Research cutoff: September 25, 2026. This guide explains a research process rather than judging any current stock. The company and figures in the example are fictional. Nothing below is a recommendation to purchase or sell a security.
Confirmation bias in investing is the habit of giving more attention or weight to facts that fit a favored thesis while minimizing facts that challenge it. The risk is not that you have an opinion; an investment thesis requires one. The risk is that a thesis becomes impossible to revise. A short, dated evidence checklist can reveal whether your conclusion still follows from the filings and the price you would pay.
Write the claim before collecting support
Start with one falsifiable sentence: “I expect this business to improve operating cash flow because customer demand and margins are improving.” Write the time frame, the evidence that would support it and at least one observation that would weaken it. “The company is innovative” is too broad to test. “Reported revenue is growing while receivables and cash collections remain consistent with sales” gives you something to compare with the next filing.
Next, separate three columns: verified fact, interpretation and unanswered question. A reported revenue number is a fact for a defined accounting period; “demand is durable” is an interpretation. A news story repeating management’s comment is not a second independent fact. Date every source. A thesis based on a 10-K from last year may need updating after a 10-Q or a material 8-K, even if the older report still appears high in search results.
Use filings to look for both sides of the case
Investor.gov’s EDGAR guide explains how to locate free public-company filings and recognize form types. For a US-listed operating company, start with the latest 10-K and 10-Q, then check whether a more recent 8-K changes a material fact. The SEC’s 10-K/10-Q guide identifies the Business, Risk Factors, management discussion and financial statements as different sources of evidence. Risk Factors describe important uncertainties; management discussion explains changes in results and liquidity; the statements show the amounts.
Read the positive case and the risk disclosure on the same day. If a company reports faster sales, ask whether gross margin, operating cash flow, receivables, inventory and debt support the implied business improvement. If the company has changed the definition of a preferred metric, reconcile it with the prior period before comparing percentages. A business can have a strong quarter and still be an unattractive investment at a given valuation. Conversely, one adverse number does not automatically disprove a multi-year thesis.
FINRA’s stock-evaluation guide explains that professional research may disclose conflicts, while online commentary may have fewer protections. The joint SEC-FINRA social-sentiment bulletin warns that social data can be stale, incomplete or misleading. Repeated posts from the same original claim are not independent confirmation. Check whether a promoter owns the security or gains from attention before treating a confident prediction as evidence.
A hypothetical test: growth with weaker cash collection
Imagine a fictional company, Northline Tools, reports that annual revenue rose from $100 million to $120 million. An enthusiastic shareholder highlights the 20% increase and a new customer announcement. In the same reporting periods, operating cash flow falls from $18 million to $4 million and receivables rise from $15 million to $32 million. These invented figures do not prove fraud, nor do they prove that the company will fail. They do create questions about the timing and quality of sales collections.
A balanced research note would record the $20 million revenue increase and the weaker cash-flow conversion side by side. It would look for explanations in the cash-flow statement and management discussion, check customer concentration and payment terms, and compare with prior periods. The note might conclude that growth is real but currently consumes more working capital, or it might find a temporary timing effect. The point is to specify the next evidence needed instead of discarding an inconvenient line item.
Reverse the exercise if you are bearish. An investor convinced a company must fail can also ignore falling debt, better margins or a refinancing that resolves a near-term obligation. The method is symmetrical: look for the best evidence against your preferred view, then decide whether it materially changes the thesis.
An evidence checklist before you act
- State the thesis: Which measurable business change, valuation assumption and time frame are you relying on?
- Find the primary record: Which current filing or company disclosure supports the central claim, and what is its date?
- Find the strongest counterpoint: What fact in the same or a later filing would a thoughtful skeptic emphasize?
- Check independence: Are several articles repeating one press release, analyst note or social post?
- Inspect incentives: Who wrote the claim, and are any financial or promotional interests disclosed?
- Set a revision rule: What specific future disclosure would cause you to reduce confidence, pause or abandon the thesis?
Keep the worksheet short enough to use. A research process does not require treating every negative headline as equally important. It requires giving a material contradictory fact a fair chance to change the conclusion. Save your original note so you can see whether you applied the revision rule consistently rather than rewriting the rationale after the price moved.
Our investing decision journal helps track decisions over time; this page focuses on which evidence deserves weight before the decision. The FOMO checklist addresses pressure from a fast price move, while the earnings-report guide gives a starting structure for checking the reported numbers.
Frequently asked questions
Is reading a bearish opinion enough to remove confirmation bias?
No. Check whether the objection uses current evidence, addresses the investment thesis and could change your decision. Merely collecting opposing opinions is not a substitute for testing facts.
What is the best primary source for a public-company claim?
The relevant company filing is a useful starting point. Use the latest 10-K, 10-Q or 8-K as appropriate, then compare the claim with the filing date, definitions and financial statements.
Does a disconfirming checklist tell me when to sell?
No. It helps identify what has changed. A decision also depends on valuation, position size, time horizon, alternatives and personal circumstances.
Educational information only. Research can reduce avoidable errors but cannot remove uncertainty or guarantee an investment outcome.