For ridge regression the degrees of freedom are commonly calculated by the trace of the matrix that transforms the vector of observations on the dependent variable into the ridge regression estimate of its expected value. For a fixed ridge parameter this is unobjectionable. When the ridge...

Holly and Magnus [1988] show that the IV estimator in a linear equation is asymptotically as efficient as the ML estimator in the model that is obtained by "completing" this equation to a complete equations system. Their proof is long and involves many matrix manipulations. A simpler approach is...