How a call is graded
Each published rating with a price target is compared with the stock’s price at the time of the note. A target more than 3% above the price is a bullish call; more than 3% below is bearish; anything inside that band is a reiteration and is not graded. The call is then scored against the subsequent closes: a bullish call “hits” if the stock rises, a bearish one if it falls.
Hit rate is the bar, return is the tiebreaker
Bars show the share of graded calls that moved in the analyst’s direction. Where two analysts have the same hit rate, the one whose calls produced the larger average move ranks higher. Both numbers are on the detail card.
Why calls need to season
A rating published yesterday cannot be judged yet. Calls are graded only once they are at least 30 days old, and this leaderboard admits only analysts with at least 20 graded calls. Even then, 6-for-6 is luck as often as skill, so the ranking uses the statistical lower bound of each hit rate (the 95% Wilson interval): 27 hits from 32 calls outranks a perfect 8 from 8. The bar still shows the plain hit rate.
The leader: Eric Heath
Eric Heath sits first with a +85.2% hit rate across 30 rated calls, and an average subsequent return of +52.9%. Click the name to open the rating history.
Junk targets are filtered out
Data vendors occasionally ship nonsense — a $700 target on a $180 stock. A note is only graded when the target sits within 0.4×–2.0× of the price on the day of the note and within three times the consensus. This keeps a data error from inflating anyone’s record.
What a good hit rate looks like
Analysts as a group are right a little more often than not — the market drifts upward and most calls are bullish. A hit rate in the 70s over a meaningful number of calls is unusual; above 80 over many calls is rare; 100% over any real sample does not exist, which is why a name showing it has simply not made enough calls yet. Sample size matters more than the headline percentage.
Sector concentration
Most analysts cover one sector. The detail card shows the sector an analyst’s rated symbols cluster in, because a strong record in a sector that happened to rally is a different thing from a strong record against the tape.
Update schedule
Ratings are collected continuously; the leaderboard is rebuilt on the first of each month, and the Updated date reflects that rebuild.
