What is the difference between Core and AI Hybrid?

Core is the primary rules-based view derived entirely from reported financial data. AI Hybrid cross-checks Core shortlisted companies against live data and current context, applying news checks and multi-source evidence validation as a second layer.

What does the Confidence score mean?

Confidence reflects how complete and consistent the underlying data is for that company. High confidence means the model had full, clean inputs. Low confidence means important inputs were missing or inconsistent — the rating may still be valid but warrants extra caution.

What is the Pick Score?

Confidence Score reflects how strongly a company passed the multi-source validation and news checks in the AI Hybrid view. A higher score indicates stronger conviction across multiple evidence sources.

How is Fair Value calculated?

Fair Value is an intrinsic value estimate derived from reported financial data using the Stedrok methodology. The Discount column shows how far the current price is below that estimate as a percentage. A 20% discount means the stock is trading 20% below the model's fair value estimate.

Can I export the ranked stock list?

Yes. Pro members can use the Export CSV button above the ranked table to download the full filtered list for offline analysis, spreadsheet tracking, or portfolio review.

How to Use the Stedrok Dashboard

A practical walkthrough for getting research value from the Core and AI Hybrid views in under five minutes per session.

Step 0: Choose a Dashboard View

Core is the primary rules-based view. Every score is derived from reported financial data using the same transparent ruleset. Start here.

AI Hybrid is a separate validation layer. Companies shortlisted by Core are cross-checked against live data and current context. Use AI Hybrid as a second confirmation, not a replacement for Core.

AI Hybrid provides an AI-assisted validation layer with live news checks, multi-source evidence, and confidence scoring. Use it to find candidates with strong cross-source conviction.

Step 1: Apply the Right Filters First

Before reading individual rows, narrow the table to what's relevant for you:

Research State: Start with BUY only. WATCH candidates are valuable for building a watchlist but shouldn't be your first action.
Country/Region: Narrow to markets you can access through your broker before spending time on ideas you can't act on.
Sector: If you have sector convictions (or restrictions), apply them here.
Market Cap: Filter by size if your strategy targets small-, mid-, or large-cap stocks specifically.
Minimum Score: The default shows all BUY candidates. Raising the minimum score threshold narrows to highest-conviction ideas.

Step 2: Read Confidence Before Anything Else

Confidence tells you how complete and consistent the underlying data is for that company. High confidence means the model has full, clean data to work with. Low confidence means important inputs were missing or inconsistent — the rating may still be valid but treat it with more caution.

Practical rule: if confidence is low and the name is unfamiliar, move on. Come back to it when confidence improves.

Step 3: Read the Four Pillar Scores

Each column tells a specific story:

Value Score: How cheap is it relative to intrinsic estimates? Higher = more discounted.
Quality Score: How good is the underlying business? Higher = stronger ROIC, margins, earnings stability.
Risk Score: How resilient is the balance sheet? Higher = lower debt, better coverage ratios.
Dip Score: Is the price at an interesting entry point in its recent history? Higher = further from recent highs.

A genuine BUY needs all four to be constructive. If Quality or Risk is weak, investigate why before acting — it may be a value trap.

Step 3b: Fair Value and Discount

Fair Value is the model's intrinsic value estimate for the company, derived from reported financial data using the Stedrok methodology. Discount shows how far the current price sits below that estimate, expressed as a percentage. A 20% discount means the stock is trading at 20% below the model's fair value estimate.

Note that fair value estimates are model outputs, not guarantees. Use them alongside the four pillar scores to assess whether the apparent cheapness has quality and resilience behind it.

Step 4: Use Ticker Insights as a Starting Point

Click any row to open the Ticker Insight panel. This shows a plain-language summary of the model's reasoning for that company. Read it as a hypothesis, not a conclusion. Your job is to verify or refute it.

The insight tells you: the main driver of the rating (is it cheap? high quality? resilient?), the sector context, and any notable limitations in the evidence.

After reading the insight, your next step is always: check recent earnings, read the last annual report summary, and check broker availability in your region.

Step 5: Export and Track (Pro)

Pro members can use the Export CSV button above the ranked table to download the full filtered list. Useful workflows:

Drop the CSV into a spreadsheet and add your own notes column.
Track rating changes over time — a company moving from WATCH to BUY is often more interesting than one that has been BUY for six months.
Build a personal shortlist from consecutive weeks of BUY candidates to identify persistent value ideas.

Common Mistakes to Avoid

Don't act on a BUY rating alone. It's a shortlist tool, not a buy signal.
Don't ignore WATCH candidates. Many move to BUY when entry conditions improve — building a watchlist from WATCH is good process.
Don't filter out unfamiliar geographies automatically. Some of the best value opportunities are in markets most investors ignore.
Don't confuse low price with cheap. The Value Score measures discount to intrinsic value, not absolute price level.