06/11/2017
1
Building Capacity for Enhancing Bioenergy Sustainability
Through the Use of the GBEP indicators
Measuring “GBEP Indicator 10” with the AGLINK-COSIMO
Model
Introductory Training
09-10 November 2017
Hanoi – Viet Nam
2
Dr. Holger Matthey
Economist and Team leaderMedium-term Outlook and Market Analysis
Trade and Markets Division (EST)Food and Agriculture Organization of the United Nations (FAO)
Familiarize participants with
• the structure and general functionality of the Aglink-Cosimo model and,
• its application in assessing the impact of the biofuel sector on domestic agricultural commodity markets.
Design and discuss simulation scenarios that closely reflect their interests and concerns in the development of the bioenergy and agricultural sectors.
3
Scope and objectives of training
06/11/2017
2
Country training programs
• Argentina
• Brazil
• China
• Czech Rep.
• European Commission
• India
• Indonesia
• Japan
• Mexico
4
• New Zealand
• Pakistan
• Paraguay
• Poland
• United Kingdom
• South Africa
• South Korea
• Thailand
Workshop agenda
Thursday• Introduction to the workshop and collaboration project
• Introduction to medium-term commodity projections
• Aglink-Cosimo model
• Model-based outlook generation and use
Friday• General scenario analyses with Aglink-Cosimo
• Biofuel applications and exercises
5
6
Introduction to medium-term commodity projections
Session 1
OECD-FAO medium-term projections methodology
06/11/2017
3
Forward looking analyses
• Forward-looking policy
and market analyses to
support evidence-based
policymaking
• Promote and foster a
model-based dialogue on
global prospects for food
and agriculture between
stakeholders and FAO
• Consensus baseline
• Scenario analyses
8
Outlook tool Analysis tool
• Medium-term outlook
model to produce
baseline projections
• Provides guidelines for
decision making
• Control in scenario
work
• Conduct studies on
policy reform
• Timely assessment of
emerging market
issues
• Disaggregating of
complex interactions
Aglink-Cosimo
OECD-FAO Agricultural Outlook
• Produced annually from December to June
• Ten-year consensus projection of global supply,
demand and trade
• Assessment of driving factors in commodity markets
• Theme Chapter – shared messages for FAO-OECD
• Collaborative effort between various organizations
06/11/2017
4
10
Going beyond the OECD-FAO Outlook…
• Detailed country outlooks and capacity building for
member countries
• Support to country and regional projects in HQ,
decentralized offices and partner institutions
• Scenario analysis
• Customized input in downstream analyses
Applications of Aglink-Cosimo
• GBEP Tools
• Fish Model
• Fruits, Fibers and Tea Model
• Indirect Land Use Change assessments
• Feed Demand System
• Bilateral Trade Flows
11
12
Questions?
06/11/2017
5
13
Introduction to medium-term commodity projections
Session 1
Highlights of the 2017-2026 Agricultural Outlook
Agricultural commodity markets in 2016/17
• Above-average capital investments in recent years
• Record production and abundant stocks of most
commodities in 2016
• Prices of cereals, meats and dairy products
continued to decline from peaks experienced in the
last decade
• Oilseeds, vegetable oils, and sugar provide
exceptions to this trend, with prices rebounding in
2016
06/11/2017
6
Key macroeconomic assumptions
• Global population growth slow down to 1% p.a., but
strong regional differences remain
• Economic recovery in most industrialized countries,
but growth prospects in large emerging countries
diverge
• Real currency appreciation expected in major
exporters
• Crude oil reference price rising from $44 to $90
Consumption Outlook
17
Main Messages
• Global consumption continues to expand and evolve
toward higher value commodities
• Growth will slow down compared to previous decade
• China, India and Sub-Saharan Africa drive global growth
• Convergence in per capita food consumption patterns
remains limited
Consumption by commodity
18
Annual growth in demand for key commodity groups, 2007-16 and 2017-26
Global consumption
growth slows down
Total demand mainly
driven by population
growth
Per-capita food
demand limited by:
• Saturation
• Access
For fresh dairy
products, vegetable oil
and sugar strong per-
capita increase
expected
-
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
200
7-16
2017
-26
200
7-16
2017
-26
200
7-16
2017
-26
200
7-16
2017
-26
200
7-16
2017
-26
200
7-16
2017
-26
200
7-16
2017
-26
Cereals Meat Fish Freshdairy
Roots andtubers
Sugar Vegetableoil
%
Due to per capita consumption growth or non-food consumption growth
Due to population growth
06/11/2017
7
Calorie consumption per capita
19
Per capita caloric availability in 2006, 2016 and 2026Increasing trend in caloric availability per capita to continue
• Bridging caloric availability
levels between developing
countries and OECD
• China reaches the overall level
of the OECD caloric availability.• India makes substantial
progress
Cereals remain to be the most
important source of calories across
the world.
Bulk of additional calories from
sugar and vegetable oil
Dietary preferences persist and
influence nutrition patterns
-
500
1,000
1,500
2,000
2,500
3,000
3,500
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
Sub-SaharanAfrica
India China SoutheastAsia
OECD
kca
l/d
ay/p
erso
n
Cereals Roots and tubers Vegetable oilsMeat Sugar DairyFish Other
Protein consumption per capita
20
Protein availability remains diverse
Key difference is the level of animal protein consumption between regions
• low levels in Sub-Saharan Africa, India, and Southeast Asia
In India, main animal protein is coming from the fast expanding fresh dairy product consumption
In Southeast Asia, fish consumption is important
Limited income growth in poor rural and urban households and the slow development of a retail infrastructure for animal protein are seen as the main constraints to animal protein consumption growth
-
20
40
60
80
100
120
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
200
6
2016
2026
Sub-SaharanAfrica
India China SoutheastAsia
OECD
g/d
ay/p
erso
n
Cereals Meat Dairy Fish Other
Per capita protein availability in 2006, 2016 and 2026
Production Outlook
Main Messages
• Yield growth will be the main driver of food crop and
feed production
• Growth due to 80% yield and 20% land
• Diverse growth in livestock and fish production
• Intensive and extensive production systems coexist
• Developing countries lead the agricultural production
growth
21
06/11/2017
8
Crop production
Yield improvements main driver of crop production growth
Wheat:
– 9% yield growth– 2% net area gain (4 million ha)
mostly in in Russia and India
Maize:
– 11% yield growth, large yield gap remains,
– 2% area growth (4 million ha), all in Africa and Latin America
Rice:
– 12% yield growth, – small net area gain globally (2
million ha), shift into LDC with low yield base
Soybean:
– 9% yield growth, yield gap small and closing,
– 14% area net growth, 16 million ha,
22
0
5
10
15
20
25
Wheat Maize Rice Soybean
Gro
wth
20
14
-16
to 2
02
6 (%
)
Area Yield
Relative growth shares of area and yield
75%
0.3
5.3
2.71.53.1
25%
12.0
7.9
6.40
5
10
15
20
25
30
35
40
45
Beef Pork Poultry Sheep Beef Pork Poultry Sheep Total
Developing Developed Increase
Meat production
23
Total meat, 13% growth
globally
46% of it in poultry,
becoming leading meat
Pigmeat production
expansion slowed by
increased environmental
regulations and animal
welfare concerns
Beef production expands
in Argentina, China, Brazil, India
Sheepmeat sector global
growth of 22%, faster than
last decade
39 million tons of additional meat produced by 2026
Mt
(c.w
.e/
r.t.
c)
Trade Outlook
24
Main Messages
• Growth in trade is closely matched with output
growth in the coming decade
• High concentration in export markets persists
• Growing import demand from developing countries
• Role of trade continues to differ by commodity
06/11/2017
9
-100
-50
0
50
100
150
2006 2011 2016 2021 2026Bill
ion
co
nst
ant
USD
(20
04-0
6 b
ase)
Regional net trade in agricultural products
Americas Sub-Saharan Africa
Asia Other
North Africa/Middle East
Trade by region
25
• The Near East and Africa increasingly dependent on agricultural imports
• Eastern Europe and Central Asia region now established as main exporter, due to high exports of cereals
• Regional net trade flow evolution reflects continuing changes in regional production and consumption patterns
Projections of commodity prices
26
Supply and demand
fundamentals keep
international reference
prices in line with inflation
Most commodity group
prices follow similar trends
due to substitutability and
complementarities
Meat prices have historically
followed a somewhat
different path, meat prices
are expected to fall in real
terms to levels similar to
those in the early 2000s
50
100
150
200
250
300
350
400
1990 1995 2000 2005 2010 2015 2020 2025
20
02-
04 =
100
Projected FAO Food Price Indexes - Nominal
Cereals Sugar Vegetable oils
Meat Dairy Total Agriculture
Long term trend of real
commodity prices
27
• Prices follow the long run declining trend, with an average price decrease of
about 1.5% per year in real terms
• Prices of agricultural commodities are subject to considerable volatility and may
show large deviations from their long-term trends for an extended period of time
• Prices eventually returned to their long-term trend
0
500
1000
1500
2000
2500
1900 1909 1918 1927 1936 1945 1954 1963 1972 1981 1990 1999 2008 2017 2026
USD
(20
10 b
ase)
Wheat Maize Rice
06/11/2017
10
Key takeaways and policy
challenges
Consumption Production
• Slowing demand growth • Few sources of demand to replace China
& biofuels• Rising demand for milk, sugar and oils
Policies to combat malnutrition
• Output gains mostly from yield increases• Important sustainability challenges – esp.
in SE Asia• Big yield gaps in poorest countriesInvest in sustainable productivity growth
Trade Prices
• Trade growth to slow• Main cause is weak demand• Export trade to remain concentrated
Need for open and reliable markets
• Real prices at or below current levels• Always a risk of specific price swings!
Market information and risk management
Southeast Asia:
Prospects and Challenges
Developments in agriculture and fisheries in Southeast Asia
• Extensive structural adjustment
• Changes in fisheries production
• Drivers of production growth in Southeast Asia
• Growing regional participation in world food markets
• Agricultural policies in Southeast Asia: A focus on rice
and self-sufficiency
• Fisheries policies in Southeast Asia: the sustainability
and food security challenge
Southeast Asia - Consumption Food expenditure shares (%)
0
1
2
3
4
5
6
7
8
2001 2006 2011 2016 2021
%
Dairy SugarOther Cereals OtherVeg Oil
0
5
10
15
20
25
30
35
2001 2006 2011 2016 2021
%
Meat/eggs Fish Rice
06/11/2017
11
Calorie availability
-50
0
50
100
150
200
250
Ch
an
ge
in K
ca
l/d
ay
Other crops Rice Animal
Protein availability
-2
0
2
4
6
8
10
g/p
ers
on
/da
y
Other crops Rice Animal
Southeast Asia - Production Changes in major production activities in Southeast Asia
Increase in production across major production activities, 2016-26
0
5
10
15
20
25
Mill
ion
to
nn
es
Viet Nam Thailand
Philippines Myanmar
Cambodia Laos
Malaysia Indonesia
06/11/2017
12
Southeast Asia - Trade
Contributions to changes in the trade balance
Changes in net trade balance 2016-26
-3
-1
1
3
5
7
Vegetable oil Rice Meat & eggs Maize Sugar Roots andtubers
Mill
ion
to
nn
es
Philippines Myanmar Laos
Cambodia Thailand Viet Nam
Malaysia Indonesia
35
Download the publication from:
http://www.fao.org/publications/oecd
-fao-agricultural-outlook/2017-
2026/en/
Access the full database at:
http://www.agri-outlook.org/data/
Contact us for further information at:
36
Aglink-Cosimo model
Session 2
Aglink-Cosimo model structure
06/11/2017
13
Vision and main objectives
•Provide consensus analyses on the future evolution of international commodity markets.
•Develop increasingly integrated systems that link short, medium and long term projections.
•Publication of the annual “OECD-FAO Agricultural Outlook” in collaboration with the OECD.
•Construct scenarios analyzing emerging market and policy issues using the Aglink-Cosimo model.
37
The Outlook: Forecast or Baseline?
• Expectation for the future
• Based on clearly defined assumptions
• A prediction, as of coming events or conditions = Forecast
• A measurement, calculation, or location used as a basis for comparison = Baseline
The Projections team hard at work
39
06/11/2017
14
40
Model
Data
Parameters
Projection Output
EndoExo
Aglink-Cosimo• Partial equilibrium model that provides structural
representations of national and global agricultural
markets where all of the major agricultural sectors are
assumed to be connected
– Non-agricultural markets are not modelled and are treated
exogenously to the model
• Specific assumptions regarding: – Macroeconomic conditions
– Agriculture and trade policy settings
– Weather conditions
– Longer term productivity trends
– International market developments
41
• Model, driven by elasticities, technical parameters and policy variables,
• provides representations of national and global agricultural markets where all of the major agricultural sectors are assumed to be connected,
• outlook simulation tool that constructs projections over a ten year period so that all of the main characteristics of the crops and livestock sectors influence the final equilibrium.
Aglink-Cosimo
06/11/2017
15
Methodological Approach
Key assumptions of the model:
1. World markets for agricultural commodities arecompetitive.
2. Domestically produced and traded commodities areviewed to be perfect substitutes by buyers and sellers.
3. Aglink-Cosimo is a partial equilibrium model for the mainagricultural commodities. Non-agricultural markets arenot modelled and are treated exogenously to the model.
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
Wheat Beef Skim Milk Powder
Coarse Grains** Sheepmeat Whole Milk Powder
Rice Pigmeat Cheese
Oilseeds** Poultry Butter
Vegetable Oils** Eggs Fresh Dairy Products
Oilseed Meals** Sugar Ethanol
Roots and Tubers
Animal Feed
Cotton Bio-diesel
Fish (Separate Model)
Aglink – Cosimo Commodities
** Indicates sectors which may be disaggregated
06/11/2017
16
46
Norway Switzerland
AGLINK Countries
Australia Canada EU-27 Japan S.-Korea
Mexico N-Zealand USA
Argentina Brazil China Russia
Aglink (OECD) Countries Cosimo (FAO) Countries and Regions
Australia Algeria Nigeria Other LDC Africa
Argentina Bangladesh Pakistan Other LDC Asia
Brazil Chile Paraguay Other LDC Oceania
Canada Colombia Peru Other Western Europe
China Egypt Philippines Other Asia Developed
European Union Ethiopia Saudi Arabia Other Eastern Europe
Japan Ghana South Africa Other Middle East
Mexico Haiti Sudan Other North Africa
New Zealand India Tanzania Other Africa
Norway Indonesia Thailand Other Asia
Russia Iran Turkey Other Oceania
South Korea Israel Ukraine Other Latin America
Switzerland Kazakhstan Uruguay
United States Malaysia Vietnam Mozambique Zambia
Geographical coverage
27 country modules were added to Cosimo
in 2017
• Angola• Burkina Faso• Chad• Democratic Republic of the
Congo• Madagascar• Malawi• Mali• Mauritania• Rwanda• Senegal• Somalia• Uganda
• Myanmar• Cambodia• Lao People's Democratic
Republic• Yemen
• Iraq• Lebanon• Libyan Arab Jamahiriya• Morocco• Tunisia
• Cameroon• Congo• Côte d'Ivoire• Gabon• Kenya• Zimbabwe
48
Model
Data
Parameters
Projection Output
EndoExo
06/11/2017
17
Annual time series for:
Historic (endogenous)• prices• supply (area, yield, animal numbers...)• demand (food, feed, crush...)• trade (exports, imports)
Projection (exogenous)• macroeconomic data (GDP, ex. rate...)• policy variables (tariffs, CAP...)
Data requirements
50
• national statistics
• international databases
• surveys, questionnaires
• literature
• calculations
Data sources for endogenous data
51
Sources (historic series and projections):• national statistics• international databases• surveys, questionnaires• literature• calculations
Exogenous projections may have to be done by you!
Methods:• trend• time series methods• regression• expert opinion
Data sources for exogenous data
06/11/2017
18
52
Model
Data
Parameters
Projection Output
EndoExo
53
• are the link between variables
• determine the properties of the model
IDNparms.inp
Parameters
54
Introduction to parameters
• types
• function
• sources
Parameters
06/11/2017
19
55
Steps to develop and modify the parameter set• example of non-estimation method• based on available estimates
VNMparm.xlsm
Parameters
Parameters
• Strategy for parameter choices
1. Use available estimates.
2. Use systems / appropriate constraints
3. Estimate: research estimation agenda
Model validation by country / by commodity
Emphasis on consultation with experts
57
Model
Data
Parameters
Projection Output
EndoExo
06/11/2017
20
• Produced annually from December to June.
• Ten-year projection of global supply, demand and trade
• Assessment of driving factors in commodity markets
• Theme Chapter – shared messages for FAO-OECD
• Growing scope – fish, cotton, land, fertilizers....
• Collaborative effort between various international organizations and national institutions
OECD-FAO Agricultural Outlook
Uncertainties and Limitations
Related to:– macroeconomic developments
– technology advances (yields, biofuel)
– energy prices,
– weather-related production shocks,
– disease outbreaks,
– agricultural policy developments
60
Questions?
06/11/2017
21
61
Model based outlook generation
Session 3
Medium term baseline process
Definition of partial equilibrium model
“A partial equilibrium is one which is based on only a restricted range of data, a standard example is price of a single product, the prices of all other products being held fixed during the analysis.”
Source: Definition by George Stigler in Jain, T.R. (2006-07). Microeconomics and Basic Mathematics. New Delhi: VK Publications. p. 28.
Advantages of PE modeling
1. Limited data requirements
2. Disaggregated and detailed sector analysis
06/11/2017
22
Assumptions of a partial equilibrium model
• The analysis only considers the effects of a given policy action on the market that is directly affected;
• Price and Quantity of a given commodity are taken as endogenous (determined) and all other goods are constant and exogenous to the analysis;
• Markets are perfectly competitive
Assumptions of PE model (2)
• Cost minimization or profit maximization on the production side;
• Utility maximization (given budget constraint) on the consumption side
AGLINK-Cosimo and partial equilibrium modeling (3)
• The functional relationships linking supply and demand to prices are, in most cases, linear in the logarithms of the variables
• Equation coefficients are partial elasticities
06/11/2017
23
AGLINK-Cosimo and partial equilibrium modeling
SummaryAGLINK-Cosimo is a recursive-dynamic, partial equilibrium, supply-demand model for the main agricultural commodities consisting of behavioural equations which represent responses of economic agents making supply and demand decisions
68
Questions?
69
COUNTRY MODULE
06/11/2017
24
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
Supply
• Crop production:
Area=f(Return/ha, Paymt/ha, Costindex,Area(-1),Rfactor)
Yield=f(Return/t,Paymt/t,Costindex,trend,Rfactor)
Production=Area*Yield
• Oilseed products
OilseedMeal production=mealyield*oilseed crush
Protein meal production=oilseed meal + others…
Oilseedoil production=oilyield*oilseed crush
Veg oil production=oilseed oil + palm oil + others…71
• Meat production:Livestock Inventory=f(Meatprice,Paymt/hd,
Feedpriceindex,Costindex,Inventory(-1),trend,Rfactor)Indigenous meat production=f(Meatprice,Paymt/tn,
feedprice index,costindex,Inventory(-1), production(-1),trend,Rfactor)
• Milk production:Cow Inventory=f(Milkprice,Paymt/hd,
Feedpriceindex,Costindex,Inventory(-1),trend,Rfactor)Cow yield=f(Milkprice,Paymt/tn, feedprice index,costindex,
trend,Rfactor)Milk production= Inventory*yieldMilk products=f(fatprice,proteinprice)
72
Supply
06/11/2017
25
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
• Food demand
Food demand= f(Ownprice, other prices, pricedeflator, Income/person,trnd,Rfactor)
» Demand constraints apply in log linear relationship
• Crop Feed demand
Feed demand=f(Ownprice, otherprices, nonruminant prod, ruminant prod,trnd, Rfactor)» Fish meal, DDGs are deducted
• Bio-fuel crop feedstock demand Feedstock demand=f(Pricebiofuel, feedstock price, mandate,subsidy,
Rfactor)
• Crop Other use– Other use=f(Price, income,trnd,Rfactor)
74
Demand
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
06/11/2017
26
• Crops
Stocks= f(Ownprice/(average(ownprice), production, Rfactor)
• Meat and dairy products
Stocks=f(Ownprice, production, Rfactor)
76
Stocks
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
• Exports
Exports= f(producerprice/exportprice, Rfactor)
• Imports
Imports=f(producerprice/importprice, Rfactor)
78
Trade
06/11/2017
27
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
• Export price
Exportprice= worldprice*(1+exportwedge)*exchangerate
• Import price
Importprice=worldprice*(1+tariff+importwedge)*exchangerate
• Producer price: domestic market clearingProduction+stocks(-1)+imports=consumption+exports+stocks
• Consumer priceconsumerprice=f(producerprice,deflator)
80
Prices
81
GLOBAL AGLINK-COSIMO MODEL
06/11/2017
28
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
How do we carry out the Outlook?1/ Country expertise is taken into account as a starting point
A questionnaire is sent to each collaborator country
2/ The AGLINK-COSIMO model is used to get
a consistent and coherent picture
Collaboration between FAO and OECD Secretariats
3/ Model Outcomes are adjusted through expert opinions
Collaboration between FAO and OECD Secretariats
4/ Final review of projections in OECD commodity groups
Projections (not forecast) are discussed with member/observer countries
Main steps to obtain the baseline
Step 1 : Country standalonesCalibrate country models on questionnaire replies
Step 2 : Merge of country standalone and of COSIMOGet the AGLINK – COSIMO model
Step 3: Computation of the baselineWork on the global model on a commodity base
06/11/2017
29
85
• All country modules are merged• Models• Data files• Parameter files
• World price endogenized
• Equilibrium solution for all countries and all commodities
Introduction to the global “all linked” model concept
Domestic price (internal market clearing)
Net trade(exports – imports)
World trade balance∑EX = ∑IM
Equilibrium world price
Production
Area/
livestock
Yield
Opening
stocks
Imports
Income
PopulationConsumption
(food, feed,
other use)
Ending
stocks
Exports
Aglink-Cosimo Model
World Prices Market Clearing
• Wheat, Coarse grain, rice, oilseeds, raw sugar, poultry, sheep meat, butter, cheese, skim powder, whole powder, ethanol, biodiesel:– Global exports = Global imports
• Bovine and pig meats:– Pacific market: exports=imports– Atlantic market: exports = imports– Rest of world: exports=imports
87
Model closure
06/11/2017
30
Aglink – Cosimo Outlook Process
Aglink
Cosimo
OECD-FAO World Agricultural Outlook
OECD Questionnaire
Responses
Country Model Development
FAO Experts
CalibratedStand-alone
Models
AdjustedStand-alone
Models
Solve for worldprices
Iterations,exchanges of modifications
OECDCommodity Groups
OECD Experts
FAO Databases
Solve for domestic prices
89
Model use
Session 4
Aglink-Cosimo model functionality
90
Introduction to Troll-based Modeling
06/11/2017
31
The Troll Simulation Software
91
• Integrated software for econometric, modeling and statistical analysis
• State-of–the-art model simulation engine designed for large systems
• Complete integration of various tasks (calculation, model editing, estimation, simulation...)
• Hundreds of built-in functions and sophisticated modelinglanguage
• Efficient interface to MS-EXCEL
• Available for any platform, MS-WINDOWS or UNIX
92
VNM_RI_FO
Country
Coding
93
VNM_RI_FO
Commodity
Coding
06/11/2017
32
94
Coding
VNM_RI_FO
Activity
95
VNM_RI_FO
CountryCommodity
Activity
Coding
96
Simplified country model
06/11/2017
33
Model File
Driver (program)
REGmod.inp
REGsys.inp
Model Definitions
SECTORS
Wheat
Coarse Grains
ACTIVITIES
SUPPLYArea
Yield
DemandFood
Stocks
TradeImports
Exports
Price Domestic
Model Definitions (2)
06/11/2017
34
DATAEXO
ENDREGexo.csv
REGhist.csv
Model Definitions (3)
PARAMETERS
REGparms.inp
Model Definitions (4)
102
• Simulation exercises with global model– Baseline generation
– create a projection with model, put output in viewer
• Adjustments of other countries– read in new r-factors,
– modify the macro variables
• Parameter modifications – modify parameters,
• Structural changes– exercises to change equation structure
• Model improvement discussion
Practical exercises using the model
06/11/2017
35
103
Questions?
Friday
Day 1 - Key takeaways
Session 1 Session 2
• Forward looking analysis to support evidence based policymaking
• Aglink-Cosimo – baseline and scenarios• Global agricultural Outlook
Model analysis, consensus projections
• Partial equilibrium model• Exogenous and endogenous data• Parameters connect variables
Partial equilibrium model capabilities
Session 3 Session 4
• Model assumptions• Equation structure• Global model
Specify equations to reflect markets
• Construct a simple model• General scenario analysis• Evaluate the results
Scenario setup is crucial
06/11/2017
36
106
Questions?
107
Basic scenarios
Session 5
General scenario analyses with Aglink-Cosimo
108
• Simulation exercises with global model– Baseline generation
– create a projection with model, put output in viewer
• Adjustments of other countries– read in new r-factors,
– modify the macro variables
• Parameter modifications – modify parameters,
• Structural changes– exercises to change equation structure
• Model improvement discussion
Practical exercises using the model
06/11/2017
37
109
Aglink–Cosimo Standalones
• Standalone model construction and baseline generation
– how modules are put together, solution process
• Analysis and scenario exercises
– demonstration exercises running these scenarios, compare, discuss one example chosen by the group
VNMys.inp step2.inp Country_Viewer
110
Introduction to the Viewer
Overview over viewer
• Review the results• Projections• Indicators• Ratios
• Introduce revisions• Endogenous• Exogenous
VNMviewer
Baseline projections for agricultural markets
Scenarios involving:
– yields
– areas
– costs
– macroeconomic variables
– domestic support policies
– trade policies
– disease outbreaks
– biofuel policies
Potential application
06/11/2017
38
112
Biofuel scenarios
Session 6
Biofuel applications and exercises
113
Bioenergy and Food Security
114
Biofuel Prices
BIOFUEL Demand
BIOFUEL Supply
Feedstock A Feedstock B Feedstock C Feedstock D
Crop Prices
Mandatory Blending
Total Energy demand
Price driven Price driven
06/11/2017
39
115
Ethanol model
Crop Area
Ethanol
Ethanol
Sugar Cane
“Sumol”
Molasses Raw Sugar White Sugar
Feed supply
Coarse grains
FeedFoodEthanol
DDG
Food Feed
Cassava
Second GenerationFood supply
Ethanol Market
116
Biodiesel Market
Crop Area Oilseeds
Meal Oil
Jatropha
Food
Food demandFeed demand
Biodiesel
Biodiesel
Biodiesel model
117
• Biofuels
• Ethanol
• Biodiesel
Country model sections
06/11/2017
40
118
• Produce an annual medium-term outlook for the biofuel
sector.
• Carry out in-depth analyses of the fundamentals
shaping the biofuel market and their impact on
agricultural markets and food security.
• Assess the impact of stated biofuel policy targets on
commodities markets.
• Evaluate the impact of second generation biofuels.
Biofuels
Objectives:
119
Ethanol:• Molasses
• Sugar cane
• Sugar beet
• Wheat
• Coarse grains
• Roots and Tubers (Cassava)
• Non Agricultural
• Energy crops and residues
(second generation)
Biodiesel:• Vegetable oils
• Jatropha
• Non Agricultural
Feedstocks:
Biofuels
120
• Transportation fuel use and retail prices (gasoline,
diesel)
• Technical conversion factors
• Taxes, tariffs, payments
• Blending requirements
Biofuels
Exogenous data:
06/11/2017
41
121
• Ethanol and biodiesel domestic production
• Feedstock use
• Ethanol and biodiesel consumption (fuel and other)
• Biofuel trade
• Ethanol and biodiesel domestic and world market
prices
Biofuels
Endogenous data:
122
Total Production = f(price of ethanol/average feedstock price,
policy variables)
Production by feedstock = f(average feedstock price/feedstock
price)
Feedstock demand = f(production by feedstock, conversion
factor)
Byproducts production = f(production by feedstock,
conversion factor)
Biofuels / Production
123
Ethanol/Biodiesel use in low blends (price driven) =
f(ethanol price/gasoline price, per-capita income, population)
Ethanol other use (non fuel) = f(ethanol price, income)
Mandated Ethanol/Biodiesel use = f(mandated blending,
gasoline/diesel use)
Total Ethanol/Biodiesel use = max (mandated, price driven
uses)
+ other use (ethanol)
Trade and domestic prices equations are same as other
commodities
Biofuels / Consumption
06/11/2017
42
124
Questions?
Scenario 1
Impact of ethanol production on domestic cassava price from 2009 to 2016
• Ex- post 2009-2016
• Assess the impact of domestic fuel blending mandate and ethanol production from cassava on the domestic cassava market between 2009 and 2016.
• The ethanol blending mandate intends to generate fuel use, but the envisaged 5% of total gasoline use have never been reached.
• It is assumed that no fuel consumption of ethanol occurs without a mandate.
• Scenario 1 removes the blending requirement and subsequently all ethanol production for fuel use.
Scenario 1: Setup
06/11/2017
43
0
50
100
150
200
250
2009 2010 2011 2012 2013 2014 2015 2016
Thn
ou
san
d C
ub
ic M
eter
s
Domestic Ethanol Market
Production Total Use Other Use Fuel Use Net Trade
Scenario 1: Background
Scenario 1: Background
Ethanol Gasoline Consumption
Blendingratio
Total Use Other Use Fuel use
Thousand cubic meters2009 86 81 5 5,501 0.09%2010 89 84 5 6,123 0.08%2011 96 86 10 6,003 0.17%2012 104 89 15 5,973 0.25%2013 122 91 31 6,179 0.50%2014 123 82 41 6,848 0.60%2015 163 97 67 7,255 0.92%2016 176 100 77 7,757 0.99%
The current blending mandate is 5% of gasoline volume sold.
Scenario 1: Background
0
50
100
150
200
250
2009 2010 2011 2012 2013 2014 2015 2016
Thn
ou
san
d C
ub
ic M
eter
s
Ethanol production by feedstocks
Cassava Molasses Sugar cane Other
06/11/2017
44
Scenario 1: Background
0
200
400
600
800
1000
1200
2009 2011 2013 2015
Tho
usa
nd
To
nn
es
Molasses
Feed Use
Use for Ethanol Production
0
200
400
600
800
1000
1200
1400
2009 2011 2013 2015
Tho
usa
nd
To
nn
es
Roots and Tubers
Industrial UseFood UseFeed UseUse for Ethanol Production
0
5000
10000
15000
20000
25000
2009 2011 2013 2015
Tho
usa
nd
To
nn
es
Sugar Cane
Sugar Production
Use for Ethanol Production
Use of ethanol feedstuffs
Scenario 1: Results
0
500
1000
1500
2000
2500
3000
3500
4000
0
50
100
150
200
250
300
350
400
2009 2010 2011 2012 2013 2014 2015 2016
Tho
usa
nd
To
nn
es
Tho
usa
nd
To
nn
es
Roots and Tubers market impact
Ethanol Production Baseline Ethanol Production ScenarioFood Use Baseline Food Use ScenarioNet Trade Baseline Net Trade Scenario
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
-0.50%
-0.45%
-0.40%
-0.35%
-0.30%
-0.25%
-0.20%
-0.15%
-0.10%
-0.05%
0.00%
2009 2010 2011 2012 2013 2014 2015 2016
Price Impact
Cassava Molasses
Scenario 1: Results
06/11/2017
45
• Cassava industry is export oriented and domestic price depends on global market.
• Ethanol production for biofuel only accounted for about 5% of domestic use in 2009, however the share increase to reach around 43% in 2016, still accounting for only 6-7% of production.
• Domestic price appeared only marginally elevated, which hardly impacted food or feed use.
• Molasses price drops significantly, but sugarcane price and production would be only slightly affected.
Scenario 1: Implications
134
Policy applications
Session 7
Biofuel applications and exercises
Impact of the ethanol blending ratio increase to 3.5% in 2018 and further
gradual raise to 5% by 2023
• 2017 – 2026
• Assess the impact of the introduction of a E5 blending mandate by the Ministry of Industry and Trade (MOIT) in January 2018 on the domestic and international markets for cassava and other relevant agricultural commodities.
Scenario 2
06/11/2017
46
• Two different paths are considered:
– Scenario 2.a assumes the enforcement of a blending ratio of 3.5% from 2018 throughout the projection period.
– Scenario 2.b assumes the introduction of a 3.5% blending ratio in 2018, its gradual increase to 5% over the following five years, and for it to remain at this level to 2026.
Scenario 2: Background
0
100
200
300
400
500
600
700
800
900
Tho
usa
nd
Cu
bic
Met
ers
Ethanol Production
Scenario 2.a Scenario 2.b Baseline
Scenario 2: Results
Scenario 2: Results
0
100
200
300
400
500
600
700
800
900
Baseline Scenario 2a& 2b
Baseline Scenario 2a Scenario 2b Baseline Scenario 2a Scenario 2b
2018 2023 2026
Tho
usa
nd
Cu
bic
Met
ers
Ethanol Market balance
Domestic Supply Total Demand Biofuels Use Other Use
Net Trade Exports Imports
06/11/2017
47
Scenario 2: Results
• Domestic ethanol production increases 75% in 2018, 98% and 154% in 2023 for Scenarios 2.a & 2.b respectively, and 95% and 154% in 2026 for Scenarios 2.a & 2.a
• Biofuels become the predominant use of ethanol in all the scenarios
• Other use remains unchanged
• Exports would decrease and imports would increase, however net trade would remain positive for all the scenarios
0
500
1000
1500
2000
2500
3000
3500
4000
4500
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
2008
20
09
20
10
20
11
20
12
20
13
20
14
20
15
20
16
20
17
20
18
2019
20
20
20
21
20
22
20
23
20
24
20
25
20
26
Tho
usa
nd
To
nn
es
Roots and Tubers Market
Scenario 2.a Net Trade Scenario 2.b Net Trade Baseline Net Trade
Scenario 2.a Demand Scenario 2.b Demand Baseline Demand
Scenario 2.a Domestic Supply Scenario 2.b Domestic Supply Baseline Domestic Supply
Scenario 2: Implications
Scenario 2: Implications
-1000
-500
0
500
1000
1500
2000
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026
Tho
usa
nd
To
nn
es
Molasses Market
Scenario 2.a Net Trade Scenario 2.b Net Trade Baseline Net Trade
Scenario 2.a Domestic Supply Scenario 2.b Domestic Supply Baseline Domestic Supply
Scenario 2.a Demand Scenario 2.b Demand Baseline Demand
06/11/2017
48
Scenario 2: Results
Change market balance in 2018 with respect to the baselineCassava Molasses
Domestic Supply 0% 2%Total Demand 34% 35%
Food Use 0% -Feed Use 0% -5%
Biofuels Use 76% 75%Other Use -1% -5%Net Trade -13% 422%
Stocks 32% 17%Producer Price 4% 32%
Scenario 2: Results
Changes in 2023 Scenario 2.a Changes in 2023 Scenario 2.bCassava Molasses Cassava Molasses
Domestic Supply 0% 3% 1% 2%Total Demand 48% 45% 76% 72%
Food Use 0% -1%Feed Use 0% -6% 0% -8%
Biofuels Use 103% 92% 160% 145%Other Use 0% -7% -1% -9%Net Trade -9% 836% -16% 1387%
Stocks 49% 43% 76% 67%Producer Price 2% 45% 4% 58%
Scenario 2: Results
Changes in 2026 Scenario 2.a Changes in 2026 Scenario 2.bCassava Molasses Cassava Molasses
Domestic Supply 0% 3% 1% 3%Total Demand 50% 44% 80% 72%
Food Use 0% 0%Feed Use 0% -7% 0% -9%
Biofuels Use 102% 86% 163% 140%Other Use 0% -8% -1% -10%Net Trade -10% 1215% -16% 2077%
Stocks 50% 42% 81% 70%Producer Price 2% 53% 4% 68%
06/11/2017
49
Scenario 2: Results
0100200300400500600700800900
Bas
elin
e
Scen
ario
2a
& 2
b
Bas
elin
e
Scen
ario
2a
Scen
ario
2b
Bas
elin
e
Scen
ario
2a
Scen
ario
2b
2018 2023 2026
Tho
usa
nd
Cu
bic
Met
ers
Vietnam ethanol domestic production by feedstock
Cassava Molasses Sugar Cane Others
0%
20%
40%
60%
80%
100%
Bas
elin
e
Scen
ario
2a
& 2
b
Bas
elin
e
Scen
ario
2a
Scen
ario
2b
Bas
elin
e
Scen
ario
2a
Scen
ario
2b
2018 2023 2026
Vietnam ethanol domestic production by feedstock: shares
Cassava Molasses Sugar Cane Others
Scenario 2: Results
0%
1%
2%
3%
4%
5%
6%
2018 2019 2020 2021 2022 2023 2024 2025 2026
World Prices Changes
Scenario 2.a Molasses Scenario 2.a Roots and Tubers
Scenario 2.b Molasses Scenario 2.b Roots and Tubers
Scenario 2: Implications
• The additional ethanol demand is expected to be supplied domestically, however the production would be partially based on imported feedstocks.
• The price increase would not significantly stimulate the cassava domestic production, as the market remains dominated by exports.
06/11/2017
50
Scenarios are based on assumptions related to:
• Market developments;
• Domestic policies;
• Trade policies;
• Weather conditions
Uncertainties concerning Scenarios
Summary
• Aglink-Cosimo system generates baseline projections and simulation analysis of policy alternatives
• Complex integration of biofuel sectors in agriculture depends on reliable policy and economic data and parameters
• Scenarios allow to trace out interactions and interdependencies across sectors and countries
150
Areas for potential cooperation
06/11/2017
51
151
• Data review
• Country module development
• Scenario work
Discussion for future co-operation
• Benefits– Participate in annual world commodity projection– Access to data, models, FAO staff – Participate in analysis projects– Access to capacity building – training– Tap global expertise
• Costs– Depends on commitment– 2-6 staff months– Provide data on markets and policies– Advice as to policy issues etc.– Input into projection– country questionnaire
Co-operators Program
• Annual Cycle – from November to April/May
• Questionnaire on historical data and projection judgements
• Participation in baseline process
• Meetings – OECD/FAO
• Model maintenance - research
Time requirements
06/11/2017
52
154
Questions?
Thank you
Visit our website:
www.agri-outlook.org
Thank you
Visit our website:
www.agri-outlook.org