# Confirmation bias

Confirmation bias is the tendency to seek, interpret, and remember information that supports an existing belief while discounting evidence against it.

## What it means

An investor can become attached to a view of a stock and give supporting evidence more attention than conflicting evidence. This is confirmation bias. It can affect which articles the investor reads, how they interpret an earnings report, and which facts they recall when reviewing a decision.

The bias is about **how evidence is handled**, not whether the initial belief turns out to be right or wrong. A well-supported view may be right; a poorly tested view may also turn out right by chance. Neither outcome alone shows whether the evidence was evaluated fairly.

## How it is calculated

Confirmation bias has no standard trading formula or universal score. A simple review tally can help reveal whether an investor investigated both sides of a proposed trade, but it does not measure the strength of the bias:

`Evidence balance = C / (C + D)`

- `C` = number of relevant, independently checked items reviewed that support the investor’s initial view, measured in items.
- `D` = number of relevant, independently checked items reviewed that challenge that view, measured in items.
- `C + D` = total relevant items reviewed. The ratio has no units and is undefined if no items were reviewed.

The tally counts items, not their quality or significance. Several articles repeating one claim are not necessarily several independent pieces of evidence. A high ratio can signal a need to look for counterevidence, but it cannot establish that the investor is biased.

## Worked example

Consider a **hypothetical** investor assessing a US stock before placing an order. Their initial view is that the company’s next earnings report will support their case for the stock.

1. The investor reviews eight distinct items: six that support the view and two that challenge it.
2. Using the illustrative tally, `C = 6` and `D = 2`, so `Evidence balance = 6 / (6 + 2) = 0.75`, or 75% supporting items.
3. One challenging item discusses weaker revenue growth. Rather than dismissing it because it conflicts with the initial view, the investor checks what period it covers and whether the underlying figure is comparable with the supporting items.

The 75% figure describes the review mix, **not** a 75% chance of success. It says nothing by itself about the stock’s value, future price, or whether the proposed trade is appropriate.

## Why it matters before you trade

Confirmation bias can leave an investor with an incomplete picture of a proposed trade. They may focus on favorable analyst commentary while overlooking unfavorable commentary, or emphasize encouraging news without checking whether it changes the risks. It can also influence how they interpret a pre-trade analysis that disagrees with their initial view.

A useful pre-trade check is to write down the original case, identify the strongest credible objection, and record what evidence would change the assessment. Separating the **investment case** from execution questions—such as spread, liquidity, and potential market impact—also helps avoid treating a compelling story as proof that an order will be inexpensive to execute.

## How PreTrAIde uses it

PreTrAIde does **not** claim to detect or score confirmation bias. Its pre-trade analysis is a separate check: a user builds a stock basket with a side, quantity, and execution conditions for each line, and receives GO, NO-GO, or MODIFY verdicts for the basket and each line, with reasoning in plain language.

The analysis draws on quantitative scores, recent news, and analyst ratings. A user can compare that output with their original reasoning and examine any disagreement. The verdicts are decision support, not recommendations; the success probability is a model estimate, not a guarantee or forecast of returns. Users place any trades themselves through their own broker.

## Common mistakes

- **Calling every disagreement bias.** Rejecting a weak argument after checking it is not the same as ignoring credible contrary evidence.
- **Counting headlines as independent proof.** Multiple stories may repeat the same underlying information.
- **Confusing an evidence tally with a probability.** The number of favorable items does not establish the chance of a favorable outcome.
- **Treating a tool’s verdict as a substitute for judgment.** A pre-trade check can surface considerations, but it does not remove the need to assess the evidence and its limitations.

---
Source: https://pretraide.com/glossary/confirmation-bias