For the analyst working solo
The independent's research desk.Filings, flow, and macro on one surface.
Sell-side analysts got a terminal. Independents got a browser bookmark bar. Convexity is the terminal built for the second group, and every number in it carries the primary source it came from.
The job of the independent analyst hasn't changed much in twenty years. Read the 10-K. Diff the risk factors against last year's. Listen to the conference call. Check the insider ledger. Triangulate macro. Write it up with citations a reader can actually follow back¹. It is, still, a craft.
The tools around that craft, however, have calcified into an absurd stack. EDGAR full-text search surfaces the wrong filing. PDF tables get copied into Excel by hand. Transcript archives get scrolled looking for one quote you half-remember. A Form 4 feed sits in one tab, a FRED chart in another, FOMC speeches in a third. At the end of the day, a Google Doc collects every citation re-pasted manually so the footnotes will hold.
Sell-side analysts got a terminal to collapse that stack. Independents got a bookmark bar. Convexity is built for the second group, not as a Bloomberg replacement, but as a research surface designed around the actual shape of the work².
“The job hasn't changed. The tools have. Read the 10-K, diff it against last year, listen to the call, check the insider ledger, triangulate macro, write it up.– the thesis of this page, in one line
A typical morning, set against the same morning on Convexity. Same job, two different shapes of effort.
A morning in twelve tabs
- EDGAR full-text search for the 10-Q you actually want.
- Copy tables out of a PDF and into Excel by hand.
- Scroll a transcript provider for one quote you remember.
- Cross-check a Form 4 against last quarter's cluster.
- Open a second browser for FRED, a third for FOMC speeches.
- Re-paste every citation into a Google Doc that night.
The same morning, one workspace
- EDGAR XBRL normalised across 10-K, 10-Q, and 20-F periods.
- 8-K timeline with FinBERT sentiment per event.
- Form 4 ledger with cluster and role weighting.
- FedSpeak NLP for policy tone and language shifts.
- Patent filings indexed by assignee, hiring signals alongside.
- Draft memo with every claim cited back to its source.
- 08:00
Morning brief.
Overnight filings, insider moves, macro prints. Auto-curated at /market-brief.
- 09:30
Open.
Scan the 8-K timeline with FinBERT sentiment for every name on the watchlist. One surface, one scroll, sentiment attached per event.
- 11:00
Cross-check.
Form 4 cluster vs. insider ledger vs. 10-K risk-factor diff. Three sources, one ticker page, one thesis forming.
- 14:00
Deep dive.
XBRL compare across periods, plus patent and hiring cross-reference. Live at /sec.
- 16:30
Memo draft.
AI stitches primary sources into a cited draft. Luna / Sonnet / Opus routed by tier at /research.
- 19:00
Distribute.
Shareable memo link (Pro Plus) or PDF export. Send a clip, not a file. Citations resolve back to the exact paragraph.
Three surfaces, each designed around a specific part of the craft.
10-K to model in an afternoon.
Every SEC filing a company has ever made is indexed and XBRL-parsed, so three periods of income statement can be normalised and compared without anyone touching Excel. 8-Ks are laid out as an event timeline with FinBERT sentiment scoring. You see inflection points, not a stack of PDFs.
Pro Plus accounts get semantic search across the entire filing corpus, powered by Voyage embeddings. "Find every time this management team described margin pressure in the last five years" becomes a query, not a project. Risk-factor diffs flag additions, deletions, and tone shifts year-over-year. The AI annotates the change, not the paragraph.
Insider, 8-K, patent, hiring on one ticker page.
An equity thesis rarely lives in the income statement alone. Convexity wires Form 4 insider flow, 13F holdings, congressional trades, USPTO patents, H-1B talent flows, ATS job postings, and 8-K material-event tracking into the same ticker view. The alt-data sits beside the financials rather than inside a separate subscription.
Insider signal intelligence reads transactions with context: cluster detection, officer-role weighting, transaction-type classification. A CEO open-market purchase doesn't look like a 10b5-1 plan sale. FedSpeak NLP layers FOMC speeches, minutes, and FRED series on the same surface, useful for the days your sector thesis is secretly a rates thesis.
A model row that already knows the answer.
The ticker page lands with the numbers most modelers stitch together by hand: trailing and forward PEG, analyst price targets, the full grade table, per-quarter EPS surprise history with beat/miss streaks, revenue segmentation by product and geography, and a DCF baseline. The peer set auto-selects from FMP's sector cohort with a yfinance fallback when coverage is thin. No manual peer hunting.
The proprietary signal is Convexity Score, a capture-ratio asymmetry vs SPY computed over a trailing 252 days, refreshed nightly. It separates names that participate in upside and reject downside from names that do the opposite. Methodology is documented in full at /methodology/convexity-score. No black box.
Memo drafts that cite their sources.
Convexity's AI stack is routed by tier: GPT-5.6 Luna on Free, Claude Sonnet on Pro, Claude Opus on Pro Plus, with cross-vendor fallbacks for uptime. Retrieval runs on Voyage embeddings over the filing corpus, so the model answers from primary sources rather than guessing. Every claim ships with a citation. If the citation isn't there, the claim isn't made.
Pro Plus also includes 20 deep-research runs per month: long-running, multi-source synthesis jobs that span the 8-K event corpus and a curated news feed alongside filings. A three-paragraph risk-factor comparison cross-checked against a two-week-old insider cluster becomes one prompt instead of a forty-minute session on EDGAR.
Convexity does not replace a Bloomberg-class terminal for a rates trader. It doesn't try to. But most independent analysts aren't trading MMDs, they're writing research, and the value delta flips hard.
Bloomberg-class terminals
- Real-time fixed-income, FX, and OTC pricing you likely don't need.
- Per-seat contracts, annual commits, procurement friction.
- Chat and messaging network effects irrelevant to a solo analyst.
Convexity Pro & Pro Plus
- The filing, insider, macro, and alt-data surface you actually use.
- $29 or $79 per month. Cancel from the settings page.
- AI memo drafting with citations, which the terminals don't have.
EDGAR plus six micro-SaaS
- Free EDGAR has no sentiment layer, no normalised XBRL, no semantic retrieval.
- Six micro-SaaS tools that don't talk to each other.
- Every citation is manual; every number lives in a different tab.
Convexity Pro & Pro Plus
- One surface, one ticker page, one bill.
- A documented REST API and MCP server (Developer & Quant tiers) to feed your own model.
- Shareable memo links on Pro Plus. Send a clip, not a PDF.
Two tiers, both honest about their quotas. See the full pricing page for every feature flag. The numbers below are the ones that shape the workflow.
$29 / month · $290 / yr
Claude Sonnet · 150 AI queries / mo
Real-time market data, real-time insider flow, 25 watchlists of 250 tickers each, 25 smart alerts, morning brief, command palette, options flow, GEX charts, 3 narrative threads, unlimited history. Enough surface for a single-analyst research practice.
Start free$79 / month · $790 / yr
Claude Opus · 1,000 queries + 20 deep-research runs
Everything in Pro, plus unlimited watchlists, unlimited alerts, unlimited workspaces, unusual flow alerts, backtest, shareable memo links, semantic search, and unlimited history.
Start freeResearch faster. Cite everything.
Create an account and you land on the same surface we use ourselves. Free tier is real: Luna, 10 queries / month, delayed quotes. No card required.