Each color in the graphs represents a certain type of asset. The relative size of each band of color represents the "weight" of that asset in the portfolio. The horizontal axes represent the risk along the efficient frontier. The composition of portfolios to the left of each graph has the minimum risk. The composition of portfolios to the right of each graph has the maximum expected return and a correspondingly higher risk. In short, the graphs display portfolios along the efficient frontier, from low to high risk, in terms of their portfolio weights.
The first graph, at left, displays the portfolio composition map of a classical efficient frontier, based on a universe of 500 stocks. Since there are no asset constraints (apart from non-negativity), the maximum expected return, or the right-most point in the graph, consists of one stock, which is represented by a column of the same color from bottom to top. These portfolios do not change.
The center graph displays one hundred simulated MV portfolio composition maps, made by entering the data used for the graph to the left into New Frontier's Optimizer. Note how remarkably different some of these composition maps are relative to the original classical efficient frontier.
The third graph, at right, displays the Michaud Resampled Efficient Frontier as the simulated frontiers from the center graph are averaged. This is the image from which we take our logo. Even with the small amount of resampling shown here, the Optimizer provides a more stable portfolio composition map.