Questions a skeptic should ask

The model is a multi-factor weighted composite. The heaviest weight sits on on-chain and valuation signals that reliably discriminate cycle tops from bottoms, with momentum and sentiment next, and macro kept light as context. Each signal family, the share of the score it carries, and how the scoring works are published on the Methodology page — the specific indicators inside each family and their calibration are proprietary. You can judge the approach yourself rather than take our word for it.

Honestly stated: the model's weights were refined by studying past cycles, so backtest results are in-sample and should be read as illustrative of the strategy's logic — not as a validated track record. What we commit to is publishing the signal weekly, in public, without retroactive edits. Over time that builds a genuine out-of-sample record you can evaluate. We'd rather earn trust slowly than claim a track record we can't defend.

No — and that's deliberate. The model outputs a risk score and zone (accumulate / hold / reduce), which is a statement about historical conditions, not a trade call. It quantifies when conditions have historically favored deploying capital versus preserving it. What you do with that depends on your situation, which we don't know. Nothing on this site is financial advice.

Alphabit is built by Sahil Kaila and a small team — the story and people are on the Our Story page. The model logic is the founder's own quantitative work, and the methodology is fully disclosed rather than hidden behind "proprietary" language.

The model draws on on-chain, market, sentiment, and macroeconomic data from reputable third-party providers, plus exchange price history. We don't generate any input data ourselves. How the signal families are weighted and scored is published; the specific indicators within each family and their calibration are proprietary.

Factor data refreshes daily as sources publish; the actionable signal is weekly, mapped to a Sunday cadence. Crypto risk conditions change over weeks and months, not minutes — a weekly cadence is a feature, not a limitation. It keeps the process disciplined and ignorable noise ignored.

The current week's risk score and zone are free, always — along with Fear & Greed, the Macro Dashboard, the DCA Calculator, and the Whale Tracker. Paid tiers (see Pricing) add depth: full signal history, per-factor breakdowns, multi-asset coverage, account-synced portfolio tracking, and a weekly digest. The core insight is never paywalled.

No. Alphabit is analytics only — no custody, no exchange API keys, no wallet connections. The Portfolio Tracker runs on manually-entered numbers, stored in your own browser (or synced to your account if you sign in). Details on the Security page.

The published default weighting is thesignal — one number, computed the same way for everyone, so it stays repeatable and accountable. An advanced mode for exploring custom weightings exists in the research tool, but we treat the default as canonical: a signal you can tune to say anything isn't a signal.

It will be wrong — any honest quantitative process is wrong on individual calls. The design goal is that the processis defensible even when an outcome is bad: factors are disclosed, changes are documented, and the signal history stays public. That's the same standard a quantitative research desk holds itself to, and it's the difference between a bad outcome and a bad process.