Welcome to Valarn#
If you've just signed up, this guide walks you through your first complete AI research report — from adding a stock to reading the results. By the end you'll know how to run a report and how to read the research view, the confidence score, and the debate between the agents.
Step 1: Add a Stock to Research#
Valarn researches individual stocks, one ticker at a time — there are no watchlists to set up first. There are three easy ways to get a stock in:
- The onboarding wizard (new accounts). On your first visit, click Run my free AI analysis, type a ticker, and Valarn runs a free report for you. It uses your one free report for the month.
- Start new research. Type any ticker or company name into the "Start new research" bar. Valarn automatically starts tracking it — you don't have to add it anywhere first.
- The Stocks page. Click Stocks in the left sidebar to add and manage the tickers you follow.
Any ticker you type is created as a tracked stock automatically, so you can go straight from a symbol to a report.
Step 2: Choose Your Report Type and Options#
Open Analysis in the sidebar — the drawer is titled "Run Stock Research." Pick your stock, then choose a Report Type:
- Quick — Market + News agents. Ready in about 30–60 seconds for roughly half a research credit. A fast first snapshot.
- Standard — up to ~10 agents. Ready in about 2–3 minutes for 2 research credits. The everyday choice.
- Deep Debate — the full roster of up to ~25 agents with a complete bull-versus-bear debate. About 4–6 minutes and 3–8 credits depending on the AI routing. The deepest research.
Costs are measured in research credits. On free and report-based plans the unit shows as "reports" instead — for example, "1 report."
A few optional settings, all with sensible defaults:
- Analysis Style — presets such as Balanced (the default), Technical Focus, News Focus, Risk Review, Earnings Review, or Valuation Review. The preset decides which agents lead; some styles are available on higher plans.
- AI Routing (Pro and above) — choose a speed/quality band (Fast, Balanced, Premium, Premium Full, or Local) and a provider (GPT, Claude, Grok, Gemini, or DeepSeek). Free plans use default routing.
- Risk Profile — Conservative, Balanced, or Aggressive.
- Advanced — Debate Depth (1–3 rounds) and Ensemble Runs (run the whole analysis 1–3× and merge the results by majority vote to reduce randomness — this multiplies both credits and runtime). Active Analysts lets you switch specific agents on or off.
Two things worth knowing: there is no "time horizon" or hold-period setting, and Ensemble is a separate multiplier, not a "Standard vs. Ensemble" choice. You can run a Standard report at 3× ensemble whenever you want a more stable read.
Step 3: Run It#
With your stock and report type selected, click Run Standard Report (or Run Basic Report / Run Deep Debate). A badge shows the exact credit cost before you commit. When the run finishes, the report opens on its own page under Reports.
Step 4: Reading the Research Report#
Valarn produces research, not trading calls. Here's what to look at first.
The Research View#
At the top you'll see Valarn's overall research view — one of Bullish, Cautious Bullish, Neutral, Cautious, or Bearish. This is a research posture, not a buy / sell / hold recommendation.
You may still see the words "buy" or "sell" elsewhere in the report — on Wall Street analyst ratings, insider filings, or Congressional trades. Those are other people's actions, clearly labeled as third-party facts, never Valarn's own view.
Confidence and Agreement#
Two different 0–100 scores sit next to the view, and they mean different things:
- Confidence — how complete and reliable the underlying data is (source coverage and freshness). Bands: High (80+), Moderate (60–79), Low (40–59), Insufficient (below 40).
- Agent Agreement — how strongly the agents converge on the same view.
Read both. High agreement with low confidence — or the reverse — tells you something on its own. Neither number is a price prediction.
The Executive Summary#
The summary band gives you the three things that matter fastest: what supports the view, what works against it, and the single biggest risk.
The Agents#
A Standard report runs up to about ten agents. Deep Debate runs the full roster of up to ~25 specialist agents across five categories — Core Research, Market Structure, Debate & Risk, Financial Quality, and Events/Sector & Macro. Each agent contributes a stance (bullish, neutral, bearish, mixed, or unavailable), a short summary, and a data-quality note, all shown as agent run cards and a full reasoning transcript. Higher-quality evidence is weighted more heavily than a simple head-count.
The Debate Summary#
This is where the research earns its keep. It lays out:
- Bull case emphasized — the strongest supportive arguments
- Bear case emphasized — the strongest counterarguments
- Research Manager decision — how the synthesis weighed the two sides
Take your time here. The disagreement is often more useful than the headline view.
Scenario Range#
Instead of price targets, Valarn shows a Scenario Range (bear / base / bull reference levels), a Reference Price, and a Risk Level. These are educational modeling inputs — not entry points, targets, or stop-losses.
Step 5: Save, Compare, and Re-Run#
Every report is saved under Reports, so you can revisit it any time. Use Compare Runs to see how the research on a stock has shifted over time, and re-run the analysis whenever something material changes — earnings, major news, or a large price move. Research goes stale; a report from last week may not reflect today.
Tips for New Users#
- Don't anchor on the view alone. The confidence score and the debate summary carry just as much signal.
- Read the disagreement. When the agents split, that tension usually points to the real question.
- Re-run before you rely on it. A report is a snapshot of a moment, not a standing call.
- It's research, not advice. Valarn is built to help you think. Every report is educational only and is never a recommendation to buy, sell, or hold.
Welcome to Valarn. Our team is available any time via Training & Support in the app.
Valarn
Customer Success
Valarn Research Team
