The exposure curves of Bernegger's MBBEFD class, used to rate property
excess of loss treaties from a risk profile. mbbefd_curve() gives the
share of a risk's expected loss that falls below a deductible of x times
its sum insured (or maximum possible loss), for the parameters b and
g. swiss_re_parameters() gives b and g of the
one-parameter curves of Swiss Re, and swiss_re_curve() the curve itself:
c = 1.5, 2, 3 and 4 are the curves Y1 to Y4, from personal lines to large
industrial risks, and c = 5 the curve used at Lloyd's for industrial risks.
Arguments
- x
The deductible as a share of the sum insured: numbers from 0; those above 1 are taken as 1.
- b, g
The parameters of the curve:
bat least 0 andgat least 1. The arguments are recycled to a common length.- c
The parameter of a Swiss Re curve, at least 0, or the name of one: "Y1", "Y2", "Y3", "Y4" or "Lloyds".
Value
mbbefd_curve() and swiss_re_curve(): the curve at x,
between 0 and 1. swiss_re_parameters(): a data frame with the columns c,
b and g, a row for each value of c.
Details
With \(g = 1\) or \(b = 0\) every loss is a total loss and \(G(x) = x\); with \(b = 1\), \(G(x) = \log(1 + (g - 1)x) / \log(g)\); with \(bg = 1\), \(G(x) = (1 - b^x) / (1 - b)\); otherwise $$G(x) = \log\left(\frac{(g - 1)b + (1 - gb)b^x}{1 - b}\right) / \log(gb).$$ The probability of a total loss is \(1/g\). The Swiss Re curves have \(b = e^{3.1 - 0.15(1 + c)c}\) and \(g = e^{(0.78 + 0.12c)c}\).
A layer of limit excess of deductible on a risk with sum insured
\(M\) takes the share \(G((d + l) / M) - G(d / M)\) of the risk's expected
loss; see property_exposure_rating.
References
Bernegger, S. (1997). The Swiss Re exposure curves and the MBBEFD distribution class. ASTIN Bulletin 27(1), 99-111.
See also
Other reinsurance pricing functions:
burning_cost(),
credibility_weight(),
pareto_rating(),
property_exposure_rating(),
technical_premium()
Examples
# a layer of 4m excess of 1m on a risk of 10m, on the Y3 curve
swiss_re_curve(0.5, "Y3") - swiss_re_curve(0.1, "Y3")
#> [1] 0.3713214
swiss_re_parameters(c("Y1", "Y2", "Y3", "Y4", "Lloyds"))
#> c b g
#> 1 1.5 12.648011 4.220696
#> 2 2.0 9.025013 7.690609
#> 3 3.0 3.669297 30.569415
#> 4 4.0 1.105171 154.470015
#> 5 5.0 0.246597 992.274716
mbbefd_curve(c(0, 0.25, 0.5, 1), b = 2, g = 10)
#> [1] 0.0000000 0.5090424 0.7285969 1.0000000
