We began writing PRF policies in the fall of 2012 for the 2013 calendar year. Our methods of optimization to arrive at an optimized set of intervals and the given percentages assigned to each interval have changed over the years. However, we have always optimized the selections and worked on providing the client with an alternative to what we call “spread and pray.” The logic among many agents is to market PRF at 90%, the highest coverage level available, and then “spread” the coverage out among the full year using six intervals at the same percentage. This type of “spread” is usually Jan-Feb at 16%, Mar-Apr at 17%, May-Jun at 17%, Jul-Aug at 17%, Sep-Oct at 17%, and Nov-Dec at 16%. This would have the client covered across the whole year and use 100% of the interval percentages. I used the 16% at the beginning (Jan-Feb) and end (Nov-Dec) of the year since many clients are hesitant to even use those intervals (without seeing the optimized results) and would not want any extra coverage assigned to them.
The logic behind this is rooted in the very beginning of PRF marketing. When PRF was introduced in 2007, the key was to have ranchers, who were not used to purchasing crop insurance, receive some type of payment to entice them to keep using the product year after year. There are times where an optimized solution may consist of only two intervals (4 months) or three intervals (6 months). Yes, if it is wet in those intervals and the final index is above the coverage level selected, that client won’t receive a payment. However, with the “spread and pray,” that same client might receive a payment in some other interval, but given the small assigned percentage to each interval, the payment won’t be very large. The very large negative to the “spread and pray” is that by using all the intervals to span the year, you are placing some coverage in very expensive intervals since each two-month interval has a unique rate.
Given our geographically diverse book of PRF policies, we wanted to compare our optimized interval solutions to the “spread and pray” method of interval selection. We went back to every different grid that we have insured through the years and compared the actual results from our optimized solution to what would have happened if that same premium would have been spent on a 90% coverage level policy with the intervals spread across the full year as discussed above. Before we get into the results, this analysis spans 94 grid-years with 53 unique grids and covers the 2013-2016 calendar years. A grid-year is a combination of intervals for a particular grid for a given year.
Over all the grid-years, the optimized solutions returned 123% of the “spread and pray” solutions. This is calculated by taking difference between the optimized indemnity and the “spread and pray” indemnity and then dividing that difference by the premium. This difference is then averaged across the 94 grid-years and converted to a percentage.
For example, if the premium was $100 and the indemnity was $150 for the optimized solution and $25 for the “spread and pray” solution, the calculation for a single grid-year would be:
Net difference: (150-25)/100 = 1.25 = 125%
Over the 94 grid-years, the optimized solution was better 73% of the time (69 grid-years). The “spread and pray” solutions returned at least some indemnity 88 of the 94 grid-years while the optimized solutions returned some indemnity 80 times. This would support the initial flawed marketing of PRF, however, the optimized solution resulted in enough indemnity to cover the premium 53 times which is 56% of the grid-years. The “spread and pray” solutions only covered the premium 37 times which is only 39% of the grid-years. So while the indemnities may have been slightly more frequent with “spread and pray,” the number of grid-years with zero out-of-pocket premiums is greater with the optimized solutions.