insurers

Report
3rd West and Central Africa Agricultural Science Week
and 10th General Assembly of CORAF/WECARD
Ndjamena, Chad, 14th – 18 th May 2012
Index based crop insurance in West Africa :
principles, existing projects and prospects
L’Assurance agricole indicielle en Afrique de l’Ouest:
principes, premières réalisations et perspectives
Bertrand MULLER1, Moussa SALL4, Antoine LEBLOIS5, Alpha BALDE2,
Moustapha FALL3, Patrice KOUAKOU3 et François AFFHOLDER1
1
2
3
4
5
Introduction : constraints/risks limit productions
 Crop productions are limited by constraints and risks
 Many constraints(problems) and risks can be controlled or
prevented: good practices, inputs, organization …
 But residual risks: rainfall variability (droughts)
massive locusts and birds attacks
extremes temperatures, winds, floodings
 And those risks are generally extremely covariant : affect
numerous people at the same time
2
Quelea Quelea
Introduction : constraints/risks limit productions
 Soudanian (600-1200mm) and Sahelian (200-500mm) areas:
very important rainfall spatio-temporal variability
 Spatially, inter-annual and intra (seasonal) annual
Very variable (uncertain) start, dry-spells/droughts

 Sharp decrease in 1970 -> 1990 : 1st sign of CChange
 Increasing since 15 years …
Pluviométrie annuelle de Bambey (1923-2010)
1100
1923-2010
1923-1969
1970-1979
2000-2010
1000
Pluviométries 2007 - Dept. Diourbel - 26 postes
(AMMA DMN-CERAAS-CIRAD)
900
Pluviométrie annuelle (mm)
600
550
Pluviométrie (mm)
500
450
400
350
:
:
:
:
586,5
667,5
465,8
593,7
mm
mm
mm
mm
800
700
600
500
300
400
250
300
2007
2003
1999
1995
1991
1987
1983
1979
1975
Années
1971
1967
1963
1959
1955
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
1951
8
1947
7
1943
5 6
1939
4
1935
3
1931
2
1927
1
1923
200
3
Introduction : constraints/risks limit productions
 Risks and economic conditions affect productivity
 WA farmers are used to climate variability and CChange : aim
at securing medium/low yields and don’t want to loose
investments in bad years : “risk aversion” (pertinent)
 Risk aversion reinforced by economic conditions : low prices,
markets, organizations “value chains”, etc..
=> extensive systems : low productivities
=> food dependency
(some exceptions : irrigated rice ..)
 But African population will increase by 2,5 – 3
 And Asian production seems to not increase more …
 Agricultural productivities must increase to feed Africa
 Under climate change with probably more climate variability
4
Introduction : constraints/risks limit productions
 Agricultural (crops, animals) insurances appear as a
possible tools to cope with residual risks by paying
indemnities in case of low productions (losses)
 May secure part of farmers incomes
 May secure credit programs : more sustainable
 Mays allow access to credit for much more farmers
 May generate credit rate decrease
 May contribute to a virtuous circle conducing to more
investments (inputs, works) and productivity increase
 Also considered as an adaptation tool to CC and CV
 Interest of backers : many money is coming
 Very ”fashion” theme for many stakeholders …
and a new market for (re-)insurers
 Just starting in West Africa but 10 years in Asia, S.Am., E.Af.
5
Insurance and index based insurance principles
 Insurance is a service :
 the insurer pays compensations in case of bad production /
losses due to one(some) problem(s) (residual ones)
 the insured farmer must pay an annual premium
 nobody know when compensations (payouts) will come ..
 Premium prices depend on:
 compensations/payouts : statistical average (on time and space)
 service management costs : same for credit and all services
but insurance specificity is losses evaluation
 commercial margin : same for credit and all services
 re-insurance costs : to allow the insurer to be able to
compensate simultaneously numerous customers if necessary
… that is often the case in agriculture since risks are
extremely covariant
6
Insurance and index based insurance principles
 Damages/losses evaluation is difficult/costly, particularly in South
Countries where fields are small, disseminated, heterogeneous, etc..
risk of conflicts and “moral hazard”
 Index based insurances (since early 2000s)
 No direct damages/losses evaluation (at fields)
 Damages/losses indirectly assessed through the value of
an index (indicator) related to some measurements

For instance measures of temperature in one reference site
=> reduced cost
 Allow to insure several farmers of an area at the same time,
and/or to have group contract => reduced cost
 Additional ways to decrease costs
 Management linked with credit
 Mobile phones technologies (sms) for contracts and payments
7
Insurance and index based insurance principles
 Calibrated to protect investments and then credit systems :
not for production/profit losses: would be too costly considering
markets and risks frequencies (1 year out of 5-10)
 “Real insurance” (private sector) is a priori reserved to “intensive
(using inputs) and market linked productions” because farmers
(or other stakeholder) have to pay premium
 Cotton, peanut, maize, rice, vegetables …
 But could be partially subsided by Governments/Backers -> PPP
 Index system can be used also for “social protection system”
 Aggregated (average) yields index : “all-risks insurance”
 Specific index allow to link losses to only one specific risk
(whatever the other problems)
 Climate index : on temperature, rainfall measurements
 Satellite data index : NDVI, biomass, ET%...
8
Insurance and index based insurance principles
 Main problem: “basic risk” i.e. risk that index values and
thus payouts aren’t correctly linked to damages
 Depend on kind and quality of index
 On spatial variability of reference observed factor ..
 On variability of other conditions: soils, sowing dates, varieties
 Development of insurance requires efforts (time, money) to
explain and convince farmers and others stakeholders:
 difficult to install “confidence” (trust) since there is no
“insurance culture” and because of “commercial” aspects …
 Also some initial investments in technologies : secured
raingauges, satellite data, yields controls …
9
Insurance based on aggregated yields
 Payouts depend on average yield compared to a reference
yield level that is a fraction of the mean inter-annual yield
 Requires a very good and confident “yields measurements
Moyenne historique
Seuil protection 80%
Indemnisations 80%
Seuil protection 30%
Indemnisations 30%
300
400
200
200
100
0
-
pluies
rendements
Lin. rendements
20
06
Rdts moyens
600
20
04
2000 2002 2004 2006
400
20
02
1992 1994 1996 1998
800
500
20
00
1986 1988 1990
1 000
19
98
200
0
1 200
600
19
96
800
600
400
700
19
86
1 200
1 000
Evolutions rendements arachide et pluies Diourbel
rendements (Kg/Ha)
800
cumuls annuels (mm)
rendements et
indemnisations (Kg/Ha)
Exemple fonctionnement assurance sur rendements agrégés
1 600
1 400
19
94

19
92

19
90

19
88

system” : cotton, sugarcane, vegetable
Quite expensive
Difficult now with national statistics : spatial cover and “quality” ..
Better if also some control of practices at farmers fields
Problem if bad yields due to human factor such as bad fertilizer for
instance … or decrease in tendency
Lin. pluies
10
Rainfall index based insurance
 Based on rainfalls = “drought insurance” : most frequent
 Many kinds of rainfall index:
 Simple : total rainfall amounts but don’t perform well
 Most complex : simulated yield (or stress index) by crop model
 Intermediate (IRI, World Bank) and most used since it is
quite good and easy to explain to farmers and insurers :
composite index based on rainfall amounts on different phases
of crop cycles
 Whatever the index : its parameters must be precisely defined
based on agro-climato-economical analysis
 Attention to pure statistical index as “payout if observed value reach
percentile x%” : not recommended
 payouts not calibrated according to crop status
 induces differences in protection level between areas
11
Rainfall index based insurance
 Composite index:



Simulation of a “virtual crop cycle” that starts within a recommended
sowing period according to a reference rainfall value (20 mm)
Fixed cycle and fixed phases (2 to 4) considered for insurance
“Trigger/Strike” and “Exit” reference rainfalls values for each phase to
pilot payouts according to rainfalls during the phases
Déficit Pl. (mm)
PHASE 1
Semis et installation
Déficit Pl. (mm)
PHASE 2
Croissance et Floraison
Déficit Pl. (mm)
PHASE 3
Développement Fruits
Calendrier Cultural
Démarrage Période
Indemnités = min (Coût prod. Total ; Somme paiements phases 1 + 2 + 3)
12
Rainfall index based insurance
 Several parameters in the contract
Peanut 90
Fonctionnement contrat arachide qualité Paoscoto-Nioro
First Dekad of Sowing Window
Last Dekad of Sowing Window
Sowing Trigger for Contract (mm)
Dekadal Cap (mm)
Phase 1 Start (dekad)
Phase 1 End (dekad)
Phase 2 Start (dekad)
Phase 2 End (dekad)
Phase 3 Start (dekad)
Phase 3 End (dekad)
Phase 1 Trigger (mm) (/4mm)
Phase 1 Exit (mm)
Phase 2 Trigger (mm) (/5mm)
Phase 2 Exit (mm)
Phase 3 Trigger (mm) (/4,5mm)
Phase 3 Exit (mm)
Total Insured Production Costs (FCFA)
Insured Prod. Costs Phase 1 (FCFA)
Insured Prod. Costs Phase 2 (FCFA)
Insured Prod. Costs Phase 3 (FCFA)
19
21
20
80
2
2
3
5
6
8
10
0
120
60
80
30
100000
70000
100000
100000
Indemnisations (FCFA/Ha)
Contract Parameters
180 000
1,00
160 000
0,90
140 000
0,80
0,70
120 000
0,60
100 000
0,50
80 000
0,40
60 000
0,30
40 000
0,20
20 000
0,10
Indemnité phase 1
0,00
Indemnité phase 2
Indemnité phase 3
Année
Série4
RDT Stat Nioro
 Main problems : potential “basic risks” due to
 Rainfall spatial variability
 No good time synchronization between contract functioning
13
and farmers crops
INDEX RDT
Nioro
How to improve “rainfall” index based insurance
 Favorable environment promoting homogeneity and
adequate practices limiting insured risks (sowing date,
variety) and other problems (diseases, etc..)
 Index are calibrated and perform well for “good practices”
 Losses due to other factors are not considered
 Mathematical solutions to partially limit basic risks:
 Use of majored dekadal (10 days) rainfall amounts
 Index related to several raingauges
14
How to improve “rainfall” index based insurance
 Most recommended : use of geospacialized data from
satellites : pixelised rainfalls or relative evapotranspirations
ET% (or mix), controlled/calibrated using some ground
observations
 But researches are required to assess
accuracy of those methods and of
their pertinence for crop insurance
Agrhymet - EARS project
 IFAD/WFP/AFD project
 Will be required also for extension/upscaling

EARS figure
 Already on-going pilot projects in Mali, Burkina and Benin by
PlaNet Guarantee and EARS, using 3km x 3km Meteosat info
15
How to improve “rainfall” index based insurance
 Other points to consider to improve insurance systems
in the future
 “Personalized” contract considering sowing date (within
sowing window): info transmitted by mobile phones
 “Sophisticated index”: simulated yield (or ET%) by crop
model : more accurate and will allow to consider also overrained period and/or integer other factors (Temp)
 Because farmers who invest need precision / quality : in focus
groups and meetings they ask very pertinent questions about
index functioning
 And since technologies allow (will allow) to do it
16
Index based insurances experiences in West Africa
 Up to now only in Burkina Faso and Mali : on cotton and maize, less than
1000 farmers in 2011 ; PlaNet Guarantee / GIIF / EARS
 20.000 farmers expected in 2012 …
 Pilots will be implemented in 2012 in Senegal on maize and peanut :
PG/GIIF+WorldBank+Cirad : 1000 farmers (??)
 Pilots will be implemented in 2012 in Bénin on maize ; PG/GIIF
 Different studies : Ghana (GTZ), Bénin (WB), Cameroun, BOAD
17
What we can learn from experiences
 USA : fully subsided (EU : prices are subsided ..)
 India : different programs sustained by Government which
subsides premium : more than 20 million farmers
 Positive for credit and allow government to help small farmers
 Malawi, other projects: strong investments
 GRET (2011) : “insurance programs could be economically
profitable but require at the beginning many investments
from Governments or Backers”
 feasibility studies (experts)
 equipments/technologies
 explanation/information, capacity building …
 Subsides generally required for small farmers
18
What we can learn from experiences
 When there is no subside (Mali, BF) : strategies consist in
proposing very cheap products to launch insurance culture :
but they poorly protect farmers … is that pertinent?
 Subsides may also contribute to create equity between areas
 Senegal : subsides higher for Northern region in order to have
same premium and protection everywhere
 Better to consider insurance in agricultural policy
 Basic risk is a problem
 IFAD/WFP (2010) : “The future largely depends on how the
industry will be able to expand the technology frontier”
(satellite and communications technologies)
19
Conclusions
 Insurance could contribute to enhance productivity by
securing credit (at least)
 Index based insurances are quite cheap and thus allow
providing insurance to much more farmers
 Major technical issues to be addressed to reduce basis risk
and improve quality of insurance products : satellite and
mobile phone technologies and crop models also
 Satellite technology also necessary for extension
 Insurance development requires expensive investments at the
beginning, including eventual premium subsides
 Subsides/other mechanisms are also required for equity
 Insurance must be considered in agricultural policy
 Government must also regulate the sector (legal issues)
Recommendations
 Let be optimistic : crop markets and agricultural policies in
Africa will allow insurance development ..
 Researchers must work on technical issues and also
on economical and policy ones
 what are the adequate Private-Public-Partnership ?
 how to integer agricultural insurances and index based
insurances in global food security management system??
 Researchers and development stakeholders must participate
to insurance projects / feasibility studies
 to technically help them
 to ensure transparency, equity and balance between
development and commercial issues
 to capitalize experiences in order to advise Governments
Thank you very much
Bertrand Muller
[email protected]
With Moussa SALL (ISRA-BAME), Antoine LEBLOIS (CNRS-CIRED),
Alpha BALDE (AfricaRice), Moustapha FALL (ISRA-CERAAS), Patrice
KOUAKOU (ISRA-CERAAS) et François AFFHOLDER (CIRAD)
« l’assurance agricole est un sujet trop sérieux pour être confiée
aux seuls assureurs, ré-assureurs et acteurs du crédit et de la
micro-finance »

similar documents