Abstract
Uncertainty shapes the trajectory of business cycles and remains a central research topic in Macroeconomics. When studying the impact of uncertainty on the economy, economists use different uncertainty measures. While all indicators approximate uncertainty along some certain dimension, none of the indicators directly captures Knightian Uncertainty. According to Knight, uncertainty represents a situation in which it is no longer possible to form expectations about the future. In this study, we propose a method to directly measure Knightian Uncertainty. Our approach relies on rm-level data and measures the share of firms that are not able to formalize expectations about their future demand. We construct the Knightian Uncertainty indicator for Switzerland and show that the indicator is able to identify times of high uncertainty and detects uncertainty shocks well. We further evaluate the indicator by comparing it to established uncertainty measures. We find that most other indicators are weakly, but statistically significantly correlated with Knightian Uncertainty.