CHATRoundhill Generative AI & Technology ETF— fair value (DCF model)
A deterministic two-stage discounted-cash-flow (DCF) estimate of this company's per-share fair value, and how far it sits above or below the current share price — with the full assumptions, a growth×discount sensitivity grid and the methodology. Every figure is engine-computed from SEC filings, delayed prices and the 10-year Treasury yield, with no LLM. This is a model estimate under disclosed assumptions, not a price target, forecast or investment advice.
How the model works
- Two-stage DCF. Stage 1 projects ten explicit years of cash flow; stage 2 caps it with a Gordon terminal value. The model runs two variants — one over free cash flow, one over earnings (net income) — whenever each is computable.
- Dollar-level projection ÷ current shares. The company-level dollar series is projected and divided by the current share count once at the end. Dollar totals are split-immune, whereas a per-share history mixes pre/post-split bases.
- Linear growth decay. Stage-1 growth starts at the historical CAGR of the base series (clamped into 0%–20%; the raw CAGR is still disclosed) and decays linearly to the terminal rate.
- Gordon terminal growth = min(10-year Treasury yield, 3%). A company cannot outgrow the economy forever; the discount rate must clear the terminal rate by a minimum spread or the value is undefined.
- CAPM discount rate = 10-year Treasury + beta × equity-risk premium (5%). Beta is clamped into 0.6–2.0 (a degenerate regression beta destabilises the model); an unknown beta defaults to 1.0. Every clamp/default is disclosed.
- Honesty gates (absent, never fabricated): at least four annual points with positive first/last values to anchor a CAGR; banks, insurers and REITs are out of model scope (an FCF/earnings DCF structurally misfits their economics); and a result outside 1/8×–8× of the current price is withheld — the assumptions do not fit that business, so no number is shown.
- Sensitivity, not a single oracle number. A 5×5 grid over (growth-start offset × discount offset) shows how the estimate moves as the two key assumptions change — the honest presentation of model uncertainty.
This page is a deterministic model estimate under disclosed assumptions — not a price target, forecast or investment advice. Source: SEC XBRL filings, delayed prices and US Treasury yields; for reference only.