The Alphave Portfolio Optimizer takes the Indian stocks you already hold and works out what each one's weight should be, given how the holdings actually move together. It reads your book, estimates a return and a risk for every name, builds the efficient frontier those holdings can reach, and shows you the weights at four points on it — minimum risk, best risk-adjusted trade-off, risk parity, and any risk level you pick yourself — against the weights you hold today.
It is a replacement optimiser, not just a re-weighter. The universe it can move money into is not only the stocks you own: it also includes each held sector's own index, the broad market, and a risk-free leg. So the honest answer to "is my bank weight wrong?" can be "move it to Bank Nifty" or "park it in debt", rather than only "buy more of the other bank you happen to own".
Tells you the risk you are not being paid for
One verdict line: the Sharpe ratio your book earns today against the Sharpe ratio the same holdings could reach at the same risk. If those two numbers are close, the optimiser says so and you stop.
Draws the efficient frontier of your own holdings
Not a textbook curve — the frontier your specific stocks span, clickable to set a risk target, with a transition map showing which names enter and leave as you turn the risk dial.
Gives four target portfolios, not one answer
Current, minimum-risk, best trade-off, risk parity and your own target, each with the twelve-month path it actually realised, so you can see what the recommendation would have felt like to hold.
Turns the target into a trade list
An illustrative rebalance for whichever strategy you picked — what to trim, what to add, in rupees — with a hold band so it does not tell you to move a position by half a percent.
Stress-tests today against the target
Value at Risk and conditional VaR, the drawdown curve, and index-shock scenarios, run on both the book you hold and the mix you are considering.
Shows where the risk actually sits
Percentage of money against percentage of risk, per holding, covariance-weighted — which is how a 6% position turns out to be carrying 20% of the portfolio's variance.
Flags redundant holdings
A correlation matrix across your holdings with redundancy flags, so two names that are one bet stop counting as diversification.
Grades each stock as a fund manager would
Per-holding beta, alpha, the t-statistic on that alpha, R² and volatility — plus whether your stock-picking actually beat that stock's own sector index.
Lets you overrule it
Lock a holding to keep today's weight, exclude one entirely, or state your own view on a stock and let Black-Litterman blend it with the market's.
Expected returns are priors, not past returns
By default each stock's expected return is a CAPM prior — its beta against the Nifty 50, an equity risk premium of 6%, and a risk-free rate taken from the 91-day Government of India T-Bill — plus an alpha that is shrunk towards zero by its own t-statistic. Raw historical averages are available as an explicit, warned-about choice, because feeding sample means into an optimiser turns it into an error-maximiser: it overweights whatever was luckiest in the window.
Covariance is shrunk, always
The sample covariance matrix of a few hundred trading days across a few dozen stocks is badly conditioned — it reports correlations it does not have the data to support. Alphave shrinks it towards a structured target before any optimisation runs, in every mode, with no option to turn that off.
Weights are long-only and capped
No shorting and no leverage, because neither is available in a normal Indian retail equity account, and a per-name cap you set so the optimiser cannot answer "put 70% in one stock".
A stock and its sector index share one budget
Sector caps are applied to economic exposure, not to tickers. HDFC Bank and Bank Nifty count against the same bucket, so the optimiser can never manufacture fake diversification by holding both sides of the same trade.
Every assumption is on the page and editable
Estimation window (125, 250, 500 or 750 trading days), return engine, risk-free rate, equity risk premium, per-name cap, hold band, forecast horizon, and a volatility multiplier for "the next year will be stormier than the last one". Change any of them and the frontier redraws. The settings persist in your browser.
Locking a name still moves the others
A locked holding keeps today's weight but stays in the covariance matrix, so the optimiser reallocates around it through the full gradient rather than simply dropping it from the problem.
What the output looks like
Say the book is twelve stocks and four of them are private banks. The exposure read will show the bank weight; the risk-share panel will show those four carrying a much larger share of the portfolio's variance than of its money, because they move together; the correlation matrix will flag them as redundant. The frontier then shows how much risk comes off for how much return given up, and the rebalance list puts numbers on it — trim two of the four, and the freed capital can go to an under-represented sector index rather than to a fifth bank. Every figure is an illustration of the maths on your inputs. It is not a recommendation to buy or sell anything.
Estimates are computed from historical daily closes and are only as good as that history: a stock with a short listing record, a thin float or a recent corporate action will produce unstable estimates, and the optimiser says so rather than hiding it. Long-only and unlevered, which matches a normal Indian retail equity account. No transaction costs, brokerage, STT or tax impact is modelled in the trade list. The cash-versus-equity decision deliberately lives in the Allocation tab, not here — this tool re-weights an equity book, it does not tell you how much of your money should be in equities.
The Portfolio Intelligence workspace at /portfolio is open to everyone and needs no account to look at. Saving holdings needs a free account. The Optimizer tab is included on both paid plans, Alpha Plus (₹499/month) and Alpha Pro (₹999/month) — it is not on the free plan.
Informational only — not investment advice. Alphave is NISM-certified and is not a SEBI-registered Research Analyst. Market data is live during NSE trading hours; company financials are as filed by the company.