What is the Black-Litterman Model
The Black-Litterman model is a model used to estimate inputs for portfolio optimization. It mixes different types of estimates, some based on historical data, others based on equilibrium conditions to arrive at updated estimates. The Mixed Estimation Model was developed by Henri Theil in the early 1960's, but was applied to financial data by Fischer Black and Robert Litterman in the early 1990's.
The beauty of this model is that one can blend a variety of views specified in different ways, absolute or relative, with a given prior estimate to generate a new and updated posterior estimate which includes all the views. The diagram below shows what the mixing might look like in a single dimension. The updated posterior estimate should be centered more closely around the unknown mean, and should also have a lower variance(higher precision) that either the prior or conditional distribution.