What is max-Sharpe portfolio optimization?
Max-Sharpe optimization picks portfolio weights that give the best return per unit of risk. It needs two inputs: how the holdings move together, and what return you expect from each one. Because the expected returns are yours rather than measured, the answer is only as good as those estimates.
How it works
Every long-only portfolio you could build from a set of holdings has an expected return and a risk. Plot them all and the good ones form an upper edge. The max-Sharpe portfolio is the single point on that edge with the steepest reward-per-risk trade-off from the risk-free rate — the mix that buys the most expected return for whatever volatility it takes on.
The risk half of the input is estimated from price history: how volatile each holding is and how the holdings move together. Estimating that from a limited sample is noisy, so the estimate is shrunk toward a structured target before it is used, which stops the optimizer from chasing a correlation that was an accident of the window.
The return half cannot be estimated the same way, and this is the honest limit of the method. Past average returns are a famously poor forecast of future ones, so the builder here does not invent them: max-Sharpe unlocks only once you have supplied your own expected return for every name in the book. That makes the output a consequence of your view rather than a prediction dressed as arithmetic.
Optimizers are sensitive to their inputs — small changes in an expected return can move weights a lot — which is why the risk-only allocators sit beside this one. Minimum-variance and equal risk contribution need no return view at all, and are the right tool when you do not have a confident one.
Where you see this in the terminal
Portfolio builder(account required)enter an expected return per name and the max-Sharpe scheme unlocks beside the risk-only ones
PORT · Risk tab(account required)the max-Sharpe point marked on a saved book's frontier, once you have entered expected returns
Risk-based allocatorsthe three schemes that need no return view, and what each one assumes
The conventions behind it
What this page explains in plain language, the methodology documents state as conventions: the inputs, what is rejected, and the stated limits.
More of these: every concept page. The full transparency index is at Methodology.