How to Read the Bull-Bear Index (BBI)

What the BBI measures

The Bull-Bear Index is a single 0–100 number answering one question: "how favorable is the current news flow for markets?" It does not predict prices — it is a live thermometer of the direction and weight of financial news published over the last 24 hours.

How it is computed

Individual calls: our AI classifies each story as bullish, bearish, or mixed, and assigns a 0–100 impact score. A Fed rate decision lands in the 90s; a minor exchange listing might be a 40. Not all news weighs the same.
Weighted-average comparison: the average impact of bullish stories over the last 24 hours is compared against the average impact of bearish ones. Story weight moves the index, not story count — one 90-impact catalyst outweighs a hundred quiet 40s on the other side.
Time decay: older stories count less (roughly a 12-hour half-life). A morning catalyst has about half its influence by evening.
Deduplication: stories covering the same event are clustered and counted once. An event reported by ten outlets does not move the index ten times.

Reading the ranges

0–35 (Bearish): bearish news clearly dominates — negative catalysts are driving the tape.
35–48 (Mild Bear): mixed, tilted negative. Slightly more risk than reward in the flow.
48–52 (Neutral): balanced flow, no clear direction.
52–65 (Mild Bull): mixed, tilted positive.
65–100 (Bullish): bullish news clearly dominates.
Insufficient sample: too few scored stories in the window for the number to mean anything. We show this instead of a band rather than manufacture confidence we do not have.

This table is not currently accurate (measured 2026-08-09)

Those five ranges are the boundaries the code actually uses. But across 180 days of measurement the index stayed between roughly 42 and 53. Bullish and Bearish have never occurred, and Mild Bull only rarely. Describing five ranges on this page was not truthful.

The cause is an asymmetry in scoring. The analyzer is instructed to judge impact independently of whether a story is bullish or bearish, but measured over samples of more than ten thousand stories per side, bearish stories average about 8 points higher. So the gap between the two scores sits near −7 rather than 0, and the index only moves within a narrow band pinned below the midpoint.

To test whether this was a wording problem, we re-scored 200 stories with a prompt containing no bullish/bearish vocabulary at all and compared them against the same stories scored normally (100 pairs matched on category and date). The gap survived — a difference-in-differences of −1.25 ± 1.66, indistinguishable from zero. It is not something better prompt wording can remove.

We will therefore recalibrate the boundaries against the measured distribution. The recalibration will not retroactively change any figure already published; it will be introduced as a new series with a stated changeover date. Until then the boundaries above remain in force, and this note stands as the disclosure of their limits.

The change often matters more than the level. If the BBI moves from 45 to 48 in half a day, something changed the flow — the Top Driver card on the homepage shows you what.

How it differs from Fear & Greed

CNN's Fear & Greed and Alternative.me's index are computed from market data — momentum, volatility, volume. The BBI is computed from news text. That means it can show a different picture in phases where the news builds before the price moves. They are complements, not substitutes.

What the BBI is NOT

Not a buy/sell signal. "BBI 70" does not mean "buy".
Not a price forecast. News sentiment and price often travel together — but not always.
Not infallible. AI calls can be wrong, which is exactly why we verify every call in a public accuracy ledger.

This guide is informational only, not investment advice. See our Editorial Policy & Disclaimer.