The recent increase in biodiesel production subsidies in Indonesia and the rollout of the B7 mandate in Malaysia show that the two countries remain committed to the pursuit of their ambitious biodiesel mandates, despite the sharp drop in oil prices which has made biodiesel use clearly unattractive. However, in spite of these developments, we continue to believe the impact on demand will be lower than anticipated, as implementation of the mandates remains problematic. We believe Indonesia and Malaysia's attempts to keep the ailing biodiesel sector afloat are more of a long-term strategy to limit their dependence on imported oil in the coming years. Indeed, Indonesia is trying to address its very large current account deficit, while Malaysia is a net importer of refined oil.
Competitive Landscape
Table: Major Agribusiness Companies (USDmn)
Company Sub-Sector Revenue
(USDmn) Fiscal
Year End Market Capitalisation
(USDmn) Employees
Viet Nam Dairy Products
JSC (Vinamilk) Dairy 1651.2 12/2014 4954.7 6244
Kinh Do Corp
Food manufacturing (confectionery &
snacks) 233.8 12/2014 464.9 7318
Vinacafe Bien Hoa JSC Coffee & food
manufacturing 140.4 12/2014 215.8 527
Hung Vuong Seafood 703.5 12/2014 168.4 818
Societe De Bourbon Tay
Ninh Sugar & real estate 109.5 06/2014 82.9 269
Tuong An Vegetable Oil JSC Edible oils & fats 194.7 12/2014 36.6 780
Lam Son Sugar Sugar & alcohol 81.8 12/2014 35.3 720
Minh Phu Seafood Seafood 712.6 12/2014 na 14395
Southern Seed Crop seeds 28.7 12/2014 37.0 432
Viet Thang Feed JSC Animal feed 208.8 12/2014 57.5 630
Source: BMI, Bloomberg
Demographic Forecast
Demographic analysis is a key pillar of BMI's macroeconomic and industry forecasting model. Not only is the total population of a country a key variable in consumer demand, but an understanding of
the demographic profile is essential to understanding issues ranging from future population trends to productivity growth and government spending requirements.
The accompanying charts detail the population pyramid for 2015, the change in the structure of
the population between 2015 and 2050 and the total population between 1990 and 2050. The tables show indicators from all of these charts, in addition to key metrics such as population ratios, the urban/rural split and life expectancy.
Population
(1990-2050)
Vietnam - Population, mn
1990 2000 2005 2010 2015f 2020f 2025f 2030f 2035f 2040f 2045f 2050f
0 50 100 150
f = BMI forecast. Source: World Bank, UN, BMI
Vietnam Population Pyramid
2015 (LHS) & 2015 Versus 2050 (RHS)
Source: World Bank, UN, BMI
Table: Population Headline Indicators (Vietnam 1990-2025)
1990 2000 2005 2010 2015f 2020f 2025f
Population, total, '000 68,909 80,887 84,947 89,047 93,386 97,057 99,811
Population, % y-o-y na 1.1 0.9 1.0 0.9 0.7 0.5
Population, total, male, '000 33,892 39,827 41,830 43,970 46,158 47,980 49,302 Population, total, female, '000 35,017 41,060 43,117 45,077 47,228 49,076 50,508
Population ratio, male/female 0.97 0.97 0.97 0.98 0.98 0.98 0.98
na = not available; f = BMI forecast. Source: World Bank, UN, BMI
Table: Key Population Ratios (Vietnam 1990-2025)
1990 2000 2005 2010 2015f 2020f 2025f Active population, total, '000 39,197 50,153 56,330 62,305 66,093 68,401 70,001 Active population, % of total population 56.9 62.0 66.3 70.0 70.8 70.5 70.1 Dependent population, total, '000 29,712 30,733 28,617 26,741 27,292 28,655 29,810 Dependent ratio, % of total working age 75.8 61.3 50.8 42.9 41.3 41.9 42.6
Key Population Ratios (Vietnam 1990-2025) - Continued
1990 2000 2005 2010 2015f 2020f 2025f Youth population, total, '000 25,778 25,543 23,038 20,918 20,950 20,690 19,395 Youth population, % of total working age 65.8 50.9 40.9 33.6 31.7 30.2 27.7
Pensionable population, '000 3,934 5,190 5,578 5,823 6,342 7,964 10,414
Pensionable population, % of total working age 10.0 10.3 9.9 9.3 9.6 11.6 14.9
f = BMI forecast. Source: World Bank, UN, BMI
Table: Urban/Rural Population & Life Expectancy (Vietnam 1990-2025)
1990 2000 2005 2010 2015f 2020f 2025f
Urban population, '000 13,957.7 19,715.6 23,174.6 27,064.2 31,383.5 35,771.3 40,027.3
Urban population, % of total 20.3 24.4 27.3 30.4 33.6 36.9 40.1
Rural population, '000 54,952.2 61,172.3 61,773.2 61,983.2 62,003.1 61,285.7 59,783.9
Rural population, % of total 79.7 75.6 72.7 69.6 66.4 63.1 59.9
Life expectancy at birth, male, years 66.1 69.0 69.9 70.7 71.7 72.7 73.7
Life expectancy at birth, female, years 75.1 78.5 79.6 80.2 80.7 81.2 81.7 Life expectancy at birth, average, years 70.6 73.8 74.8 75.5 76.2 77.0 77.8
f = BMI forecast. Source: World Bank, UN, BMI
Table: Population By Age Group (Vietnam 1990-2025)
1990 2000 2005 2010 2015f 2020f 2025f
Population, 0-4 yrs, total, '000 9,314 7,127 6,897 7,228 7,012 6,574 5,922 Population, 5-9 yrs, total, '000 8,606 9,253 7,023 6,790 7,180 6,968 6,535 Population, 10-14 yrs, total, '000 7,856 9,162 9,117 6,898 6,757 7,147 6,936 Population, 15-19 yrs, total, '000 7,359 8,492 9,050 9,011 6,865 6,725 7,116 Population, 20-24 yrs, total, '000 6,644 7,672 8,332 8,873 8,936 6,802 6,664 Population, 25-29 yrs, total, '000 6,005 7,065 7,470 8,111 8,772 8,837 6,717 Population, 30-34 yrs, total, '000 5,138 6,351 6,909 7,285 8,021 8,680 8,747 Population, 35-39 yrs, total, '000 3,888 5,803 6,241 6,763 7,207 7,939 8,596 Population, 40-44 yrs, total, '000 2,462 4,994 5,719 6,147 6,684 7,127 7,856 Population, 45-49 yrs, total, '000 2,016 3,753 4,935 5,647 6,054 6,588 7,031
Population By Age Group (Vietnam 1990-2025) - Continued
1990 2000 2005 2010 2015f 2020f 2025f
Population, 50-54 yrs, total, '000 1,968 2,345 3,699 4,855 5,521 5,926 6,457 Population, 55-59 yrs, total, '000 2,045 1,885 2,237 3,541 4,677 5,330 5,733 Population, 60-64 yrs, total, '000 1,668 1,790 1,734 2,068 3,352 4,443 5,079 Population, 65-69 yrs, total, '000 1,411 1,770 1,609 1,562 1,906 3,104 4,134 Population, 70-74 yrs, total, '000 1,027 1,322 1,530 1,399 1,379 1,695 2,776
Population, 75-79 yrs, total, '000 752 984 1,080 1,263 1,166 1,159 1,437
Population, 80-84 yrs, total, '000 429 596 731 814 964 900 903
Population, 85-89 yrs, total, '000 223 336 385 482 545 653 617
Population, 90-94 yrs, total, '000 71 132 177 209 267 306 372
Population, 95-99 yrs, total, '000 15 40 52 74 89 115 133
Population, 100+ yrs, total, '000 1 6 11 16 23 30 38
f = BMI forecast. Source: World Bank, UN, BMI
Table: Population By Age Group % (Vietnam 1990-2025)
1990 2000 2005 2010 2015f 2020f 2025f
Population, 0-4 yrs, % total 13.52 8.81 8.12 8.12 7.51 6.77 5.93
Population, 5-9 yrs, % total 12.49 11.44 8.27 7.63 7.69 7.18 6.55
Population, 10-14 yrs, % total 11.40 11.33 10.73 7.75 7.24 7.36 6.95
Population, 15-19 yrs, % total 10.68 10.50 10.65 10.12 7.35 6.93 7.13
Population, 20-24 yrs, % total 9.64 9.49 9.81 9.97 9.57 7.01 6.68
Population, 25-29 yrs, % total 8.72 8.73 8.79 9.11 9.39 9.11 6.73
Population, 30-34 yrs, % total 7.46 7.85 8.13 8.18 8.59 8.94 8.76
Population, 35-39 yrs, % total 5.64 7.17 7.35 7.60 7.72 8.18 8.61
Population, 40-44 yrs, % total 3.57 6.17 6.73 6.90 7.16 7.34 7.87
Population, 45-49 yrs, % total 2.93 4.64 5.81 6.34 6.48 6.79 7.04
Population, 50-54 yrs, % total 2.86 2.90 4.36 5.45 5.91 6.11 6.47
Population, 55-59 yrs, % total 2.97 2.33 2.63 3.98 5.01 5.49 5.74
Population, 60-64 yrs, % total 2.42 2.21 2.04 2.32 3.59 4.58 5.09
Population, 65-69 yrs, % total 2.05 2.19 1.90 1.75 2.04 3.20 4.14
Population, 70-74 yrs, % total 1.49 1.63 1.80 1.57 1.48 1.75 2.78
Population, 75-79 yrs, % total 1.09 1.22 1.27 1.42 1.25 1.20 1.44
Population, 80-84 yrs, % total 0.62 0.74 0.86 0.92 1.03 0.93 0.91
Population By Age Group % (Vietnam 1990-2025) - Continued
1990 2000 2005 2010 2015f 2020f 2025f
Population, 85-89 yrs, % total 0.32 0.42 0.45 0.54 0.58 0.67 0.62
Population, 90-94 yrs, % total 0.10 0.16 0.21 0.24 0.29 0.32 0.37
Population, 95-99 yrs, % total 0.02 0.05 0.06 0.08 0.10 0.12 0.13
Population, 100+ yrs, % total 0.00 0.01 0.01 0.02 0.03 0.03 0.04
f = BMI forecast. Source: World Bank, UN, BMI
Methodology
Industry Forecast Methodology
BMI's industry forecasts are generated using the best-practice techniques of time-series modelling and causal/econometric modelling. The precise form of model we use varies from industry to industry, in each case being determined, as per standard practice, by the prevailing features of the industry data being examined.
Common to our analysis of every industry is the use of vector autoregressions. Vector autoregressions allow us to forecast a variable using more than the variable's own history as explanatory information. For
example, when forecasting oil prices, we can include information about oil consumption, supply and capacity.
When forecasting for some of our industry sub-component variables, however, using a variable's own history is often the most desirable method of analysis. Such single-variable analysis is called univariate modelling. We use the most common and versatile form of univariate models: the autoregressive moving average model (ARMA).
In some cases, ARMA techniques are inappropriate because there is insufficient historic data or data quality is poor. In such cases, we use either traditional decomposition methods or smoothing methods as a basis for analysis and forecasting.
BMI mainly uses ordinary least squares estimators. In order to avoid relying on subjective views and encourage the use of objective views, we use a 'general-to-specific' method. BMI mainly uses a linear model, but simple non-linear models, such as the log-linear model, are used when necessary. During periods of 'industry shock', for example, if poor weather conditions impede agricultural output, dummy variables are used to determine the level of impact.
Effective forecasting depends on appropriately selected regression models. We select the best model according to various different criteria and tests, including but not exclusive to:
■ R2 tests explanatory power; adjusted R2 takes degree of freedom into account;
■ Testing the directional movement and magnitude of coefficients;
■ Hypothesis testing to ensure coefficients are significant (normally t-test and/or P-value);
Human intervention plays a necessary and desirable role in all or our industry forecasting. Experience, expertise and knowledge of industry data and trends ensure analysts spot structural breaks, anomalous data, turning points and seasonal features where a purely mechanical forecasting process would not.
Sector-Specific Methodology
Within the Agribusiness industry, issues that might result in human intervention could include but are not exclusive to:
■ Technology development that might influence future output levels (for example greater use of biotechnology);
■ Dramatic changes in local production levels due to public or private sector investment;
■ The regulatory environment and specific areas of legislation, such as import and export tariffs and farm subsidies;
■ Changes in lifestyles and general societal trends;
■ The formation of bilateral and multilateral trading agreements, and political factors.
The following two examples show the demand (consumption) and the supply (production) of rice. Note that the explanatory variables for both are quite similar, but the underlying economic theory is different.
Example Of Rice Consumption Model
(Rice consumption)t = β0 + β1*(real private consumption per capita)t + β2*(inflation)t + β3*(real lending rate)t + β4*(population)t + β5*(government expenditure)t + β6*(food consumption)t-1 + εt
Where:
■ β are parameters for this function.
■ Real private consumption per capita has a positive relationship with rice consumption, if rice is a normal good in a particular country. If rice is an inferior good in a country, the relationship is negative. So the sign of β1 is determined by a specific product within a specific country.
■ When inflation is high, people with rational expectations will consume today rather than wait for tomorrow's high price to come. Higher rice demand in year t due to higher inflation in that year leads to an assumed positive sign of β2.
■ The relationship between real lending rate and rice consumption is expected to be negative. When real lending rates increase, disposable incomes, especially for those with mortgage burdens, etc, will decrease.
So the sign of β3 is expected to be negative.
■ Of course, other things being equal, growth in rice consumption can also be caused by growth in population. Consequently, positive sign of β4 is expected.
■ Government expenditure typically causes total disposable incomes to rise. So the sign of β5 is expected to be positive.
■ Human behaviour has a trend: A high level of food consumption in previous years means there is very likely to be a high level of food consumption the next year. So the positive sign of β6 is expected.
■ ε is the error/residual term.
Example Of Rice Production Model
(Rice production)t = β0 + β1*(real GDP per capita)t + β2*(inflation)t + β3*(real lending rate)t + β4*(rural population)t + β5*(government expenditure)t + β6*(food production)t-1 + εt
Where:
■ The same as above: the relationship between real GDP per capita and rice production depends on whether rice is normal or inferior good in that country.
■ If high inflation is caused by food prices increasing, farmers will be more profitable. Then they will supply more agricultural product (eg rice) to increase their marginal (extra) profit, although this is tempered by the rising cost of other inputs in line with inflation.
■ There is a global move towards corporate farming, away from small holdings, in order to achieve greater agricultural productivity. Corporate farming means more investment in the modes of production, ie agricultural machinery. Higher real lending rates discourage investment, which in turn reduce production.
■ BMI assumes that only the rural population has a positive effect on agricultural product supply.
■ With supportive government policy, other things being equal, rice production is expected to go up. Government expenditure is likely to play some role in supporting agribusiness.
■ Again, previous food production positively affects this year's prediction.