# QuantStark > QuantStark is portfolio risk analytics for US equity portfolios, built for RIAs, SMA and model-portfolio managers, family offices and emerging managers. It measures factor exposure, risk and return attribution from actual holdings, looking through ETFs and funds to the stocks inside. QuantStark is not related to QuantStart, QuantSpark or QuantStats. ## What it does - **Combined view**: several accounts (brokerage, IRA, 401(k)) are merged at the dollar level, trade histories included, so return, profit and concentration are measured across everything you own. - **History rebuilt from trades**: trades, dividends, splits and transfers are replayed to reconstruct how the money actually moved. Deposits are never counted as profit. Positions that cannot be reconciled against the account are labelled approximate rather than guessed. - **Fund look-through**: ETFs and supported US equity mutual funds are opened up from their regulatory holdings filings (SEC Form N-PORT), so the real top holdings and factor exposures include what sits inside the funds. - **Findings**: the Overview ranks what deserves attention (the position that carried the period, what lagged, hidden concentration, unintended factor bets, idle cash) and reports strengths too. Written explanations only restate figures the engine computed; the AI never does the arithmetic. - **Performance**: time-weighted and money-weighted return over any window, drawdowns, and per-position profit in dollars from actual fills, including positions already sold. Benchmark: the S&P 500 over the identical window. - **Composition**: true holdings after look-through, the whole book described as one company (valuation, profitability, leverage against the index), per-stock factor scores, and systematic factor exposures to value, size, momentum, volatility, leverage and growth, sized against the index. In this model a high size exposure means larger companies. - **Scenarios**: past market episodes (for example the 2008 crisis, the 2020 Covid crash, the 2022 rate rises, the Q4 2018 selloff) replayed on what you hold today. - **What-if**: add, trim or drop names and watch concentration and risk respond before any money moves; a suggested fix arrives already drafted; Fit check tests how one stock would fit before you buy; a Watchlist ranks candidate names the same way. - **Markets research**: a stock screener (by factor, sector and size) and a two-ETF comparison (holdings, exposures, overlap). Both work with no portfolio loaded. - **AI connector (MCP)**: ask Claude, Cursor or ChatGPT about your own portfolio in plain words. Ten read-only tools; included with every account. ## Professional and Free - **Professional** (by invitation, https://quantstark.com/professional): everything in Free, plus the Risk chapter (predicted volatility, beta and tracking error; factor versus specific risk), factor-based return attribution, side-by-side portfolio Compare, the Optimizer (least risk, least tracking error, or the smallest change that fixes one problem), PDF and HTML report export, books for every client, and the Factor Monitor and Backtest Engine research tools. - **Free dashboard** (sign up at https://quantstark.com/portfolioanalytics): Overview findings, Performance (returns, drawdowns, per-position profit), Composition (holdings, profile, scores, factor exposures), Scenarios, What-if with Fit check, Watchlist, the stock screener, ETF comparison and the AI connector. ## Data and methodology - About 5,400 US stocks and about 2,000 equity ETFs with parsed holdings, refreshed daily at end of day. - Point-in-time fundamentals (no look-ahead, no survivorship bias); exposures are built from what companies reported, not just inferred from price history. - A 132-factor research library; an 18-factor daily risk model (6 style, 11 industry, 1 market). - Coverage is shown on every view, including what could not be covered. ## Connecting accounts - Broker sync through SnapTrade: Charles Schwab, Fidelity, Robinhood, Vanguard, Interactive Brokers, E*TRADE and 20+ more. You log in on the broker's own page; QuantStark never sees broker credentials and access is read-only. - File import: a Schwab transaction export (full history) or any holdings CSV or Excel file. ## Limitations - US-listed equities and US equity funds only. Bonds, commodities, crypto, non-US markets and multi-currency portfolios are not analysed. - End-of-day data, not intraday; QuantStark is not a trading system and places no trades. - Options positions are valued but sit outside the factor model and are labelled approximate. - The factor/specific risk split needs roughly 10 or more holdings to be stable. - Informational and educational only; not financial, investment, legal or tax advice. ## Privacy Portfolio data is stored encrypted at rest, tied to your account, never sold and never shared for advertising, deleted after 180 days of inactivity, and hard-deleted within 30 days of account deletion. Details: https://quantstark.com/privacy and https://quantstark.com/security. ## Company QuantStark Analytics Private Limited. Co-founders: Pranvendra Chaturvedi and Pratyaksh Mangal. Contact: https://quantstark.com/contact. LinkedIn: https://www.linkedin.com/company/quantstark ## Pages - [Home](https://quantstark.com): What QuantStark does and who it is for. - [Live demo](https://quantstark.com/demo): See QuantStark analyze a sample US equity portfolio end to end: factor exposure, ETF look-through, risk decomposition and stress scenarios. No sign-up. - [Under the Hood](https://quantstark.com/under-the-hood): Holdings-based vs returns-based factor analysis, point-in-time data, risk decomposition and the 18-factor fundamental risk model behind QuantStark. - [AI connector (MCP)](https://quantstark.com/mcp): Connect QuantStark to Claude, ChatGPT or Cursor and ask about a portfolio's factor exposure, risk and concentration in plain words. Ten read-only tools. - [Professional](https://quantstark.com/professional): An 18-factor fundamental risk model, return attribution, client books and PDF risk reports for US equity SMA, model-portfolio and RIA books. By invitation. - [Security](https://quantstark.com/security): How QuantStark protects portfolio data: encrypted at rest, never sold, read-only broker connections, and deletion after 180 days of inactivity. - [Insights](https://quantstark.com/blog): Research notes on factor exposure, hidden concentration and portfolio risk for advisors and US equity managers. - [Privacy Policy](https://quantstark.com/privacy): What is stored, for how long, and how it is deleted. - [Terms of Service](https://quantstark.com/terms): Terms governing use of QuantStark. - [Contact](https://quantstark.com/contact): Request QuantStark Professional for your firm, or ask a question. A person reads every message and replies. ## Blog - [Five Funds, One Bet](https://quantstark.com/blog/five-funds-one-bet): Five funds. Five fund families. Five categories on a holdings page. One bet. The principle is obvious. The verification is not. - [You Don't Have a Concentrated Portfolio. You Have an Unmeasured One.](https://quantstark.com/blog/unmeasured-concentration): Holding NVIDIA from 2016 would have made you rich. So would holding Cisco from 1998. The honest question isn't which stock you'd have picked. It's whether the same process that led you to one would have also led you to the other.