Glossary
Every term GemRadar puts on screen, in the fewest words that are still true.
On a stock row
Divergence Score — one number from 0 to 100 combining three independent signals. Higher means more of them point the same way. It is a sorting mechanism, not a verdict, and it always says how many of the three it had data for. See how it is built.
Strong / Moderate / Low — the tier a score falls into: 66 and up, 33 and up, below 33. The same thresholds apply to each pillar and to the composite, so a colour means the same thing everywhere.
Unknown — no usable data for that signal. It does not count toward the score at all, rather than counting as zero.
fwd PE — forward price/earnings: the share price divided by what analysts expect the company to earn next year. Lower means you pay less per dollar of expected profit. Below 8 scores full marks in our valuation pillar, above 40 scores nothing. Blank means nobody publishes an estimate.
vs 200d avg. — how far today's price sits above or below its own 200-day average, in percent. A rough measure of whether a stock has already run. It appears in the explanation of the market pillar but does not feed the score.
turnover/mcap — average daily trading volume as a percentage of the company's market value. It answers "can I get in and out of this without moving the price myself". Around 0.5% a day is thin; 3% and up is heavily traded.
52-week range — the lowest and highest closing price of the past year. The position within it says where today sits between those two.
On the signals tab
Insider conviction — whether people inside the company bought shares with their own money on the open market in the past twelve months. Grants, vesting and tax withholding are excluded; only genuine purchases count.
Days to cover — how many normal trading days it would take short sellers to buy back everything they have borrowed and sold. Higher means a crowded short position.
% of float — the share of freely tradable stock currently sold short.
Thin evidence (the amber dot) — the rating rests on two analysts or fewer, or on none at all. The signal still counts, but at a reduced weight.
On the trade setup
Entry — the last close. Nothing clever: the price you would realistically have paid at the point you were looking.
Stop — a price below entry at which the idea is wrong, placed at a multiple of the stock's own volatility rather than a round percentage.
ATR(14) — average true range over 14 days: how far this stock typically moves in a day, in dollars. A wide-moving stock gets a wider stop.
Wide / standard / tight — the stop at 3, 2 or 1.5 times ATR. Wider stops are hit less often but cost more when they are; tighter stops are the reverse. There is no safer option, only a trade-off.
Risk / reward — the distance to the target divided by the distance to the stop. 1:3 means the target is three times as far away as the stop. It improves automatically as the stop gets tighter, which is why the next number is there.
Stop hit — how many times in the past 250 trading days a fall of this depth actually happened. It is the honest counterweight to risk/reward: a tempting ratio on a stop that would have been hit nine times is not a good trade.
Target — the average analyst price target, capped at the 52-week high. If no analyst covers the stock, there is no target and no risk/reward — we do not invent one.
On the track record
Call — a gem that scored at least 60 on the day we flagged it. Rows we surfaced with a low score are listings, not judgements, and counting them as calls would measure something other than what we claim.
Proven — a call that is up 15% or more since the day we flagged it.
Sweep — the daily pass over every sector and exchange that produces the list. It runs before the US market opens.
Flagged — the first day a stock appeared in a sweep. Everything about a call is measured from that day, not from the day before.