Individual forecasting behavior

Participants of a laboratory experiment judgmentally forecast a time series. In order to support their forecasts they are given a highly correlated indicator with a constant lead period of one. The subjects are not given any other information than the time series realizations and have to base their forecasts on pure eyeballing/chart-reading. Standard economic models do not appropriately account for the features of individual forecasts: These are typically affected by intra- and inter-individual instability of behavior. We extend the scheme theory by Otwin Becker for the explanation of individual forecasts by simple schemes based on visually perceived characteristics of the time series. We find that the forecasts of most subjects can be explained very accurately by only a few schemes.