Often, signaling research in the strategy and economics literature postulates the existence of an ostensible signal and then empirically tests its veracity, utilizing cross-sectional data. We argue that this static approach does not allow researchers to fully incorporate the concept of equilibrium in their analysis, thereby potentially violating a key axiom of signaling theory. We propose that a dynamic analysis of signals can address this omission, and then conduct such an analysis. We use empirical data on warranty coverage offered by automobile manufacturers in the U.S. market extending from the first warranty offered by the industry in 1960 through to 2008. Our findings support the notion that signaling behavior differs in periods of equilibrium and disequilibrium, in turn influencing signal accuracy. Copyright © 2013 John Wiley & Sons, Ltd.