 2005 2010 Year
Chart 6 Truck population trend in China.
Relationship between the truck population and GDP indices by data 1978-1993 500 ~ 400-'
y = 1,4705x - 46.963 R2 = 0,9958 Serie1 - - - -Trendline 300 ZOO--Tr uc k po pul at io n 19 78 -1 99 3 (1 0 00 0 ve hi cl es ) 100-0 l i l . . 100 150 200 250 300 350 400 GDP indices 1978-1993 (1978=1 00)
Chart 7 Relationship between the truck population and GDP.
Turnover volume of road freight traffic in China (100 000 000 ton-km) 4500 4000 -3500 P 3000 2500
-Ton-km total statistics Ton-km transport agencies 2000 To n-km (1 00 00 00 00 ) 1500 ~ 1000 -
l
O > 19 60 -19 49 19 50 -19 51 19 52 > 19 53 1954 1955 1956 1957 1958 1959 ~ 19 61 ~ 19 62 ~ 19 63 19 64 -1965 r 19 66 ~ 19 67 19 68 -19 69 -19 70 -19 71 19 72 -19 73 ~ 19 74 -19 75 19 76 -19 77 ~ 19 78 -19 79 -19 80 -19 81 -19 82 19 83 -19 84 19 85 -19 86 -19 87 -19 88 -19 89 -1990 -19 91 > 19 92 -19 93 F 19 94 r 19 95 YearChart 8 Chinese turnover volume ofroadfreight tra ic 1949-1993.
Trend in road turnover volume of freight (total) 7000 y = 2,467915x2 - 9613,026382x {9260836609375 R2 = 0,961686 6000 5 ; 5000 . .. . 4000 5 I
Ton-km total statistics
__ ... . .. t .. . t .. .. A. To n-km (1 0, 00 0, 00 0) - - - Trendline 2000 ,. ' 1000 ,.
1975 1980 1985 1990 1995 2000 Year
Chart 9 Trend ofroadfreight turnover volume (1979-1993).
Regression between population of trucks and volume of road freight turnover
(1984-1993) y = 0,108978x + 12.93642 R2 = 0.920189 0 0 Y Q Radicde po pul at ro n 01 tr uc ks (1 00 00 ve hi cl es ) O I '5. § 6'}. . . . . -. - Linear regression O 1000 2000 3000 4000 5000 60m 7000 8000 Ton-km(100 000 000)
Chart 10 Regression between population of trucks and totalfreight turnover volume (ton-km) 1984-1993).
Ton-kms each truck per year (total trucks average) 120000 y = 291 ,36X - 491361 100000 80000 -E x a 60000 ~-0 I.
Ton-km per truck per year(tota| average) 40000 -- -- -- -- ----TTend ne 20000 -~ 0 T 1 1 1 1 1984 1986 1988 1990 1992 1994 1996 Year
Chart 11 Ton-km per truckper year (total trucks average).
Regression between GDP and road turnover (ton-km) 1980-1993 80 0O y = 13,784x - 669,9 7000 ~ R2 = 0,9339 7.3 6000
~-3 a
E cc: 5000 0 Y '1' § 4000 __ I Predicted Y o o Linear regression E; 52 3000-S V
,_ 2000 -_ 1000 -0 : : : : : 0 100 200 300 400 500 600 GDP indices by years 1980-1993 (1978:100)Chart 12 Regression between GDP ana1 roaa1 turnover volume.
1206 Trend in GDP 1000-_mw.wmn.WHWWL.MmmmMHmedeHMW nmm.h .m y = 0.716121x2 - 2826,075968x +27§88267,085938 R? = 0.989687 x 800 600 G D P in di ce s (1 97 8: 10 0) 400 u ...Q ...u. .« GDP indices .' - - - Trendline 200
1975 1 980 1 985 1990 1995 2000 2005 2010 Year Chart 13 Trend in GDP.
Change in number'and total capacities of trucks
240 220 ~ Num be r of tr uc ksin di ce s To ta l ca pa ci ty in di ce s (1 98 5: 10 11 ) _a Ao l 120 -m o O l _A (I)O I _A O) O I 1
The number indices of Trucks Total capacity indices of trucks
L
100
1985 1986 1987 1 988 1989 1990 1991 1992 1993 1994 1995
Year
Chart 14 Change in number and total capacities 0ftrucks.
Ca pa ci ty pe r tr uc k (t on ) 5.5 4,5 -3.5
--Change in average capacity per truck in China
Average capacity per truck
/\/\/
1985 1989 1990 1991 1992
Year
1986 1987 1988 1993
Chart 15 Change in average capacity per track in China.
Relationship between the number of trucks and GDP by regions
(1993)
60 E 50 -~ 0 SE
.2 o Y 51 2 Predicted Y O M .E 0 E l O is. . . : i 1 : : 0 500 1000 1500 2000 2500 3000 3500 GDP of 30 regions in 1993 (10 000 yuan)Chart 16 Relationship between trucks and GDP by regions (I993).
The relationship between number of trucks and GDP in region A (1993) 60 A50 k 0 3 8m g 40 -~ g - Y 9 IC e =3 .3 20 -E 5 o . o l- 510 ~-M , 0 Q . 1 : i 2 i 0 500 1000 1500 2000 2500 3000 3500
GI]3 of provinces (10 000 yuan) 1993
- Chart 1 7 Relationship between trucks and GDP in region A (1993).
The relationship between number of trucks and GDP in region B (1993) 40 . 35 -- O 30 I. .m" 25-» . . Y 20 -- . . 15 -- . * . Predicted Y 10 - . O Tr uc ks in us e by pr ovi nc es (1 0 00 0) 0 200 400 600 800 1 000 1200 1400 1600 1800 2000
GI? of provinces (10 000 yuan)1993
Chart 18 Relationship between trucks and GDP in region B (1993).
The relationship between number of trucks and GDP in region C (1993)
N 01 N O \ um-- Redicted Y 61 o L O i Tr uc ks In us e by pr ovi nc es (1 0 00 0) 01 00 O O 200 . 400 600 800 1000 1200
GIN3 of provinces (10 000 yuan) 1993
O
Chart 19 Relationship between trucks and GDP in region C (1993).
The relationship between the number and capacity of trucks in region A in 1993 60 > A50 -- 0
a s
as 2 40 --
E 1: 0 Y 3". w 30° 3 PredlctedY.saw
E 9. g 9.10 --0 . . . : l : . O 20 40 60 80 100 120 140 160The total capacities of trucks by provinces in region A (10 000ton)
Chart 20 Relationship between the number and capacities oftrucks in region
A (1993).
The relationship between the number and capacity oftrucks in region B in 1993 40 > A35 ~-0 o 25 E 1: ' Y .5 g f: " Predicted Y "' .E
éaw
3 Q 5 -0 . . : : : . : I O 20 4O 60 80 100 120 140 160 180The total capacities of trucks by provinces in regionB
(1o 000ton)
Chart 21 Relationship between the number and capacities oftrucks in region
B (1993).
The relationship between the number and capacity of trucks in region C in 1993 25 3 820 "'g, 0 J. x o E E15 " ° Y E g 10 u ... predicted Y
a a 5 w
0 . . I I . . . t . O 10 20 30 40 50 60 7O 80 90 100The total capacities by provinces in region C (10 000ton)
Chart 22 Relationship between the number and capacities of trucks in region
ang
6 References:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Shuxia Li, Xin Zhao and Xibao Zhang, A composite model of .macro-economic functioning of China, Computers & Industrial Engineering ISSN: 0360-8352 Vol: 27 Iss: 1-4 Date: Sept. 1994 P: 313-316.
Harrold Peter, The future of China s economic reforms, China Business
Re-view, Vol: 20 Iss: 4 Data: Jul/Aug 1993 p: 538-559.
Imai Hiroyuki, China s business cycles, China Business Review Vol: 21 Iss: 1 Data: Jan/Feb 1994 p: 14 16.
China s automotive industry and market: Powerhouse of the 2lst century, International Motor Business Data: first Quarter 1994 p: 50-71.
China s automotive sector: Prospects to 2000, International Motor Busi-ness, Data: first Quarter 1995 p: 52-81.
Lam Alexa, Roads, China Business Review, Vol: 21 Iss: 5 Data: Sep/ Oct
1994 p: 49.
Harwood Elizabeth, China: The dragon awakes, Banking World, Vol: 12 Iss: 2 Data Feb 1994 p: 38.
Xu Qingbing, Rong Chaohe and Ma Yun, Introduction to Transport Eco-nomics, China Railway Publisher, Beijing, 1995.
Rong Chao He, A theoretic analysis of the problem of quickening the pace of
economic transportization in China, Journal of China Railway Society, Vol: 3, p: 59 66, & Vol: 4, p: 44-51, 1991.
(10) Damodar N. Gujarati, Basic Econometrics, McGraw-Hill Book Company,
44
Singapore, 1988.
APPENDIX
RESIDUAL OUTPUT OF CHART 5
Predicted Y Residuals year 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 Total trucks 211069 227352 229086 246197 281112 371446 449379 564044 625793 711849 810765 903464 1070154 1266385 296704.9367 277861 .8935 273060.814 282301 .6983 305584.5464 342909.3582 394276.1338 459684.8731 539135.5762 632628.243 740162.8736 861739.468 997358.0261 1 147018.548
RESIDUAL OUTPUT OF CHART 6 Years Total trucks
1978 1980 1983 1984 1985 1986 1987 100.17 129.9 169.44 188.37 223.2 246.57 281.21 Predicted Y 99.415741 122.877753 174.556911 196.179601 220.000443 246.019437 274.236583
RESIDUAL OUTPUT OF CHART 9
Appendix Sid 1 (1 1) 85635.94 50509.89 43974.81 36104.7 24472.55 85635.94 -55102.9 -104359 86657.4 -79220.8 -70602.1 -41724.5 -72796 -1 19366 Residuals -0.75426 7.02225 5.1 1691 1 7.809601 -3.19956 -0.55056 -6.97342 year 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Total trucks 1371733 1479993 1629822 1767455 1907795 2069940 2373999 2813377 3127643 3539655 3899410 4171855 4349218 1310721 .03 148846548 16802519 188608027 210595061 233986292 258781719 284981342 3125851 .61 3415931 .77 37200539 403821798 437042403
(The trend of the number of trucks in China) Years Total trucks Predicted Y
1988 1989 1990 1991 1992 1993
Year Road ton-km Predicted Y Residuals Year 1979 745 429.709912 315.29 1987 1980 764 666.339015 -97.661 1988 1981 780 907.943948 127.9439 1989 1982 949 1154.524711 205.5247 1990 1983 1084 1406.081304 322.0813 1991 1984 1536 1662.613727 126.6137 1992 1985 1693 1924.12198 231.122 1993 1986 2117.99 2190.606063 72.61606
RESIDUAL OUTPUT OF CHART 13 (The trend of GDP) Year GDP indices Predicted Y Residuals Year
1978 100 100.775798 0.775798 1988 1980 1 16 117.437698 1.437698 1989 1983 145.5 153.172363 7.672363 1990 1984 166.9 167.948402 1.048402 1991 1985 188.2 184.156683 -4.04332 1992 1986 203.5 201 .797206 -1.70279 1993 1987 225.7 220.869971 483003
VTI rapport 413A
317.85 346.37 368.48 398.62 441.45 501 2660.39 3220.39 3374.8 3358.1 3428 3755.39 4070.5 251.2 262.1 272.7 295 334.5 378.7 304.651881 337.265331 372.076933 409.086687 448.294593 489.700651
(The trend of road turnover volume of freight) Road ton-km Predicted Y
2462.06598 2738.50172 3019.91329 3306.3007 3597.66393 3894.00299 4195.31788 241 .374978 263.312227 286.681718 31 1.483451 337.717426 365.383643
(The trend of total trucks in use of China data 1965 1991) Predicted Y Residuals ~61012 8472.483 50429.9 1186253 1981556 2699229 2138182 36436.42 -1791.39 ~123723 -179356 -133637 21206.03 Residuals -13.1981 -9.10467 3.596933 10.46669 6.844593 -1 1.2993 Residuals ~198.324 -481.888 -354.887 -51.7993 169.6639 138.613 124.8179
GDP indices Predicted Y Residuals -9.82502 1.212227 13.98172 16.48345 3.217426 -13.3164
SU M M A R Y O UT P U T of C HA R T 7 I (T he re la ti on sh ipbe twe en th epo pul at io n oftr uc ks an d GD P in di ce s by da ta 19 78 -199 3) Re gr es si onSt at is ti cs Mul ti pl e R 0. 99 79 1167 R Sq ua re 0. 9958 27 70 2 Ad jus ted R Sq ua re 0. 9954 48 40 2 St anda rd Er ro r 8. 31 02 27 87 8 Ob se rva ti on s 13 A N O V A df 88 M S F Re gr es si on 1 1813 12 .3 58 1 18 13 12 .3 6 2625 .4 36 63 2 Re si dua l 11 75 96 5876 12 69 .0 5988 7 To ta l 12 1820 72 .0 16 9 Si gn ific an ce F 1. 92 406E -1 4 Co effi ci en ts St anda rd Er ro r -4 6. 9629 55 2 6. 88 7277 83 8 1. 47 04 58 64 5 0. 02 86 98 02 7 tS ta t -6 .818 79 78 51 .2 3901 5 P-va /ue 2.88 09 2E -05 1. 9240 6E -1 4 Lo we r 9 5 % 6 2 1 217 5 9 1 4 1. 40 7294 68 1 Up pe r 95% Lo we r95 .0 % -3 1. 80 41 51 ~6 2. 12 1759 1 1.53 36 22 61 1. 40 72 94 68 Up per 95 .0 % -3 1.80 41 51 18 1. 53 3622 60 9 In te rc ep t X Vari ab le 1
VTI rapport 413A
RE SI DUAL O U T P U T Ob se rva ti on v NWVLDCDNCD (DOV NC ) w Y-r-Y Pred ic te d Y 100. 08 29 09 4 12 3. 6102 47 7 16 6. 9887 77 7 19 8. 4565 92 7 22 9. 7773 61 9 25 2.27 53 79 2 28 4.91 95 61 1 32 2. 4162 56 5 33 8.44 42 55 8 35 4.03 1 11 74 386. 82 23 45 2 444. 90 54 61 7 50 9.89 97 33 8 Re si dua ls 0. 0870 90 63 2 6. 28 9752 30 8 2. 45 12 2227 3 -1 0. 0865 92 7 6 5 773 6 1 8 8 5 7 053 7 9 1 5 -3.7 09 56 10 8 -4 .566 25 65 3 7.92 57 44 23 6 14.4 48 88 26 11.7 97 65 48 1 34 5 5 4 6 1 6 8 8 899 7 3 3 8 Appendix Sid 2 (11)
VTI rapport 413A Re gr es si on St at is ti cs Mul ti pl e R 0. 96 63 73 80 7 R Sq ua re 0. 93 38 78 335 Ad jus te d R Sq ua re 0. 92 7266 16 8 St an da rd Er ro r 29 9. 88 7855 Ob se rva ti on s 12 A N O V A df Re gr es si on 1 Re si dua l 10 To ta l 11 SS 12 70 17 7089 M S 12 70 17 71 89 93 27 .2 55 7 8993 2. 72 6 13 6010 98 15 F 14 1 .236 36 Si gn ific an ce F 3. 19 995E -0 7 S U M M A RY O U T P U T of C H A R T 10 (T he re gr es si on be twe en G D P an d ro ad tur no ver vo lum e to n-km ) Co ef fi ci en ts 6 6 9 9 0 2 6 7 9 9 13.7 83 62 66 8 In te rc ep t X Va ri ab le 1 RE SI DU AL O UTP U T St an da rd Er ro r 28 7. 52 34 57 6 tS ta t 2 329 9 0 6 1. 15 9819 06 2 11 .8 84 29 1 Pr ed ic tedY 92 8. 99 80 14 7 13 27 .3 44 82 6 16 18 .1 79 34 9 19 13 .1 48 95 9 21 32 .3 08 62 4 24 43 .8 18 58 7 27 9392 27 04 29 44 .164 23 5 9 30 84 .757 22 7 10 33 85 .2 40 28 9 11 39 36 .5 85 35 6 12 45 54 .0 91 83 1 Ob se rva ti on w NC DVLOCDNCD Re si dua ls -1 64 .9 98 01 47 2 43 3 4 4 8 2 5 7 -82. 17 93 48 58 -2 20 .148 95 95 ~1 4. 31 8623 65 21 6. 57 1413 4 42 6. 46 7295 8 43 0. 63 57 65 1 27 3. 34 27 72 9 42 .7 59 71 13 8 -181 .1 95 35 57 -4 83 .591 83 09 P-va/ue 0. 0420 56 9 3. 2E -0 7 L o we r 9 5 % Up pe r 95 % -1 31 0. 5449 77 2 9 2 6 0 3 8 2 3 6 11 .1 99 38 83 2 16 .3 67 86 50 4 L o we r 9 5 . 0 % -1 310. 54 49 77 11 .1 99 3883 2 Up per 95 .0 % 29 26 0382 36 16 36 78 65 04 Appendix Sid 3 (1 1)
VTI rapport 413A S U M M A R Y O U TPU T of C H A R T 12 Re gr essi on St at is ti cs Mul ti pl e R R Squa re Ad jus te d R Sq ua re St an dard Er ro r Obse rva ti on s A N O V A Re gr essi on Re si dua l To ta l In te rcep t X Va ri ab le1 RE SI DU AL O U T P U T 0. 9592 64 71 3 0. 92 01 88 79 0. 91 02 12 389 29.8 07 05 74 3 10 df FQO) Co ef fi ci en ts 12.9 36 42 80 7 0. 1089 78 39 88 81 948. 54 37 8 71 07.6 85 38 3 89 05 6. 22 916 Stan da rd Er ror 34.4 64 36 275 0. 0113 47 20 4 Ob se rva ti on r-NCOVLOCDNQDCDO1 Pr ed icte d Y 18 0. 32 7235 4 19 7. 43 68 42 7 24 3. 75 15 68 7 30 2.86 14 47 6 36 3. 88 9346 1 38 0. 71 66 99 3 37 8. 89 67 60 2 38 6. 51 43 49 7 42 2. 19 27 84 9 45 6. 53 29 65 4 Resi dua ls 8. 0427 64 57 8 25 .7 63 1573 2 2. 81 84 31 264 216 5 1 4 4 7 5 8 -4 6. 0393 46 1 -3 4. 34 66 9933 -10. 41 67 6021 12 .1 05 65 031 19 .2 57 21 514 44 .4 67 03 46 1 M S 81 94 8. 5437 8 88 8. 46 06 72 8 tSta t 0. 37 5356 65 9 9. 60 39 8594 7 F 92.2 36 54 60 7 P-va /ue 0. 71 71 5192 8 1.14 67 6E -0 5 Si gn if ic ance F 1. 14 67 6E-0 5 Lowe r 9 5 % (R eg re ss io nbe twe en po pul at io n of tr ucks an d vo lum e of ro ad fr eigh t tur no ve r) Up pe r 95 % Lowe r 9 5 . 0% 6653 85 86 35 92 .4 11 44 25 66 53 85 8635 0. 08 2811 67 3 0. 1351 45 11 0.08 28 11 673 Up per 95 .0 % 92.4 11 44 248 0. 1351 45 10 7 Appendix Sid 4 (11)
