Data and methods guide
A clear result starts with a traceable input.
Use this guide to find source documents, prepare CSV files and understand what StockMags calculates. A number without a date, definition or source is not ready for your model.
Where to start looking
| Need | Start here | Check before using it |
|---|---|---|
| Company financials | SEC EDGAR and the issuer’s investor-relations site | Latest relevant filing, period, units, currency, restatements and diluted shares. |
| Fund holdings | The exact fund’s issuer page and downloadable holdings | As-of date, coverage, security IDs, weight definition and share class. |
| Fund fees and mandate | Fund prospectus and shareholder reports | Expense definitions, waivers, investment objective and risks. |
| Historical observations | A data source whose terms permit your intended use | Price versus total return, split and distribution adjustments, currency and observation dates. |
| Dividend research | Issuer dividend declarations and financial reports | Declared amount and dates; distinguish a declaration from an estimated future payment. |
Keep a simple source note: document URL; company/fund; financial period; as-of date; currency; units; metric definition; your adjustments. Save an explanation when you change a reported figure.
What each Research Lab tool actually does
| Tool | Calculation | Important limit |
|---|---|---|
| EPS fair value | Projects diluted EPS, applies an exit P/E, then discounts that future share price. | Positive EPS only. Interim dividends are excluded. Growth must reflect per-share dilution or buybacks. |
| 10-year DCF | Discounts equity cash flow per share for ten years and a continuing-growth terminal value. | Use cost of equity, not WACC. Terminal growth must be below the discount rate. Do not subtract debt twice. |
| Fund overlap | Sums minimum weights in matching holdings; separately computes allocation-weighted exposure. | Long-only data. Missing holdings are not normalized away. No automatic look-through of nested funds. |
| $10,000 comparison | Normalizes each series to $10,000 and calculates total change, CAGR and observed drawdown. | Matching dates and currency required. Selecting total return does not create dividend adjustments. |
Prepare a CSV the tools can read
Fund holdings format
Use exactly these header names. Every fund needs a consistent date in YYYY-MM-DD format. Weights are percentages: enter 4.5 for 4.5%. Stable IDs must identify the same security across funds.
fund,holding_id,holding_name,weight,sector,country,as_of DEMO_A,EXAMPLE_ALPHA,Alpha Example,60,Technology,US,2026-09-30 DEMO_A,EXAMPLE_BETA,Beta Example,40,Industrials,US,2026-09-30 DEMO_B,EXAMPLE_ALPHA,Alpha Example,20,Technology,US,2026-09-30 DEMO_B,EXAMPLE_GAMMA,Gamma Example,80,Healthcare,US,2026-09-30
All names, IDs and values above are fictional. Holdings weights must total no more than 100% per fund. Partial files stay partial. Duplicate IDs within a fund, negative weights and inconsistent dates are rejected.
Optional allocations use fund,allocation, with one row per fund and percentages totaling 100. The tool accepts 2–10 funds and files up to 2 MB, subject to its 10,000-row limit.
Historical comparison format
Start with date, followed by 2–20 unique symbol columns. Every row needs a YYYY-MM-DD date and a positive value for every symbol.
date,DEMO_A,DEMO_B 2025-01-01,100,100 2025-07-01,95,105 2026-01-01,110,108
Fictional teaching data, not market prices. Keep one currency and a consistent adjustment basis. Missing observations and duplicate dates are rejected. Dates are sorted; prices are not filled forward.
CAGR is shown for periods of at least one year. Drawdown uses the observations provided; sparse dates can miss deeper declines between observations. Chart points are equally spaced, so use consistent observation intervals for visual comparisons.
Understand where your work stays.
The four Research Lab calculators process their inputs in the page. They do not upload CSV files or save analyses into an account. Reloading clears their input state. The free Decision Journal is a separate tool that stores entries in this browser; it is not automatic account synchronization.
The surrounding website may run its existing analytics and advertising. See the privacy policy. Keep account numbers and other private identifiers out of uploaded research files and support messages.
Troubleshoot an unexpected result
The calculator rejects my DCF assumptions.
Use positive equity cash flow per share and make terminal growth lower than the cost of equity. Check percentage signs and units. An unstable or negative-cash-flow business may need a different model.
Overlap looks too low.
Check whether the files contain all holdings, share the same dates and use matching IDs. Similar company names do not automatically match. Different share classes and nested fund positions may need additional research.
My comparison does not match another chart.
Check dates, currency, split adjustments, reinvestment assumptions and the starting observation. A price chart and a total-return chart can correctly show different answers.
I cannot download an export.
Try the visible export control in your normal browser and check its download settings. If the file does not appear, record your inputs and results in the Decision Journal and report the tool and browser.
Further reading: SEC: How to Read a 10-K · Investor.gov: ETF bulletin.
StockMags educational guide · Updated October 4, 2026. Research tools do not determine whether an investment is suitable for you. Investments can lose value.
