01
Step one: read the axes and the curve
Risk, usually measured as volatility or standard deviation, sits on the horizontal axis; expected return sits on the vertical. Each dot is a possible portfolio. The upper-left boundary of the cloud is the frontier: for any risk level, no available mix offers a higher expected return. Points below the curve are inefficient, since the same risk could earn more or the same return could carry less risk. Remember that expected return and volatility come from assumptions about the future, usually estimated from history, so the curve is a model of choices rather than a forecast of results.
02
Step two: place your current portfolio
Plot the portfolio you hold today. Take an example with 80 percent equities and 20 percent bonds that sits below the curve at moderate risk. That position suggests a better-diversified mix might offer higher expected return for the same volatility, or similar return with less. Look at the vertical distance to the curve and the horizontal distance to it, because they describe two different improvements. Then check which assets drive the gap. Concentration in a single sector, currency or issuer often explains why a portfolio falls short of the boundary, and it points to the cheapest change to test.
03
Step three: choose a point and test its stability
Pick a target on the curve by asking how large a loss you could tolerate, not by chasing the highest return. In the example, the investor might select a point near current volatility and review the weights it implies. Then test stability: change return assumptions slightly, use a different historical window and watch whether the weights swing sharply. Large swings mean the frontier is fragile and should be treated cautiously. Finally add constraints such as position limits, taxes and costs, and document the reasoning. A professional adviser should review the result before any real allocation decision is made.