Speaker
Dr
Tobias Fissler
(University of Bern)
Description
Competing point forecasts for functionals such as the mean, a quantile, or a certain risk measure are commonly compared in terms of loss functions. These should be incentive compatible, i.e., the expected score should be minimized by the correctly specified functional of interest. A functional is called *elicitable* if it possesses such an incentive compatible loss function. With the squared loss and the absolute loss, the mean and the median possess such incentive compatible loss functions, which means they are elicitable. In contrast, variance or Expected Shortfall are not elicitable.
Besides investigating the elicitability of a functional, it is important to determine the whole class of incentive compatible loss functions as well as to give recommendations which loss function to use in practice, taking into regard secondary quality criteria of loss functions such as order-sensitivity, convexity, or homogeneity.
Primary author
Dr
Tobias Fissler
(University of Bern)
