Audit news coverage and source reliability for your tracked assets
Answers how well the news feed covers the assets you track, and how trustworthy those sources are. It counts, for each tracked asset, how many articles appeared in the recent window and whether any coverage exists at all, then scores the collected articles against a curated publisher-reliability list, rating coverage as high- or low-reliability, flagging when a meaningful share comes from rumor-prone or social sources, and listing the top sources behind each read. Use it to know which assets' sentiment reads are well-supported and trustworthy versus thin, blind, or built on shaky sources. It matters because a sentiment score built on no articles, or on unreliable ones, is not a reliable read.
A data-quality check on the news feed itself - unglamorous, and logically prior to every other sentiment analysis here.
WHAT IT DOES. For each tracked asset it counts articles in the recent window and whether any coverage exists at all, then scores the collected articles against a curated publisher-reliability list, flagging when a meaningful share comes from rumour-prone or social sources.
WHY THIS COMES FIRST. Every sentiment analysis in this catalog computes a score from a corpus. If the corpus is empty, thin, or dominated by low-quality sources, the score has the same format and far less content - and nothing in the score itself reveals that. Establishing what the input looks like before interpreting an output is the discipline, and it is the same principle as the perpetual-liquidity check that precedes the funding analyses.
THE ABSENCE PROBLEM, which is the most important idea here. An asset with no coverage has MISSING sentiment data - not neutral sentiment. Those must never be averaged together: treating absence as a zero pulls a portfolio-level reading toward neutral for assets nobody has written about, which is a statement about journalism rather than about the asset. Reporting whether any coverage exists at all is what keeps the two apart.
COVERAGE IS NOT RANDOM, and the bias has a shape. News coverage tracks market capitalisation, recent price movement, and existing prominence. So the assets with the richest sentiment data tend to be the large and the recently volatile, and smaller or less prominent holdings often receive least - a plausibly non-random missingness, and the per-asset article counts this analysis reports show its shape in the run's own news rather than assuming it. Any cross-asset sentiment comparison is partly comparing how much is written about each.
THE RELIABILITY LIST IS A HUMAN JUDGEMENT, and worth recognising as such rather than as a measurement. Someone decided which publishers are dependable. That is defensible and it is not neutral, it requires maintenance as outlets change, and a new or foreign-language source absent from the list is unrated rather than unreliable - an important distinction that a scoring system can easily collapse.
WHY THE RUMOUR-SOURCE FLAG MATTERS PARTICULARLY IN CRYPTO. Unverified claims propagate quickly and are frequently re-reported, so a single origin can appear as many independent articles. That inflates both coverage counts and sentiment scores from what is really one source - the same non-independence problem that runs through every synthesis analysis, here operating inside a single input.
The reliability list and the score bands are our own convention.
Published reports are readable without an account. The newest is Ether (ETH) and XRP extend a month-long slide as Bitcoin falls below $75,000: trend, momentum and news (Nov 9, 2026). Create an account to run research on the stocks and crypto you choose.