Understanding the Trend Adjusted APH Yield Endorsement Bruce J. Sherrick University of Illinois September 12, 2013 Mankato, MN TA-APH Yield Endorsement Originally Sponsored by Illinois Corn Growers Research and technical support provided by University of Illinois and ifar 1
Trend Adjusted Yield Endorsement Problem addressed: Corn and soybean yields are increasing through time APH used in crop insurance guarantees is simple average of past yields (up to 10) Average of past trending yields lags actual expected yield, which is what APH is meant to measure INTRODUCE: Trend Adjusted APH yield Trend Adjusted APH: Main Idea Ave APH APH Window Insurance Year 2
Trend Adjusted APH: Main Idea TA APH Ave APH APH Window Insurance Year TA APH Endorsement Applies to most APH-based yield and revenue policies (except organic, transitional, specialty) Does not apply to CAT Does not apply to written agreements except High-risk and UA 2012 Pilot, 14 states, selected counties 2013 expansion to larger area; added wheat, canola, cotton, grain sorghum, and rice 2014 includes sunflowers, barley, peanuts Future expansions planned for other crops ranked primarily by insured value and trend rates 3
TA APH Data Eligibility Must be in eligible county/crop (has trend in ADM) One or more actual yield in last four years Make election by sales closing date (March 15) Election made on crop/county basis Continuous unless cancelled or terminated by FCIC Transferred policies must re-elect TA APH Full Adjustment Must have at least 4 actual yields to be eligible for full adjustment Percentage of full trend adjustment: 4 or more yields in last 12: 100% 3 yields 75% 2 yields 50% 1 yield 25% Max trend APH yield can not exceed highest actual yield plus trend (e.g. max actual 190, TA =2, max TA APH = 192) 4
Yields Counting Toward Actual For 1 in 4 determination A -- actual yield AY A but less than 60% T NA AY did not elect PA prorated production DA duplicate actual prod NW -- wgt avg PW wgt. avg WY wgt. avg AX simple avg P assigned yields J temporary yield Receive adjustment A AY NA PA DA NW PW WY Details Cups and Floors do not apply to TA APH Rare case could result in better floors than actual yields adjusted for trend T (and S,E,N) yields not adjusted Yield reductions Excessive yield reductions still apply Reductions in approved APH still apply Other restrictions intended to encourage additional accumulation of data 5
TA APH yield Applicable to Enterprise, Basic, and Optional units Does not impact yield database TA-APH calculated for each database Trend rates updated annually Trend Adjustment Examples Generated with farmdoc Crop Insurance Decision Tool Microsoft Excel spreadsheet, annual release, calculates premiums, what-if scenarios, historic performance and other features of current programs. Available for download in FAST section of farmdoc website. 6
/cropins/index.asp 10-Year Consecutive All Actual 7
10-Year Consecutive 2 nonactual yields 10-Year Skip All Actual 8
4-year Not all actual Establish Trend Rates Based on NASS county and Crop Reporting District yields from 1976 though 2012 in 2014CY. County data screens for number of yields, acreage and other quality conditions. Fit trends through data with controls for: Weather Geographical weighting (spatial weights) Trend rates are capped at 90 th percentile or lower based on crop, and statistical properties 9
Trend Rates 2014 - Corn Trend - bu./acre 0.00 to 0.88 0.88 to 1.25 1.25 to 1.60 1.60 to 2.00 2.00 to 2.36 No data Trend Rates 2014 - Soybeans Trend - bu./acre 0.00 to 0.26 0.26 to 0.40 0.40 to 0.46 0.46 to 0.50 0.50 to 0.57 No data 10
Trend Rates 2014 Spring Wheat Trends bu./ac - to 0.20 0.20 to 0.40 0.40 to 0.60 0.60 to 0.80 No data Trend Rates 2014 - Canola Trend - lbs./acre 0 to 27 27 to 33 33 to 37 37 to 40 40 to 42 No data 11
Premium effects of TA Total premium is the same for identical rate yield guarantee given APH or TA-APH Use of TA-APH will lower farmer-paid premium at given yield because subsidy levels are lower for higher coverage levels Always to advantage to elect TA endorsement if not on floor Risk Subsidies (Percent of Total) Coverage Level Basic, Optional Enterprise 50% 67% 80% 55% 64% 80% 60% 64% 80% 65% 59% 80% 70% 59% 80% 75% 55% 77% 80% 48% 68% 85% 38% 53% Example of Actual Premiums, Corn APH Yield = 178.0 TA APH Yield = 189.9 Coverage Bu. Trigger Premium Bu. Trigger Premium bu/acre $/acre bu/acre $/acre 50% 89 0.51 94.5 0.69 55% 97.9 0.78 104 1.00 60% 106.8 1.22 113.4 1.82 65% 115.7 2.01 122.9 2.97 70% 124.6 3.26 132.3 4.61 75% 133.5 5.74 141.8 8.16 80% 142.4 11.61 151.2 15.26 85% 151.3 22.82 160.7 29.08 12
TA-APH Q/A and Summary Ongoing efforts by RMA to reflect actual yield within insurance contracts Intended to create incentive for accurate data/reporting. No impact on premium structure, only on amount of insurance Applies to actuals in added land, not to SA-T. PTY not trended. No limit on age of records. TA-APH Q/A and Summary Each database treated separately. Max TA does not force lower than APH in cases where T-yields are higher than actual. 2012 yields are included in estimation of 2014 trends. Weather adjustment mitigated influence on trend. Trend E(level of APH). Trend rates can be high where yields are lower. Always room to add Good ideas. 13
Questions? Thanks! Feel free to email questions/comments to: sherrick@illinois.edu Visit us on the web at: www.farmdocdaily.illinois.edu Acres Covered by Crop: 14
Growing Program: Total Premium Total Liability (amount of insurance) 15
Total Payments by crop through time Total Payments by crop through time 16
Ratings System Explained in a Nutshell Rating System (Overly)Simplified Based on a Loss Cost Ratio (LCR) system initiated in 1980s for a single product (65% yield) fixed indemnity-price policy. Modified for product types, extended to several dimensions, numerous internal controls, tied to exceptionally complex data through time, but still Idea each year t : Losses/liability = rate t then ave(rates) liability = premiums. Over time, premiums should equal losses. Loss ratio target =1 Main components used as rate components: farmer risk relative to county, reference yield, exponent, coverage level differential, and loads for CAT, PP, RP, and QA; and price level, vol. & deviates (correlation) for RP related. 17
Ratings system - so how are we doing? Recall Basic Idea: n Loss Ratio Should be equal to 1.0 if rates are correct, rates correct if losses correctly converted to rates. Should have no discernible patterns across geography or crops 18
19
20
21
22
23
24