Where the numbers come from
Every figure on the site, where it originates, and how old it is.
GemRadar does not collect this data itself. Prices, filings and fundamentals reach us through a commercial market-data provider that aggregates them from the sources below. What matters for reading a number is where it originates — a Form 4 filing and a scraped forum post are not the same kind of fact, whoever delivered them.
| Figure | Originates from | Freshness |
|---|---|---|
| Prices, charts, 52-week range | Exchange end-of-day closes | Previous close, updated by the daily sweep |
| Intraday chart (1D) | Exchange 5-minute bars | During and after the session |
| Insider transactions | SEC Form 4 filings | As filed; filings lag the trade by up to two business days |
| Analyst rating and price target | Published broker ratings, aggregated | Current snapshot only — see the warning below |
| Forward P/E, market cap, financials | Company filings (10-Q, 10-K) | Quarterly reporting, so it lags the price |
| Short interest | Exchange short-interest reports | Twice monthly, as exchanges publish it |
| Earnings dates | Company announcements | As announced |
| News | Licensed newswires | Continuous |
| Market open or closed | Schedule plus a live trade check | See market hours |
The one caveat worth knowing
Analyst data has no history. The rating and price target arrive as they are today, with no dates attached. Insider filings and prices can be reconstructed for any past date; analyst consensus cannot.
That matters for the track record. We record the analyst figures with each sweep from now on, so the point-in-time record builds forward. What we cannot do is reconstruct what analysts thought in the past — and we will not use today's rating to score a past day, because analysts upgrade after a stock moves. Doing so would flatter exactly the calls that went up, in the one direction that makes a record worthless.
What we do not use
No sentiment scraping, no social media signals, no alternative data, no paid-for research. Public filings and exchange data, joined and weighted, with the weighting written down.
When data is wrong
It happens. Prices arrive in the wrong currency, share counts belong to a different class, a provider corrects itself a day later. GemRadar refuses to score a stock whose price it does not trust, re-checks stored rows when it reads them, and drops implausible jumps from the track record rather than counting them as spectacular results.