Measuring Uncertainty in the Stock Market

We propose a real-time index of time-varying uncertainty in the stock market. The index is constructed after removing the common variations of the series, taking into account recent advances in the literature, emphasizing the difference between risk (expected variation) and uncertainty (unexpected variation). To this end, we use data from 1926 to 2014 of 25 portfolios sorted by size and book-to-market value, which lowers considerably the information requirements and the modeling design costs, compared to previous proposals in the literature. Results show that, even when our estimates should be thought of as an uncertainty measure of the stock market (i.e. financial uncertainty), they perform very well addressing the uncertainty of the economy as a whole. This is apparent from comparisons with macro-uncertainty indicators and from estimations of the economy’s dynamics, after facing an uncertainty shock, measured by our index.