About Sun Insight Lab
Sun Insight Lab is an independent educational project at suninsightlab.com. It pairs a browser-based historical investment simulator with plain-English writing on risk, valuation, diversification, and time horizon. It's built for everyday investors who want to test ideas against real historical data — and understand the reasoning — before committing capital.
What you'll find here
| Section | What it does |
|---|---|
| Quant Lab | Backtest rule-based factor strategies — market cap, momentum, low volatility, value, and quality — across the whole S&P 500. |
| Simulator | Three modes — Single, Compare, and Portfolio (weighted, with optional monthly rebalancing) — over historical monthly data. |
| Retirement | Backtest a TSP or 401(k) payroll-contribution plan against an S&P 500 benchmark. |
| AI Bubble Meter | A breadth-and-concentration gauge for how narrow the current market rally has become. |
| Analysis | Market notes across four tracks: Market Trend, Deep Dives, Breaking, and Basics. |
| Resources | A library of guides organized by level: Beginner, Intermediate, and Advanced (including options strategies). |
The aim
My aim is to make historical investing questions testable. I state assumptions, distinguish examples from observed outcomes, and publish corrections when a claim cannot be supported.
How the simulator is built
The simulator is a Web Component written in vanilla JavaScript. It fetches monthly closing-price data (sourced from Yahoo Finance) from the database I maintain over a small read-only JSON API, runs a deterministic month-by-month allocation across your selected period, and draws the result with Chart.js. The whole calculation runs in your browser, so you can read the source and reproduce every number.
- Coverage: US-listed stocks, ETFs, the S&P 500, and Bitcoin, with coverage varying by asset; inspect the returned observation dates.
- Update cadence: separate scheduled pipelines; inspect each tool’s latest observation rather than assuming an update succeeded.
- Reproducibility: the price API is public and read-only; script tags use
?v=cache-busters whenever the logic changes.
Who's behind it
Sun Insight Lab is written and built by one person. I've been investing my own money for about twenty years, and over that time my interest moved steadily away from stock tips and toward rule-based, quantitative questions — the kind you can actually settle with data instead of argument. This site is where I share what that experience taught me, and the tools I built to keep testing it.
Everything here I built myself, and that is the part I'd point to rather than any credential. The Quant Lab ranks the full S&P 500 by market cap, momentum, low volatility, value, or quality and backtests the rotation month by month. The simulator replays historical monthly closes across single positions, comparisons, and weighted portfolios with optional rebalancing. The Bubble Meter scores market breadth and concentration against its own history. Behind them sits a data pipeline I maintain: daily prices into a local database, reduced to monthly closes, pushed to an edge database that the site reads over a public read-only API. The studies publish frozen inputs, formulas, and executable Python so you can re-run them and disagree with the results.
What I am not: I hold no professional investment license or adviser credential, I manage no one else's money, and I'm not a financial professional. Twenty years of investing my own capital is experience, not authority — which is exactly why the tools show their assumptions and the studies ship their code. Check the work rather than taking my word for it. The project provides no brokerage services and does not manage visitor portfolios; the private portfolio feature is separate from the public tools. See editorial standards and disclaimer.
Reach me at suninsightstudio@gmail.com.
How content is made
Every article here starts with me. I decide what's worth writing about — usually a question I've hit in my own investing — and I choose the sources it has to be built on: company filings, Federal Reserve and regulator releases, index-provider documents, or output from my own backtests. I give direction on the argument the piece should make and what it must not claim, then I read the result and revise it before it publishes. Nothing goes up that I haven't gone through.
AI tools assist with drafting, editing, and code. Editorial direction, source selection, and final review are mine, and I'm responsible for what's published.
The tests I apply are the ones I'd want applied to me. Numerical claims have to be reproducible: the studies publish input snapshots, formulas, and scripts so you can re-run them yourself. Factual statements cite a filing or primary source next to the claim, not a link to some organization's homepage. Calculated examples are labeled hypothetical. A company's forecast is labeled as its forecast. Where a claim can't be supported, it comes out or gets corrected in public — the editorial policy keeps a dated log. One honest limit: historical vendor price data can contain errors, so reproducing a calculation proves the arithmetic, not the accuracy of every input. See Methodology for how the data is assembled.
Found something wrong? Send me the URL, the disputed statement, and your source at suninsightstudio@gmail.com or via the contact page. Corrections get made.
What this site is not
We do not provide personalized financial advice, brokerage services, portfolio management, or guaranteed-return claims. The simulator does not model taxes, trading fees, slippage, bid/ask spreads, or account-specific rules — real-world results will differ. See our Disclaimer.
How to verify our work
- Reproduce a simulation: open the simulator, pick the same asset and date range used in any article, and confirm the totals match.
- Inspect the data: the price API at
/api/prices?symbols=returns the same monthly closes the simulator uses. - Read the source: the simulator file
stock-calculator.jsis unminified. - Report a discrepancy: use the contact page with the URL, the parameters, and your cross-check source.