2. Basis Risk: The primary hurdle
Imperfect relationship between index & targeted loss
Distance from the settlement metrological station
Differing scales of risk faced by insurer & farmers
Time averaged meteorological indices makes farmers
responsible for relating them to production losses 2
3. 3
Exogenous Meteorological Indices
Not adjusted to farms yields (cumulative rainfall)
Extremely easy to design
Carries high basis risk
Yield Tailored Meteorological Indices
Designed using regression of yield on weather pattern
Maximum average risk reduction.
More than one weather parameter can be used
4. 4
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 2 4 6 8 10 12 14 16 18
waterusgae
week after emergence ->
Av corn water use based on maximum daily air temp.
50° - 59°
70° -79°
90° -99°
Water Stress based loss models are the good aids for
their ease of use.
5. 5
Risk assessment factors at a development stage
Vulnerability to a particular weather peril
Prevalence of the said weather peril
Microclimate factors
Corn Soil Moisture Stress Criteria
Growth Stage Period Vulnerability
Emergence – Onset of
tassels
40 days Can withstand up to 60% soil
water depletion
Tassel onset - Blister
kernel stage
40 – 80 days Root zone not more than 50%
water deficient
Post Blister kernel stage > 80 days Can withstand 60% water
depletion with out yield reduction
6. Growth stage
Evapotranspiration
(inches per day)
Percent yield loss per
day of stress
(min-ave-max)
Seedling to 4 leaf 0.06 ---
4 leaf to 8 leaf 0.1 ---
8 leaf to 12 leaf 0.18 ---
12 leaf to 16 leaf 0.21 2.1 - 3.0 - 3.7
16 leaf to tasseling 0.33 2.5 - 3.2 - 4.0
Pollination 0.33 3.0 - 6.8 - 8.0
Blister 0.33 3.0 - 4.2 - 6.0
Milk 0.26 3.0 - 4.2 - 5.8
Dough 0.26 3.0 - 4.0 - 5.0
Dent 0.26 2.5 - 3.0 - 4.0
Maturity 0.23 0
Estimated corn evapotranspiration & yield loss per stress
day during various stages of growth.
7. s7
The index is triggered when a pre defined weather
event occurs which is potentially damaging.
A good index should capture all the damage points.
Should account for the demanded protection by farmers
The triggers for a product will depend on the index
type. Eg. In a water deficit index distribution of
rainfall is more important consideration than the
total volume
9. Premium
= Expected Loss + Risk Margin + Admin Charges
9
Expected Loss is the average payout for the product in any
year
Risk Margin is to compensate the insurer for taking the risk of
unexpected payouts
Administrative charges like marketing cost, taxes, reinsurance
& brokerage charges.
10. 10
Pricing using Burn Analysis
Straightforward & simple. Good for 1st look.
Few assumptions, hence lesser error points
Low level of accuracy
Pricing using Index Modeling
More complex.
A distribution is fitted to underlying weather indices
To be preferred when long data series is available.
Modeling error can lead to greater level of inaccuracies
11. 11
Risk Margin: Dependencies
Trends in weather pattern
Missing historical weather data
Unprecedented weather events
Current pricing methodologies uses VaR of the
contract to determine the risk margin.
12. 12
“Actual” Premium depends on many factors
Actuarial premium
Economic profile of the target farmers
Marketing Cost
Marketing cost comprise almost 10-15% of
premium in retail sales.
Willingness to purchase weather insurance
Deductibles
Insurance schemes associated with contract
farming operations have lower rates.
13. 13
The promise of Weather Insurance
Fast claim settlement
Independently variable loss data
Hassle free claim settlement process
Roadblocks to efficient claim handling
Settlement Data issues
Administrative inefficiency
Very few bank account holders amongst the farmers
14. 14
Settlement Data Issues
Failure of primary weather stations
Inadequate indemnity when compared to actual loss
Primary St.
Secondary St.
15. 15
Admin. Inefficiencies in claim handling
3rd party observers take time to release verified data
Data is processed at multiple levels
Admin issues (claim verification, cheque printing)
Cheque distribution to individual farmers
16. Obstacles to Acceptability
Complex models can be considered opaque by the farmers
The farmers may view the process to be vulnerable to
manipulation
Farmers perceive premium amount lost in case of no payout.
Needed a balanced trade off between basis risk
and farmer’s ease of understanding.
17. Obstacles to a Sellable product
Poorly trained ground staff often unable to explain the
product to farmers
Limited numbers of “Trusted” Point of Sales.
Below par post sales service.
Needed a balanced trade off between basis risk
and communication challenges. 17
18. Handling Product Complexity
Split product according to multiple weather variable for ease
18
Lesson Learned from India
Handling Trust Issues
Local agro-input dealers, cooperatives, NGOs as POS
Handling Service Quality Issues
Daily weather forecast, actual data & agro-advisory messages
19. 19
GIS in weather index insurance
Vulnerability map of a region
Can help cut down location specific basis risk
Can help in claim settlement process
Rainfall & topography to calculate runoff
Usage in loss calculation
To model evapotranspiration for arid regions
Flood Inundation
20. 20
TERMSHEET FOR WEATHER INDEX INSURANCE FOR LAC
Peril Type 1 Damage due to high fluctuation in Max temperature
Peril Period 1 Feb 1 to Feb 28
Varying Temp Index (VTI) Cumulative value of deviation in daily temperature over the Peril period
VTI Strike 35
Notional 2%
VTI Index ( HDD - Strike ) * Notional
Peril Type 2 Damage due to high temperature
Peril Period 2 Mar 16 to Apr 15
High temp Index (HTI) HDD of Max temp above 37* C
Loss rate in % HDD
1st 15 days 2nd 15 days
15 4% 2%
25 8% 4%
Loss Cap 70% 30%
Index VTI + HTI
Index Threshold 30
Notional (Rs./unit) 20
Max Sum Insured 900
Premium (incl Service Tax) 125
21. 21
TERMSHEET FOR WEATHER INDEX INSURANCE FOR SALT PRODUCTION
Risk Index Excess Rainfall
Wet Spell Strike 2 consecutive days of rainfall greater than 15 mm
Wet Spell Exit 3 consecutive dry days
Policy Period
Phase 1 Phase 2
10 Jan – 15 Mar 16 Mar – 31 May
Index_1 (Production losses due to wet spell)
Loss index in %
15mm < Rain ≤ 50mm 3% 3%
50mm < Rain ≤ 100mm 8% 8%
100 mm < Rain 15% 25%
Index_2 (consequential losses for each rainy day)
Extra loss for each rainy day (Eloss) 0.40% 0.80%
Index_2 = Wet Days Count X Eloss
TLI { Index_1 + Index_2 - 25% } Notional (amount paid for each point of TLI)
0 < TLI ≤ 25 50
25 < TLI 105
Payout = TLI X Notional
Sum Insured = Rs. 6500 Premium = Rs. 800
22. 22
Excess Rainfall Index (ERI)
Stages Initial Phase Dev Phase Mid Phase End Phase
Dates 15 Jul – 13 Aug 14 Aug – 17 Sep 18 Sep – 22 Oct 23 Oct – 21 Nov
Excess Rainfall Index
ERS 1 150 150 125 100
ERS 2 25 25 15 15
ERC 1 Cumulative rainfall of two consecutive days – ERS1
ERC 2
If ERC_1 > 0, sum of rainfall of subsequent days till rainfall < ERS2 for two
consecutive days
Payoff ∑ {(ERC 1 + ERC 2) PHASE X Notional PHASE }
Maximum payoff 2000
TERMSHEET FOR WEATHER INDEX INSURANCE, COTTON (AP)
23. 23
Water Deficit Index (WDI)
Stages Initial Phase Dev Phase Mid Phase End Phase
Precipitation Weight 1.5 1.5 0.5 NA
Index Strike 400
Notional (Rs.) 4
Payoff formula Max [0, (Index Strike – Actual Index) X Notional]
Max Payoff 2000
Loss Calculation
Deductible Rs. 200
Max S.I. 4000
Total Payoff Max (4000, payoff for ERI+ payoff for WDI – Deductible)
Premium 400