Skip to content
BullOrBS

How BullOrBS Works

How Our AI Works

BullOrBS uses a staged AI workflow rather than one model doing everything. Planning, web research, article writing, and verification each have separately configured models and API keys. The research stage gathers a source packet; the writer is restricted to that packet.

We publish the evidence, comparison tables, uncertainties, and conclusion. We do not present private model reasoning as evidence. A draft remains private until it passes the quality gate and receives human approval.

Data Sources

Our AI pulls from a wide range of publicly available sources:

  • Market data — real-time and historical price action, volume, and technicals
  • Analyst reports — consensus ratings, price targets, and estimate revisions
  • News & sentiment — recent headlines, earnings calls, and market commentary
  • Regulatory filings — SEC (US) and SEDAR+ (Canada) filings, insider transactions
  • Company financials — revenue, earnings, margins, debt levels, cash flow

The Scoring System

Every stock recommendation we audit gets a score from 1 to 10. Here's what each range means:

8-10

Strong. The data backs it up, the timing makes sense, and the risks are manageable. This is solid advice worth paying attention to.

6-7

Decent.There's a case here, but there are real concerns or better alternatives out there. Do your own homework before jumping in.

4-5

Weak. More questions than answers. The recommendation has some merit, but the risks are hard to ignore. Proceed with caution.

1-3

Bad.The data doesn't support this recommendation. Key claims are misleading, the risks are too high, or there are much better options available.

The Elimination Tournament

For our weekly stock picks, we don't just name a winner — we show the full bracket. Here's how it works:

  1. Candidate selection — 10 to 15 stocks are identified based on market momentum, sector rotation, earnings catalysts, and reader interest.
  2. Head-to-head analysis — each stock is evaluated on fundamentals, valuation, growth potential, risk factors, and timing.
  3. Elimination rounds— weaker candidates are cut with full explanations of why they didn't make it.
  4. Final verdict — one stock emerges as the top pick, with a complete breakdown of every decision along the way.

Every eliminated stock is shown. Every reason is published. You see the full process, not just the conclusion.

Transparency Commitment

Every material data point in newly approved analysis must be sourced. We publish the evidence and decision framework instead of asking readers to trust a proprietary score. If the verifier cannot connect a claim to the saved research packet, the draft is blocked.

This is what separates BullOrBS from traditional financial newsletters. We believe transparency is more valuable than mystique.

Limitations

We believe in being upfront about what AI analysis can and cannot do:

  • AI can be wrong.Models can misinterpret data, miss context, or draw incorrect conclusions. That's why we show our work — so you can spot errors.
  • Past data doesn't predict the future. No analysis — human or AI — can guarantee stock performance.
  • This is not financial advice. BullOrBS is an educational and informational tool. Always do your own research and consult a licensed financial advisor before making investment decisions.
  • We are not affiliated with any financial institution, brokerage, or publication we analyze.

Use our analysis as a starting point — never as your only source.