Record whether each company's reported EPS and revenue beat, missed or matched expectations at recent quarterly releases, as news coverage stated it
Lists how each company's recent quarterly results compared with expectations, as news coverage of each release stated it. For each release on the company's earnings calendar it reads articles published from the release date to ten days after, and keeps the results an article states: EPS and revenue, beat, missed or matched, with the reported and expected figures as written. A result is kept only when the article's own sentence states it, and several articles on one release count as one result. It shows the latest release, the releases on record before it, beat and miss counts, any streak of consecutive releases with the same outcome, and the releases for which no result was read. It records past results only: it uses no analyst consensus feed and says nothing about the next report. The record grows as companies report, and holds only releases whose coverage has been read, so a company may have few releases on record.
A record of reported results against expectations - and why it stops at the past.
WHAT IT READS. For each release on a company's earnings calendar, the articles published from the release date to ten days after it. A language model reads each headline and summary and keeps the result an article states: EPS or revenue, beat, missed or matched, with the reported and expected figures exactly as written. A result is kept only when the article's own sentence states it and every figure in it appears in that sentence.
ONE RELEASE, ONE RESULT. Five articles about one quarter are one observation, not five. The record takes the outcome most of them state, with the figures of the earliest article that states it, and files each article under the release whose window it falls in - so a preview that recalls last quarter's result is not taken for this quarter's.
EPS AND REVENUE ARE SEPARATE, AND SO ARE BASES. Companies often beat on adjusted EPS while GAAP EPS misses, because adjusted figures exclude items such as stock compensation and one-off charges. Consensus is usually quoted on an adjusted basis, so the record uses the adjusted figure where an article names it, otherwise the figure as stated, and a GAAP figure only when nothing else is stated.
A STREAK NEEDS CONSECUTIVE QUARTERS. A run of beats is counted only across consecutive releases; a release with no result read ends it, because an unread quarter is evidence of neither.
THE PHENOMENON, WITH ITS SCOPE. In recent US large-cap samples, reported beats have substantially outnumbered misses - a pattern that varies by market, period, metric and data vendor. It is not evidence that companies routinely outperform: guidance and expectation management contribute, and consensus is a socially produced benchmark rather than a neutral forecast. A long beat streak says a company has cleared the bar it was given; it does not say the bar was high.
WHY IT STOPS AT THE PAST. What moves a price at a report is the result against what was expected at that moment. Without a current consensus there is no bar for the next report to read, so this analysis makes no statement about it. The expectations it records are the ones the coverage stated at each release.
THE SMALL-SAMPLE CAVEAT. A record is typically a handful of quarters. 'Usually beats' from four observations is a weak prior, and a business or a management team can change between them.
The release window and the outcome rules 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.