This paper presents a new non-parametric methodology for the description of the evolution of the asset cycle in the stock market. It uses the empirical distribution of the data; in particular the structures of the tails of return distributions to build Boom-Bust Indicators (BBI) that describe whether a given market is a bull or a bear. These indicators, for three different time horizons, perform better than the usual binary sequence of financial crises because they measure both direction and intensity, they have stronger variability than a binary variable, they are strongly associated to the original data and keep some of its underlying characteristics such as serial auto correlation, and they identify at least the same bull and bear markets as other methodologies. There is no evidence that favors one of the BBI specifications above the others.