The analysis is done.
Now what will you do?
The numbers you just saw are not answers by themselves. Two investors looking at the same analysis can come to entirely different conclusions — rebalance, add a holding, or simply wait. This page helps you make that decision.
Why look at relationships, not individual holdings
Harry Markowitz's insight that earned him the 1990 Nobel Prize in Economics is surprisingly simple. Don't ask which stock is good; ask how stocks move together. If one falls when another rises, holding both together reduces risk on its own — without giving up return.
Your analysis result dissects your portfolio from exactly this perspective. Volatility, Sharpe ratio, the efficient frontier, correlations — all of them are tools derived from that single line of insight.
Reading your position on the Efficient Frontier
On the efficient frontier chart in your analysis page, how close was your portfolio (red dot) to the curve? That distance tells you almost everything.
Well diversified
You're already getting close to the maximum return possible for the risk you carry. Stick to periodic checkups and gentle tweaks rather than big changes.
Room to improve
There is a different weight combination that delivers more return for the same risk. Rebalancing can lift you onto the curve.
Structural rethink needed
Weight tweaking alone may not be enough. Reconsider the holdings themselves, or add an asset class that brings new diversification.
Four ways to use the result
Markowitz analysis won't make decisions for you. It gives you the basis to make them. These are the four most common applications.
Are your current weights truly optimal?
Will adding this stock improve diversification?
How do you read the market right now?
How much can you lose in a bad year?
What Markowitz doesn't tell you
The better a model is, the more carefully you need to know its limits. Four things to keep in mind before trusting the output blindly.
Past data is no guarantee of the future
Expected return, volatility, correlations — all of them are estimates from past prices. When market structure shifts (rate-cycle turns, industry paradigm changes), the numbers go stale quickly.
Tail risk hidden by the normal assumption
The model assumes returns are bell-shaped, but real markets produce extreme events (2008, March 2020) far more often than theory predicts. This is why VaR is not a reassurance.
Input sensitivity
A 1-percentage-point change in expected return can shift the optimal weights dramatically. "Reasonable weights under current assumptions" is a more honest phrase than "optimal portfolio."
Real-world frictions are absent
Commissions, capital-gains tax, slippage, market impact — the model ignores all of them. Frequent rebalancing to chase a tiny Sharpe gain can have those costs swallow the gain entirely.
If Markowitz isn't enough
Markowitz answers "how should I diversify" but not "why does my return look this way." Two portfolios with identical returns can have entirely different sources — broad-market beta, small-cap exposure, value premium. Decomposing the source of return calls for a different model.
This article was written with the help of AI.