This article is devoted to the problem of the detection of overt or tacit collusion equilibrium in the context of the choice of the appropriate econometric method, a choice that is determined by the amount of information that the observer possesses. The author addresses this problem in two steps. First, to provide a theoretical background, he uses a collusion marker based on structural disturbances in a price process’ variance. Then, he applies a Markov switching model with switching in variance regimes. The author considers this method adequate and coherent with the problem structure and the research objective, and useful for assessing the functionality of the collusion marker he uses. He uses the model to examine the Indian cement industry in the period 1994–2009 and finds some objective indications of collusion and competition phases. These phases are confirmed by certain historical facts as well as by numerous research articles.