About Marko-Viz
A calculator, not an adviser — by design.
Marko-Viz computes the efficient frontier — the set of portfolios with the least risk for each level of expected return — for any assets you choose, over a historical window you choose, under expected returns you supply.
It is deliberately not an "optimizer" that hands you a portfolio. One rule decides every feature of this tool:
Every forward-looking assumption comes from the user. The application contributes arithmetic and clearly-labeled historical data — nothing else.
That is why the output is a whole curve and never a point: selecting a portfolio on the frontier is a judgment, and the judgment is yours to make.
Your assumptions, explicitly
Expected returns are a mandatory input, empty by default. The realized historical return is shown next to it as a labeled fact — adopting it as your forecast is a deliberate act, because past performance is not a prediction. That friction is the product, not a bug.
A curve, never a point
No portfolio comes preselected. Notable points — global minimum variance, tangency — are marked and labeled by their mathematical definition, never as "the best" or "recommended".
Fragility made visible
Mean-variance optimization is an error maximizer (Michaud, 1989). The sensitivity panel perturbs your expected returns within their estimation error and shows the resulting range of "optimal" weights — usually wide enough to break the illusion of precision. Few tools show this; it is the most honest number on the site.
Reproducible by construction
Windows are absolute dates, and a shared link stores only the inputs — opening it recomputes the result from scratch. Same question, same answer, for anyone, with no account and no stored portfolios.
Under the hood
A pure-Python engine (pandas, PyPortfolioOpt, SciPy) behind a FastAPI + HTMX interface — no accounts, no client framework, no state beyond a cache. Prices come from Yahoo Finance and are cached locally. Covariance is estimated from the historical window (sample or Ledoit-Wolf shrinkage, your choice) — it describes the past; the expected returns describe your view of the future. Keeping those two apart is the entire point.
Built by Matías Tudela — get in touch.