The stochastic channel models typically abstract away the details of the paths that carry energy in the radio channel. While these have been universally acceptable for decades due to their ease of use and reasonable accuracy in most practical cases, the appearance of steerable, narrow-beam antennas in mmWave bands makes the exact path information very valuable, primarily for beam tracking algorithms. Currently, only deterministic channel modeling (e.g. ray tracing) provides the required level of details, but at prohibitive computing cost. This limits the study and design environments for such algorithms to the confines of existing ray tracing data, which is bulky and rarely available for free. In this paper, we consider an approach to stochastic channel modeling that allows to achieve the level of details equivalent to ray tracing, but at a fraction of the computing costs. The proposed approach may be immediately applied to any system operating at 20-100 GHz. It allows the researchers and engineers to perform quick testing of elaborate mmWave MAC and PHY algorithms with a system-level simulation, without having to obtain exhaustive measurement or ray tracing data.