- 2,600+ US companies screened daily
- A few flagged
- Every flag audited
Read the company,
not the crowd.
Each morning our engine screens the US market and publishes research on the few companies that pass: a plain-English memo, the numbers behind it, and the earnings call, ownership and sentiment in one place. Then the curator audits every flag: how strong the thesis is, where it could break, and what would change his view.
Flags publish automatically, before anyone reviews them. The audit comes after.
Free during beta · No credit card · US residents 18+
Written after the flag publishes. The structure every audit follows:
- 01Thesis strength, with a confidence level
- 02The case, in plain English
- 03What the pitch leaves out
- 04How I tried to break it
- 05The three tests
- Is the downside limited by something real?
- Is the upside meaningfully larger, with a way to get there?
- Is there a reason the market has it wrong?
- 06What would change my view
Most stock research is either hype with no rigor, or raw data with no judgment.
Ours works in three steps, and it shows its work at each one.
Screen the market
Every morning, seven days a week, fixed rules screen 2,600+ US companies for size, liquidity and the numbers. The rules are set in advance and don't change with the news.
Research the few that pass
The top-scoring companies get an 8-section memo in plain English: the business, the numbers, the catalysts, the risks, and both sides of the case.
Publish, then audit
Every flag is published automatically. No one screens the engine's flags for you first. What the curator does comes next.
Then the curator goes to work.
He audits every flag, tries to break the idea and publishes what he finds, whatever the rating. The ideas that hold up go to the Watchlist, and the best of those to Curator's Highlights.
And then we measure whether his judgment was worth anything →What's inside every flag.
One company, researched four ways, then audited. These are real screens from the platform.
$BSX · Curator's Audit · 5 October 2026
Thesis Strength: Neutral · confidence: low. Boston Scientific still throws off real cash and looks cheap if its two troubled franchises stop shrinking, but I can't yet say what the market has wrong. Confidence is low because a cyberattack has withdrawn guidance and an $11 billion cash deal is about to land, and the 28 October update quantifying both would raise it.
What the pitch leaves out. On 25 August hackers disrupted manufacturing, ordering and shipping. Operations are largely back, but in early September the company said it was unlikely to meet the July guidance and would update on 28 October. So the pitch's catalyst is now a quarter with a known hole in it. …
What changed: Management cut its 2026 outlook because US WATCHMAN demand slowed sharply and EP lost share to competitors, and it warned that 2027 will stay weak.
Tone: Confident on the base business and the 2028 pipeline; apologetic and cautious on WATCHMAN, EP and 2027.
Pressed on: Analysts pressed on whether the reset is conservative enough, on WATCHMAN referral trends, and on 2027 growth and EPS; management gave only a directional 2027 view, no figures, and promised more detail in January.
“I do take responsibility for the guidance reduction”
— CEO, analyst Q&A“we did undercall the competitive pressures in the U.S.”
— CEO, analyst Q&AAuto-summarized from the earnings call · quotes are short excerpts shown for commentary, not a transcript.
Insider and institutional activity — led with the read.
13F · Q2 2026 · ~45-day reporting lag
154 new funds initiated positions and insiders net-bought $9.5M last quarter — the accumulation behind the flag.
Institutional figures are 13F — self-reported and lagged ~45 days. Insider figures are open-market buys and sells only.
News sentiment is an input, not a verdict. It's one lens on the flag, weighed alongside fundamentals and smart money.
Net sentiment · 30d
Coverage
Trend
30-day sentiment trend · daily net score
Headline breakdown
Aggregated from news-sentiment scoring (Alpha Vantage) — headline signal only, not a recommendation.
Example screens: Boston Scientific (BSX), flagged Oct 9, 2026. We may hold positions in securities we cover and do not publish which.
Asymmetric Opportunities
A few top-scoring flags a day from a market-wide screen, each with the memo, the three tabs and a Curator's Audit.
Curator's Audit
How the curator tried to break the idea, the three tests it had to pass, and what would change his view.
Watchlist Intelligence briefs
Follow-ups on names on the curator's Watchlist, produced under conditions set in advance.
Every flag scored on four parts, all of them shown.
Sentiment, fundamentals, and insider and institutional activity, side by side.
The market closes. The screen doesn't.
Every past flag, by ticker, name or keyword.
The engine finds. The curator chooses. We publish both records.
Whether the choosing adds anything is an open question. Here is the number.
Every name we publish carries two clocks. The first starts when the engine flags it, and measures the engine. The second starts only if the curator promotes it to Curator's Highlights, and measures him. We publish both, side by side.
Performance from the moment the engine publishes a flag. This is the screen's record, and the curator has no hand in it.
Performance from the moment the curator promotes a flag. He can only make it worse than the engine's, or better.
The engine's record
since July 7, 2026 · 108 daysEvery name the engine has flagged, measured from its first flag. One per ticker, never removed.
111 flagged · 1 too recent to measure, excluded above, kept in the record.
Held on a clock
The record above never sells, so a flag that ran and gave it back reads as flat. These rows close each flag a set number of calendar days after it was raised.
| Held | n | At horizon | Still open | Difference | vs SPY |
|---|---|---|---|---|---|
| 30 days | 93 | +2.4% | +1.7% | +0.7pp | +0.6pp |
| 60 days | 69 | +0.8% | +2.6% | -1.8pp | -1.6pp |
| 90 days | 27 | +3.7% | +6.6% | -2.8pp | +0.3pp |
| 180 days | 0 | — | — | — | — |
| 365 days | 0 | — | — | — | — |
| never closed | 110 | — | +2.0% | — | +1.5pp |
If since added trails since first flag, you'll see it on the same screen, in the same table. We'd rather publish a number that can embarrass us than a claim that can't be checked.
Editorial record, not a portfolio or advice. Hypothetical, equal-weighted returns; removed names stay in. SPY for reference only. Every rate shows its n. Past results, not a forecast. Figures as of October 23, 2026.
For investors who want to understand what they own.
The reasoning behind each flag, in plain English, and an audit that tests how strong the idea is. What you do with it is up to you.
Self-directed investors
The reasoning behind each flag, in plain English, without the scanning.
Analysts and professionals
A second read on a company in minutes, with the numbers shown.
Simple pricing. Start free.
No trial that quietly converts. No retention call if you leave. Full access, free, while we're in beta.
Full access during beta. No credit card. Your first flags are waiting the moment you sign in.
Read today's flags free →US residents 18+
Join before the paywall and lock the founding rate: $25 / month or $240 / year, while your membership stays active.
Become a founding member- Renews at your plan's rate until you cancel in settings.
- Refunds within 30 days, no questions asked.
- Founding rate closes when the paywall goes up.
Answered plainly.
Is this financial advice? +
No. Ticker Mavericks is an educational research publication of general circulation. Nothing here is a recommendation to buy or sell any security, or advice personalized to your situation. You always do your own research.
Do you touch my money? +
Never. We publish research. We don't manage money, take positions on your behalf, or connect to your brokerage.
Do you own the stocks you write about? +
We may, and we don't publish which. Assume a conflict of interest exists in everything we publish. Our Conflicts of Interest statement sets out the limits our founder has voluntarily placed on his own trading.
How is this different from an AI stock screener? +
A screener hands you raw data and leaves you to it. Every flag here arrives with the research done: the memo, the earnings call, the ownership and the sentiment. Then a person audits it after publication and publishes what he finds, whatever the rating. A screener is a tool. This is a research desk.
What do I get each day? +
A few flags, each with an 8-section memo, the Transcript, Ownership and Sentiment tabs, and a Curator's Audit that follows, plus Watchlist Intelligence briefs on names already flagged. Seven days a week. Nothing disappears: the History page holds a searchable archive of every past flag.
What's the engine? +
The automated system behind the research, powered by the XvoroAI engine. It screens the market, scores what passes and writes the memos, then publishes on a fixed schedule. No one reviews a flag before you see it, and that is deliberate: it lets one desk cover the whole market, and it lets the curator's record be measured against the engine's honestly. He reads the engine's flags when you do.
What kind of companies does it cover? +
US companies between about $180 million and $700 billion in market value, from small firms to some of the largest in the market. What they share is that their numbers cleared the screen's rules that morning. The smaller ones are more volatile, less liquid and quicker to fall, so the risk is real.
Can the AI be wrong? +
Yes. The agents that write the memos can misread data or state things that aren't so, and the market is sometimes right about a company it doubts. That's why every flag gets an audit, and why we publish the reasoning so you can check it. Do your own research on every name.
What does it cost? +
Free during the beta. At launch it's $29/month or $290/year (2 months free). Join before the paywall and you lock the founding rate, $25/month or $240/year, while your membership stays active. Plans renew at your plan's rate until you cancel in settings, and refunds are available within 30 days.
Who can subscribe? +
US residents aged 18 or older, for now.
You don't need a bigger research budget. You need a research desk.
Ours runs every morning, seven days a week, whether or not anyone's watching.
Read today's flags free →Free during beta · No credit card · US residents 18+