Skip to main content
    Back to the glossary

    constraint

    Efficient Frontier

    The set of portfolios that offer the highest possible return for each level of risk.

    The efficient frontier is the set of portfolios that offer the highest expected return for each level of risk, or equivalently the lowest risk for each level of expected return. Plotted with risk on the horizontal axis and return on the vertical axis, it forms a curve that rises and bends to the right. Any portfolio below the curve is inefficient, because another mix offers more return for the same risk.

    Each point on the frontier answers one optimization: minimize portfolio variance, subject to the weights summing to 100% and the expected return reaching a given target. Repeating that across a range of targets traces the curve. The inputs are those of mean-variance optimization: expected returns, volatilities and correlations, plus constraints such as no short selling or position caps, which can only pull the frontier down and to the right. Adding a risk-free asset creates a straight line from the risk-free rate, the capital allocation line; the point where that line just touches the curve, the tangency portfolio, has the highest Sharpe ratio.

    Moving up the frontier buys more expected return at the cost of more volatility, and the curve flattens as you go, so each extra point of return costs more risk. The leftmost point is the minimum-variance portfolio. The frontier is only as good as its inputs, and because expected returns are hard to estimate, the true frontier is uncertain: a portfolio slightly below the estimated curve may be just as good in practice.

    Example with illustrative numbers: Portfolio A expects 7% with 12% volatility, Portfolio B expects 7% with 9% volatility, and Portfolio C expects 8.5% with 12% volatility. A is inefficient, because B delivers the same expected return with less risk and C delivers more return for the same risk. Only portfolios like B and C can sit on the frontier.

    On GNG, the Portfolio Optimizer's Mean-Variance, Black-Litterman, Target Return and Target Risk engines each aim for a point on an efficient frontier built from your universe, limits and estimates. Target Return looks for the lowest-risk mix that reaches your return floor, Target Risk for the highest-return mix inside your volatility ceiling, and Mean-Variance for the highest Sharpe ratio. After solving, GNG applies your single-name cap and minimum position size, and a failed solve falls back to simpler weights with a notice, so read the final Expected return and Volatility in the results. For universes of 200 names or more, the optimizer skips sampling the full frontier to stay fast and says so; the recommended portfolio is not affected.

    Mean-Variance Optimization is the method behind the frontier, and Target Return shows how one point on it is chosen.

    More constraint terms

    Get full access to GNG Research

    Create a free account to access portfolio tracking, advanced tools, and more.