Your X thread
Paste your X thread. Nothing to upload.
Make a financial analysis
Every figure, with the author who posted it and whether they showed their source.
Free to try, no card. Your X thread opens in Ancher.
Paste your X thread. Nothing to upload.
Every one of these stays separately addressable instead of collapsing into one block of text.
Figures checked against the sentence or table that supports them.
Financial numbers on X arrive two ways: typed into a post, where they have been rounded for effect, and screenshotted from a terminal or a filing, where they have not. Ancher collects both, marks which is which, keeps each figure attached to the author and the post it came from, and separates figures backed by an outbound link from figures that arrived with nothing behind them.
The honest version
By the time a number reaches a feed it has been through at least one person's summarising, usually from memory and often to make a point. Threads are frequently right and frequently first, and neither of those makes them citable. The analysis exists to produce a list of things worth verifying against the primary, not to replace the primary.
One post in one thread, at post 3/9, read two ways.Illustrative. The figure is invented; the gap it falls into is not.
3/ people keep quoting the wrong figure. the actual number is
In the attached terminal screenshot — not readThe specific never reaches the draft
Lands in the draft, with its timecode
What comes out
Figures checked against the sentence or table that supports them.
The figures the threads lead with, each marked typed or screenshotted.
One line per figure: the author, the post number, and the link behind it if there was one.
Only comparisons an author made themselves, never a series assembled across posters.
Caveats and hedges as written, including the ones buried in a later post.
The outbound links, which is where the actual filings and datasets are.
The link on its own gets a summary. This is the instruction that produces the 5 sections above, in that order, with the rules that keep them honest. Paste it with your X threads — in Ancher, or in whatever assistant you already use.
Extract a financial analysis from these threads. (1) List headline figures, each marked TYPED or SCREENSHOT. (2) For every figure give the author's handle, the post number, and the outbound link if there was one. (3) Record only comparisons an author made themselves; do not assemble a series across posters. (4) Capture caveats wherever in the thread they appear. (5) List the outbound links as sources. Do not annualise, convert, extrapolate or calculate anything.
The detail, if you want it
| From the X thread | Into | Why |
|---|---|---|
| Attached images | Headline figures | The precise figure is usually in a screenshot of a terminal while the typed version is rounded, so the image is the more accurate of the two. |
| Author and profile | Evidence per figure | Who posted a number is part of the number's provenance, and on a feed that is often all the provenance there is. |
| Outbound links | Sources | An outbound link to the filing turns a quoted figure into something a person can check in a minute. |
| Posts in order | Risks | Hedges tend to arrive several posts after the headline claim, where a reader who quoted post one never sees them. |
| Likes, replies, reposts | Risks | A figure that drew heavy replies was disputed, and that dispute is a risk worth recording next to it. |
A thread is a sequential argument; read out of order it stops meaning anything.
On X the chart usually carries the claim while the text carries the opinion, so the image is the part worth keeping.
Whether the author builds the thing or sells against it changes how the claim should be weighted.
Threads cite primary sources more often than they get credit for, and those links are the way past the hot take.
Reach separates a thread the field actually argued about from one nobody read.
You can build the list of things to verify, which is the half that saves time. Every figure comes with its author, post number and link status, so checking a number against the primary takes a minute rather than an afternoon.
Because on a feed it is frequently the only provenance a figure has. The same number from a named analyst and from an anonymous account should not enter your thinking with the same weight, and the analysis refuses to flatten that.
Usually, and that is exactly the risk. Speed on a feed comes from skipping verification, so the analysis is built to make the unverified figures visible rather than to let them blend in.
Everything you save lives in one workspace, so the financial analysis is built from your sources — not from a model's memory of the internet.
Open Ancher →Not an X thread? The same financial analysis also comes from YouTube video, PDF, recording.