
497Pricing Multiple Lines of Business and Risks
the pth percentile of Y. Hence, the percentile values can be paired with straight lines that
go from one variable to the other.
Figure 28.5b illustrates the case where the two distributions are completely independent,
in which case the sample pth percentile of X can be matched with any percentile of Y with
equal probability. As a result, the dots representing the percentiles of X and Y seem to be
connected at random.
Figure 28.5c shows that if the two distributions are completely dependent and discor-
dant (e.g. if X and Y′ are two shares of a quota share arrangement, X and Y = Y′ will be
fully d