Abstract
Risk concepts in financial planning models for pavement rehabilitation are discussed. Because risk is generally related to variance, in this case the variance associated with pavement performance, it is necessary to model performance in a manner that preserves statistical relationships. In that regard, a finite Markov Chain model of year-to-year movement in pavement ratings in constructed. The method in which risk- (variance-) related and probabilistic constraints can be imposed on an aggregate financial planning model is described. Such constraints permit budgets that are influenced by both average performance and variance in performance to be set.