Bornhuetter-Ferguson reserves whose expected loss ratio is estimated from the triangle: the latest claims over the premium already earned by development (the used-up premium).
Arguments
- triangle
A triangle of cumulative claims, as for
chain_ladder, or the result ofchain_ladder(), whose development pattern is then used.The premium (or other exposure) of each origin.
- decay
The weight of an origin one period away in an origin's loss ratio, from 0 to 1 (the default, the classical Cape Cod with one loss ratio).
- ...
Arguments passed to
chain_ladder()for the development pattern, such asaverageandtail, whentriangleis a triangle.
Value
An object of class netsimr_reserve, as for
bornhuetter_ferguson, whose loss ratios are the estimated ones.
Details
The expected loss ratio is
\(LR = \sum_i C_i / \sum_i P_i p_i\), with \(p_i = 1 / CDF_i\) the share
developed of origin \(i\), and each origin's ultimate is
\(C_i + (1 - p_i) P_i LR\). With decay below 1 (the generalised Cape
Cod of Gluck), each origin has its own loss ratio, in which origin \(k\)
weighs \(decay^{|i - k|}\), so that nearby origins count most; with decay 0
the method is the chain ladder.
References
Stanard, J. N. (1985). A simulation test of prediction errors of loss reserve estimation techniques. Proceedings of the Casualty Actuarial Society 72, 124-148.
Gluck, S. M. (1997). Balancing development and trend in loss reserve analysis. Proceedings of the Casualty Actuarial Society 84, 482-532.
Examples
premium <- c(50000, 52000, 54000, 56000, 58000, 60000, 62000, 64000, 66000, 68000)
cape_cod(raa_triangle, premium)
#> Cape Cod
#>
#> origin latest premium loss_ratio developed prior_ultimate ultimate ibnr
#> 1981 18,834 50,000 0.387 1.000 19,350 18,834 0
#> 1982 16,704 52,000 0.387 0.991 20,124 16,888 184
#> 1983 23,466 54,000 0.387 0.974 20,898 24,002 536
#> 1984 27,067 56,000 0.387 0.943 21,672 28,302 1,235
#> 1985 26,180 58,000 0.387 0.905 22,446 28,311 2,131
#> 1986 15,852 60,000 0.387 0.813 23,220 20,197 4,345
#> 1987 12,314 62,000 0.387 0.694 23,994 19,662 7,348
#> 1988 13,112 64,000 0.387 0.546 24,768 24,359 11,247
#> 1989 5,395 66,000 0.387 0.336 25,542 22,349 16,954
#> 1990 2,063 68,000 0.387 0.112 26,316 25,429 23,366
#> Total 160,987 590,000 0.387 228,334 228,334 67,347
#> ultimate_loss_ratio
#> 0.377
#> 0.325
#> 0.444
#> 0.505
#> 0.488
#> 0.337
#> 0.317
#> 0.381
#> 0.339
#> 0.374
#> 0.387
cape_cod(raa_triangle, premium, decay = 0.75)$by_origin$loss_ratio
#> [1] 0.3953039 0.3978196 0.4091580 0.4145372 0.4073526 0.3876962 0.3734595
#> [8] 0.3673637 0.3576607 0.3551429
