Bad-News Exhaustion Index
Bad-news exhaustion index: tracks how loudly the news repeats “the bad news is out” — peak noise as a bottom reference.
Computing the bad-news-exhaustion index…
Details
The Bad-News Exhaustion Index turns a market saying into a testable contrarian sentiment gauge: the louder the media repeats “the bad news is exhausted” (利空出尽), the more thoroughly bad news has been chewed over and the closer panic may be to an extreme — but in a grinding downtrend the phrase gets shouted again and again, so whether it works is a question for data, not folklore. The page tracks six phrase groups (the main phrase 利空出尽; extensions 利空落地 “bad news landed”, 坏消息出尽 “all bad news out”, 该跌不跌 “won't fall when it should”; and a despair group 割肉离场 “cut losses and leave”, 爆仓潮 “liquidation wave”) as daily shares of all news volume in the GDELT global news database. Each group is scored against its trailing 90-day baseline, combined by weight into a 0–100 index with 3-day smoothing; above 70 is the climax zone. The main chart overlays Bitcoin, Ethereum, the S&P 500 and the Nasdaq Composite, and the event table lists every climax in the sample (one peak per excursion, peaks less than 10 days apart merged) with each asset's actual +7d / +30d returns and summary statistics — up share and averages included, nothing cherry-picked. Plainly stated: the source is news-media mentions, not full social-media volume — full X (Twitter) search requires a paid API and is not connected in this version. The index is a percentile of relative heat, so the climax zone simply marks the loudest days inside the sample. This tool only compiles and presents public data; it is not investment advice.
Who it's for
Investors who believe “once the bad news is exhausted, it's bullish” and want to know whether the saying has actually worked in history; anyone who likes to check a sentiment thermometer before sizing a position.
18btc Crypto Tools Navigator · Our Take
“Honestly first: the easiest way to misuse this index is treating the climax zone as a buy signal by itself — a percentile method guarantees a batch of days above 70 in any sample, and mid-downtrend bounces shout just as loud. Its real use is with the event table: weigh the +30d up share and average return across all past climaxes before deciding whether to trust this one. And remember it measures media volume, not retail sentiment — the two are separated by a layer, so don't conflate them.”