Theme leader
The stock a theme is traded through: when the theme moves, it moves.
The purest way to own a story the market is following. The research tracks the theme as closely as the company, because a leader rarely outruns a theme that is fading.
Glossary · 55 defined terms
Stock research glossary.
Every term you meet on the stock pages, the portfolio and the daily journal, in plain English: archetypes, market regimes, conviction levels, and how the portfolio decides and is measured.
Archetypes
Visual taxonomy →Picks and shovels
Sells what a theme consumes, and earns from the build-out rather than the outcome.
Suppliers of the chips, equipment, materials or services a boom needs. The research follows orders, backlog and the spending plans of the customers doing the building.
Second-order beneficiary
Gains from a theme at one remove: the theme's spending or demand reaches it, though the story is not about it.
Power and cooling suppliers to AI data centres are the best-known case. The profile also covers a hotel owner filling rooms on Silicon Valley demand or a retailer selling into a housing recovery. The research follows the link: the contracts, orders or traffic that tie the business to the theme, and how much of its results the theme explains.
Business pivot
An established company re-aimed at a new source of demand. The case is the turn, not the incumbency.
The research checks whether the new business is growing fast enough to change how the whole company is valued, and what the old business still contributes.
Event-driven
A dated event decides it: a trial readout, a deal closing, a contract, a ruling.
The outcome of one event matters more than the path to it. The research follows the milestones and the terms, and names the date that settles the case.
Crowd-driven
Moved more by trading flows, short interest and float than by the business.
These stocks can move far in either direction on positioning alone. The research says so plainly, and the kill line is usually tight.
Distinct case
Fits none of the other six cleanly, so it is read on its own terms.
Rather than force a stock into the nearest profile, the research keeps it separate and states the specific evidence that will decide it.
Regime classifier
Visual taxonomy →Risk-on
The S&P 500 is above its trend, most stocks are in uptrends and volatility is contained.
The daily read with the index more than 0.5% above its average, more than 60% of stocks above theirs (55% will do with the index at least 5% above), and the VIX below 28. Published as context on every journal entry. No rule acts on it; Claude sees the same three inputs when it decides the portfolio.
Neutral
The three inputs disagree: no clear call either way.
The read whenever neither risk-on nor risk-off applies, for example the index above its trend while too few stocks are in uptrends. Published as context. No rule acts on it; Claude sees the same three inputs when it decides the portfolio.
Risk-off
The index has broken below its trend, most stocks are falling, or volatility is high.
The daily read with the index more than 2% below its average, under 40% of stocks above theirs, or the VIX above 28. Published as context. No rule acts on it; Claude sees the same three inputs when it decides the portfolio.
Risk-off, sustained
A risk-off read that has lasted: the index below its 200-day average for ten sessions in a row or more.
Published as context, like every regime read. It ends on the first close back above the average, and the count of sessions below it starts again from zero.
Conviction levels
Conviction is each note's call against the S&P 500 over the next three months: High and Medium expect the shares to lead it, Low expects them to lag it.
High
More likely to lead the S&P 500 over the next three months, on strong, dated evidence and a clear catalyst.
The strongest of the three calls. The note names its evidence and the dated event that tests it soon. Its kill line is a close below a stated price: a fall through it would prove the call wrong. Conviction is the note's call; whether the portfolio owns the stock is shown separately on each page.
Medium
More likely to lead the S&P 500 over the next three months, by a modest margin.
The case for the shares beating the index is better than even, but not by much: part of it is still open. Like a High call, its kill line is a close below a stated price.
Low
More likely to lag the S&P 500 over the next three months. A firm call, not the lack of one.
The note expects the shares to do worse than the index. A fall would fit that call, so its kill line is a close above a stated price: a rise through it would prove the call wrong. Where a Low note still names only a price below, its stock page says so. The stock stays covered and its note is kept current.
Research, market and portfolio terms
Thesis
The call and the reason for it, in a paragraph, at the top of every research note.
What Claude expects and why, with the sourced case for and against. It is updated whenever the note is re-read, and the update date is shown on the page.
Kill line
A research note's own test, stated in advance, of what would prove the note wrong: a price level, a business fact, or both.
The price level is a daily or weekly close through a stated price: below it for a note rated High or Medium, which expects the shares to lead the S&P 500, and above it for a note rated Low, which expects them to lag. The business fact is a result that would break the call, such as growth below a stated rate for a High or Medium note. It is published with the note before the market decides, and every edition shows where the stock stands against the price level. It tests the note and is never the portfolio's sell rule. A move through a kill line is not by itself a reason for the portfolio to sell. Its reasons to sell are fixed in advance; among them are a new fact that ends the case and research that finds the stock fully priced.
Catalyst calendar
Dated events a research note expects to move the stock.
Earnings dates, regulatory deadlines, sector events, Federal Reserve meetings, inflation releases. Each entry is dated and references its source. Each stock's page lists the events ahead, usually through the next results report; a date the company has not confirmed is marked as an estimate.
Theme
A group of stocks that move together on shared news, such as a new technology, a policy change or a shift in demand.
Each stock is read next to the others in its group, because a story that lifts one name usually lifts its neighbours. The current themes are on the themes page.
Higher-low
A price low that holds above the previous low — a sign a recovering trend is holding.
Research notes use it to describe a trend that is rebuilding; a close that breaks the sequence of higher lows is often a note's kill line.
Price setup
Where a stock's price stands now: its trend, the price levels it has held or lost, and how heavily it trades.
Each research note's Price setup section describes it in plain words: whether the shares are above or below their 50-day and 200-day averages, the recent highs and lows that matter, and whether trading is heavier or lighter than usual.
Momentum score
How strongly a stock is trending, from its daily closes: 1-month price trend 50%, 3-month trend 20% and acceleration 30%.
It ranks the liquid US market every trading day for the hot-stocks board and helps decide which stocks the research covers. It never decides a purchase: Claude decides every position in the portfolio.
Theme radar
The weekly scan for groups of stocks rising together.
A research view of which themes are moving and which stocks lead them, updated weekly. It is not a list of positions.
Universe
The liquid US stocks the daily scan covers.
Two counts are printed, each with its word. Scanned: every US-listed stock with enough trading volume for the daily scan to cover. Scored: the scanned stocks with enough price history to be given a momentum score and ranked, always a few fewer. The scan decides which stocks the research notes cover and how often they are re-read; it does not pick stocks for the portfolio. The portfolio's weekly ranking uses a stricter set: every US-listed common stock and ADR that trades at least $5 million a day, on the median of its last 63 sessions (about three months), so one day of heavy trading does not count as steady trading. The first weekly ranking has not run yet.
Portfolio
The public portfolio Claude runs: what it holds, since when, and how it has done against the S&P 500.
Claude decides every position. It is meant to hold a few stocks, most for quarters and some for a few weeks around a dated event. Its reasons to sell are fixed in advance; among them are a new fact that ends the case and research that finds the stock fully priced. Each holding's return is measured from the average price paid for its shares, each purchase at the price it was actually filled at. Each holding is bought on a written case, with the fact it was bought on and the facts that would end it; its size, its weight and dollar amounts are never published.
Investment case
Why the portfolio owns a stock, written down by Claude before it buys.
One short paragraph per holding: what has to keep being true about the business for the stock to beat the market. It is written down with the dated fact the stock was bought on and its exit conditions.
Exit conditions
Up to three facts about the business, written down at purchase, that would end the case.
For example a missed product launch, a lost customer or a guidance cut. When one is met, the portfolio sells or reduces the holding. A price on its own is never an exit condition. Separately, the portfolio may sell or reduce a holding that is fully priced, when research written that day puts the value of the investment case at or below the share price.
Dated fact
The fact behind a portfolio change, with the day it became known.
Every buy, add, trim and sale is published with its date and its reason. Most reasons are a dated fact about the company, such as a results release. Two are fixed sentences the code writes: a sale or trim because the share price reached the value of the investment case, checked against that day's research, and an add from cash because the investment case is unchanged. The kinds are listed under Reason for a change.
Shorter trade
A position taken for a few weeks around a dated event, such as a results release or a scheduled decision.
Most positions are meant to be held for quarters while the business case holds. A shorter trade is bought when Claude judges that a dated event has not yet moved the price as far as the facts imply. A shorter trade is bought around one dated event, with its end set when it is bought, at most about three months ahead; the end can move once, by at most about a month, on a new dated fact. It is sold at that end, and the planned end is not published. It is kept small, so that a wrong call costs the portfolio little, and it is published with its case and the fact it was bought on.
Buy, add, trim, sell
The four kinds of portfolio change: open a position, raise it, lower it, or sell all of it.
Each change names its dated fact and is recorded with its date and direction. Sizes are never published: the record shows what was decided and why, not how much.
Reason for a change
The kind of fact a buy, add, trim or sale was made on, as the portfolio and track record label it.
One of twelve labels: Results (a results release), Company news, Industry news, Case broken (a new fact that breaks the case), Better use of the money, Portfolio risk (a risk to the portfolio as a whole, such as one stock grown too large), Weekly review plan (a change the weekly re-reading planned, carried out in the following sessions), Opening purchase (one of the portfolio's purchases in its opening build, on the case already written), Added, case unchanged (added to a holding from cash on the case already published), Event passed (a shorter trade sold at the end set when it was bought), Fully priced (sold because the share price reached the value of the investment case, checked against that day's research) and Acquired (the company was acquired for cash, which ended the position).
Invested share
How much of the portfolio is in stocks rather than cash, rounded to the nearest 5%.
The portfolio builds its positions over several sessions. The rest of it is cash, and there is no borrowing. Whenever more than 2% of the portfolio would be left in cash, Claude is asked to say why; a decision without one still trades and is flagged for review. The reason is not published.
Portfolio letter
Claude's weekly letter on the portfolio, written after the close of the last full trading day of each week, usually a Friday.
Written after the weekly review, in which every holding is re-read: what changed in the week, what the portfolio did about it and what it is watching next. It never states sizes or planned trades.
Fill price
The price a stock was actually bought or sold at.
Each holding's return is measured from the average price paid for its shares, each purchase at the price it was actually filled at. A sold position's return runs from the average price paid for its shares to the price it was sold at, not from a quoted price at another time of day. The comparison of the whole portfolio with the S&P 500 is a separate figure: an index on which both start at 100 just before the record's first session.
Pick
A stock the portfolio bought, measured against the S&P 500 over the days it is held.
Claude decides every purchase in the portfolio, and each one is a pick. While it is held, its return is measured against the same money in the S&P 500 from the minute it was bought; when the portfolio sells the whole position, its return against the S&P 500 over exactly the days it was held becomes its final result. Research notes are not picks: only the portfolio's purchases are recorded.
Benchmark
The S&P 500 (via SPY), the yardstick every pick and the portfolio are measured against.
A pick only counts as a win if it beat the S&P 500 over exactly the same days. The S&P 500 side includes dividends and is adjusted for splits; a pick's own return is measured from the prices it was bought and sold at.
Versus the S&P 500
A pick's return minus the S&P 500's over the same days, in percentage points.
If a pick rose 8% while the S&P 500 rose 3% over the same days, it is 5 points ahead. This is the number the track record grades.
Beat or trailed
Whether a sold pick did better or worse than the S&P 500 over the same days.
A sold pick beat if its return was higher than the S&P 500's over exactly the days it was held, and trailed otherwise. A gain can still trail, and a loss can still beat.
Track record
Every pick the portfolio has made since 10 November 2026, measured against the S&P 500.
The public record of every stock the portfolio has bought since the record began on 10 November 2026: when it entered, how it has done, and, once sold, what the S&P 500 did over the same days and whether the pick beat it. A pick is recorded when it is bought, before its outcome is known, and every pick since that date stays on the record.
Owner's review
The one rule written down before the first trade: after a year, a portfolio far behind the S&P 500 goes back to the owner.
From about a year into the record, and about every three months after it, the owner decides whether Claude keeps running a portfolio more than 15% behind the S&P 500. The rule in full, with its status, is on the track record. Its status is updated after every close; before the 252nd trading day it reads as pending.
Market regime
The daily read of the market backdrop, set by fixed rules.
One of four reads (risk-on, neutral, risk-off and risk-off sustained), set from the S&P 500 against its 200-day average, market breadth and the VIX, and stated on every journal entry. It informs how each stock's price is read; Claude sees its three inputs, not the label, when it decides the portfolio, and no rule sizes positions on it.
Market breadth
The share of the 1,000 most-traded US stocks closing above their own 50-day average.
It shows whether a move in the index is broad or carried by a few large companies. Above 60% is a broad uptrend; below 40% means most stocks are falling. It is one of the three inputs to the market regime.
VIX
The market's expected volatility of the S&P 500 over the next 30 days, read from options prices.
Often called the fear index: it rises when investors pay more to protect against a fall. A reading above 28 is one of the conditions for a risk-off regime.
200-day average
An exponential moving average of the S&P 500's daily closes over 200 trading days: every close counts, recent ones most.
The long-term trend line the market regime is read against. An index well above it is in an uptrend. One that has closed below it for 10 sessions or more is in a sustained risk-off regime.
Open methodology
The reasoning, the portfolio and the record are all public.
Research notes, regime calls, the portfolio and its track record are all published, dated and sourced, along with how each is produced.
Falsifiability
A claim only means something if some outcome could prove it wrong — the idea behind every kill line.
A call that no outcome could refute tells you nothing. That is why every research note publishes its kill line, the test that would prove it wrong, before the market decides.
Survivorship bias
Judging a record by its winners after the losers have quietly disappeared.
When losing calls are dropped and only winners are shown, any record looks skilled. The track record is built against it: since the record began on 10 November 2026, every pick is recorded when the portfolio buys it, judged when it is sold, and stays listed either way.
Accelerating
Theme stage: buying is spreading across the group, and the news about it is new.
The first of the three stages a current theme is given. It has to be dated rather than asserted: a named catalyst, new coverage, a new kind of buyer. If attention thins while prices hold, the stage moves to maturing.
Maturing
Theme stage: the easy move is behind it. The story is well known and the stocks still hold up, but the news repeats.
Fewer new buyers arrive and the headlines have stopped being new: analysts adjust targets rather than start coverage. What separates it from saturated is that buyers are still holding the stocks rather than selling them. Most of the themes we cover are at this stage.
Saturated
Theme stage: the story is largely in the price, and few new buyers are left.
Mainstream coverage, analyst targets at or below the price, and good results that no longer move the stocks. The story can still be true and the businesses can still grow: saturation is about who is left to buy, not about the companies. If the price trend breaks, the research may call the theme over.
Dead
The research called the theme over: the story failed or its price trend broke.
Not a fourth stage but the end of the three: the theme leaves the current list. It describes the theme and its price trend, not the companies, which can keep growing. A theme is called over when a catalyst passed and produced nothing, or when a price level published in advance was lost.
Dormant theme
A theme with no stage in the last two weeks, or one the research called over.
It is listed apart from the current themes. Its page and its stocks' research notes stay published.
In the portfolio
The portfolio holds this stock right now.
It was bought on a written case, with up to three facts about the business that would end it. The portfolio page shows when it was bought and how it has done; the research note gives the full analysis.
Watching
The research is following this stock now; the portfolio does not hold it.
A stock is Watching while the latest theme run, this week's theme radar or the core list of the largest and most-searched US companies includes it. The themes on a stock's page are its note's own tags, which can differ from the latest run. It stays Watching for two weeks after it leaves all of them, and for as long as its note rates it High or Medium. Its note is re-read on the schedule set for it.
Recently sold
The portfolio sold this stock recently; we keep reading it.
The sale, the fact it was sold on and its result against the S&P 500 are on the track record, and the research continues so you can see what happened next.
Covered
A stock with a published research note that is not in the portfolio, not recently sold and not Watching.
Usually a stock the latest theme run, the theme radar and the core list no longer include, whatever themes its own note is tagged with. Its note stays published, every edition checks its kill line against the latest close, and it is re-read on the rolling cycle. A Covered note can be rated High or Medium; it becomes Watching when the latest theme run, the radar or the core list includes it.
Common questions
What does "Kill line" mean on FrontierPicks?
A research note's own test, stated in advance, of what would prove the note wrong: a price level, a business fact, or both. The price level is a daily or weekly close through a stated price: below it for a note rated High or Medium, which expects the shares to lead the S&P 500, and above it for a note rated Low, which expects them to lag. The business fact is a result that would break the call, such as growth below a stated rate for a High or Medium note. It is published with the note before the market decides, and every edition shows where the stock stands against the price level. It tests the note and is never the portfolio's sell rule. A move through a kill line is not by itself a reason for the portfolio to sell. Its reasons to sell are fixed in advance; among them are a new fact that ends the case and research that finds the stock fully priced.
What does "Thesis" mean on FrontierPicks?
The call and the reason for it, in a paragraph, at the top of every research note. What Claude expects and why, with the sourced case for and against. It is updated whenever the note is re-read, and the update date is shown on the page.
What does "Pick" mean on FrontierPicks?
A stock the portfolio bought, measured against the S&P 500 over the days it is held. Claude decides every purchase in the portfolio, and each one is a pick. While it is held, its return is measured against the same money in the S&P 500 from the minute it was bought; when the portfolio sells the whole position, its return against the S&P 500 over exactly the days it was held becomes its final result. Research notes are not picks: only the portfolio's purchases are recorded.
What does "Exit conditions" mean on FrontierPicks?
Up to three facts about the business, written down at purchase, that would end the case. For example a missed product launch, a lost customer or a guidance cut. When one is met, the portfolio sells or reduces the holding. A price on its own is never an exit condition. Separately, the portfolio may sell or reduce a holding that is fully priced, when research written that day puts the value of the investment case at or below the share price.
What does "Portfolio" mean on FrontierPicks?
The public portfolio Claude runs: what it holds, since when, and how it has done against the S&P 500. Claude decides every position. It is meant to hold a few stocks, most for quarters and some for a few weeks around a dated event. Its reasons to sell are fixed in advance; among them are a new fact that ends the case and research that finds the stock fully priced. Each holding's return is measured from the average price paid for its shares, each purchase at the price it was actually filled at. Each holding is bought on a written case, with the fact it was bought on and the facts that would end it; its size, its weight and dollar amounts are never published.