The hidden cost of blurred evidence
Every investment decision rests on a collection of statements about the world. Some of those statements are facts: observable, verifiable, and falsifiable by evidence. Others are interpretations: judgments derived from facts, colored by the investor's mental models and prior beliefs. Still others are assumptions: claims about future states that cannot currently be verified but must be true for the investment thesis to hold.
Most investors mix these categories freely, storing them together in the same mental ledger labeled "what I know about this company." The mixing creates a specific failure mode: when a fact changes, interpretations built on that fact often fail to update, and assumptions that depended on those interpretations are never flagged for review. The investor continues to hold the original conviction without noticing that the foundation it was built on has shifted.
This is not a failure of intelligence or experience. It is a structural problem. The human mind is optimized for building coherent narratives rather than for tracking the epistemic status of every component in a belief system. Making the structure explicit, separating what is known from what is inferred, is the only reliable defense.
What the Evidence Ledger does
The Evidence Ledger is the core framework of the Evidence and Assumption Lab. It is a structured record that separates every statement supporting an investment thesis into four categories: verified facts, interpretations derived from facts, explicit assumptions about the future, and identified unknowns that are material but currently not determinable.
The ledger serves two functions. First, it makes assumptions explicit. An assumption that is written down can be monitored, tested against new evidence, and used to define a sell trigger. An assumption that lives only in the investor's head cannot. Second, it provides a structured review mechanism. When new evidence arrives, the investor reviews which ledger entries are affected and updates them accordingly. The ledger becomes a living record of the investment's epistemic state rather than a static document from the time of purchase.
Supporting the ledger are three utilities: the Fact/Inference Classifier, which challenges the investor to categorize every claim before adding it to the ledger; the Assumption Sensitivity Ranker, which identifies which assumptions have the most leverage over the investment outcome; and the Evidence Freshness Tracker, which schedules review cycles based on the expected rate of change of each evidence source.
Source quality and assumption sensitivity
Not all evidence is equally reliable. The Evidence and Assumption Lab organizes sources into four tiers based on reliability, accountability, and conflict-of-interest exposure. Primary sources (SEC filings, regulatory announcements, audited financial statements) form the strongest foundation. Secondary analytical sources (research reports, data providers) provide useful synthesis but introduce the analyst's own assumptions. Tertiary sources (financial media, market commentary) are useful for understanding market sentiment but should not serve as sole support for material claims. Quaternary sources (forums, social commentary, unattributed summaries) are the weakest and carry the highest contamination risk.
Assumption sensitivity is the other axis of evidence quality. Not all assumptions matter equally. An assumption about a company's gross margin trajectory in its core segment may determine 60% of the investment outcome. An assumption about the behavior of a minor product line may determine 5%. The Assumption Sensitivity Ranker helps investors identify which assumptions have the highest leverage over the conclusion, so review resources are concentrated where they have the most impact.
The combination of source quality and assumption sensitivity gives investors a prioritized evidence audit: high-sensitivity assumptions supported by weak sources receive the most scrutiny; low-sensitivity assumptions supported by primary sources require the least.
How the Evidence and Assumption Lab is organized
The lab is organized around five evidence categories, each treated as a distinct skill with its own curriculum structure. The five categories are: Fact vs. Interpretation, Source Strength, Source Freshness, Assumption Inventory, and Assumption Dependencies. Each category is covered by five content types: a conceptual explanation of what the category means for investment evidence quality, a how-to guide for evaluating it before acting, a checklist for collecting the right evidence, a failure-mode analysis describing how errors in this category manifest, and a worked case study moving from raw evidence to a structured investment conclusion.
The five categories are not independent. Source strength affects how much weight an assumption should carry. Source freshness determines how quickly an assumption should be flagged for review. Assumption dependencies reveal which assumptions are load-bearing for others, creating a hierarchy that directs review attention. An investor working through all five categories on a single investment thesis is building a complete, auditable evidence foundation rather than a collection of partially-examined claims.
Investors already working with the Investment Thesis Lab will find the Evidence and Assumption Lab covers the evidence layer that sits beneath every thesis element. The thesis defines what must be true; the evidence ledger documents the basis for believing it is true and tracks what would change that belief.
Every guide in this lab
- Fact vs. Interpretation covers each aspect of Evidence and Assumption analysis.
- Source Strength covers each aspect of Evidence and Assumption analysis.
- Source Freshness covers each aspect of Evidence and Assumption analysis.
- Assumption Inventory covers each aspect of Evidence and Assumption analysis.
- Assumption Dependencies covers each aspect of Evidence and Assumption analysis.
Frequently asked questions
Why do investors confuse facts with interpretations?
Facts and interpretations have different epistemic statuses, but they are stored together in the mind and expressed in similar language. A fact is an observable, verifiable event: earnings declined 12% last quarter. An interpretation is a judgment derived from facts: management is losing operational control. Both can be true, but only the fact can be verified directly. Investors confuse them because the interpretations that feel most certain are often built on facts that are genuinely solid, which gives the interpretation an unearned certainty. The consequence is that when a fact changes, the interpretation built on it often fails to update at the same rate or in the same direction, creating a persistent blind spot.
What is an assumption inventory and why does it matter?
An assumption inventory is a written list of every statement in an investment thesis that is not directly observable but must be true for the thesis to play out. For any investment conclusion, there are typically three to seven load-bearing assumptions: about the business, the competitive environment, the management team, or the macro backdrop. The inventory matters because most investment errors are not caused by incorrect facts but by assumptions that were never made explicit. An assumption that is never written down can never be monitored for change, never tested against contrary evidence, and never used to define a sell trigger. The assumption inventory is the prerequisite for falsifiable investment theses.
How does source quality affect investment conclusions?
Every piece of evidence has a source, and sources differ in reliability, timeliness, and conflict-of-interest. Primary sources, such as SEC filings, earnings call transcripts, and regulatory data, are the strongest: they come directly from the reporting entity and carry legal accountability. Secondary sources, such as analyst summaries or financial media, are weaker: they filter and interpret primary data and introduce the analyst's own assumptions. Tertiary sources, such as social media commentary or aggregated sentiment data, are weakest and should never stand alone as the basis for a material investment claim. Using a weak source where a primary source is available introduces unnecessary uncertainty into the investment conclusion.
When should evidence freshness trigger a review?
Evidence freshness matters because investment theses are based on conditions that change over time. A quarterly earnings assumption becomes stale after one earnings cycle passes. A macroeconomic assumption may degrade faster in a volatile environment. A competitive-position assumption may hold for years in a stable industry. The appropriate review trigger is not a fixed calendar interval but a change in the conditions underlying the evidence. If the key fact supporting an assumption was the company's gross margin trajectory, then any new earnings release with gross margin data triggers a review. The risk of over-scheduling reviews is low; the risk of letting stale evidence support a current conviction is high.