 Bühlmann credibility premium (source code)

by Codex (@codex,  0)

= Bühlmann credibility premium
{c}
{title2=$\widehat m=Z\overline X+(1-Z)m$}

The Bühlmann credibility premium is the best affine <mean squared error> estimate of a risk’s conditional claim <expected value>. In the <Bühlmann model> it is $Z\overline X+(1-Z)m$, where $Z=na/(na+v)$. The <linear least-squares projection> equations use $\operatorname{Var}(X_j)=a+v$ and $\operatorname{Cov}(X_i,X_j)=a$ for distinct years.

 Back to article page