Investing Basics · Decision Process
Investment Decision Record
Write down the reasoning before the outcome is known.
An investment decision record captures what was known, assumed, and expected at the time a decision was made, before the outcome is known. Recording decisions in advance prevents the retrospective distortion that reshapes memory to fit outcomes, and enables honest review of the process rather than the result.
Direct Answer
An investment decision record captures what was known, assumed, and expected before the outcome is known. It should include the decision date and content, alternatives considered, key evidence, assumptions, expected outcomes, major risks, falsifiers, review triggers, implementation constraints, and what was not known at the time. The purpose is to enable honest retrospective review of the process, not the outcome.
Why decisions should be recorded in advance
Investors naturally remember their decision-making process through the lens of the eventual outcome. A successful investment is remembered as obviously correct in hindsight; a failed one is often explained as uniquely unforeseeable. Pre-committing the reasoning in writing before the outcome is known prevents this retrospective distortion and enables genuine improvement.
The record is not about predicting correctly. It is about thinking clearly and being able to review that thinking honestly later. An investor who wrote down "I expect this company to grow revenue at 15% for three years because..." and then reviews that record against actual outcomes has a basis for learning. An investor who reconstructs the reasoning after the fact is reviewing a story, not a decision.
The minimum record
A complete decision record includes these components:
- Decision date: when the decision was made, not when it was implemented
- The decision: specifically what was bought, sold, or held, and in what size relative to the overall portfolio
- Alternatives considered: what other options were evaluated and why they were rejected
- Key evidence: the verified facts, reports, or primary data that supported the decision
- Assumptions: what must be true for the thesis to work, that is not already established by evidence
- Expected outcomes: a range, not a single point, covering upside, base case, and downside scenarios
- Major failure modes: which specific risks could cause the thesis to be wrong
- Falsifiers: what observable events or data points would indicate the thesis is failing
- Review triggers: specific events or time intervals that prompt re-evaluation
- Implementation constraints: position size limits, tax considerations, timing constraints
- Unavailable information: what was not known at the time and would have changed the analysis if it had been
What makes evidence different from assumption
Evidence is a verified fact at a specific date: the company's reported revenue from a quarterly filing, the current credit rating from a rating agency release, the regulatory requirement from a published rule. An assumption is a belief about how something will change or continue: the expectation that revenue will grow at a historical rate, the belief that a market will expand.
Confusing the two is the most common error in pre-investment analysis. A narrative that blends the two reads coherently but creates false confidence. Separating them explicitly forces the writer to acknowledge how much of the investment case rests on beliefs about the future rather than facts about the present. A record that shows ten items of evidence and two assumptions is structurally different from one showing two items of evidence and ten assumptions, even if both produce the same conclusion.
How to specify falsifiers
A falsifier is an observable event or data point that would indicate the investment thesis is not playing out as expected. For an earnings-growth thesis, a falsifier might be three consecutive quarters of declining operating margins despite revenue growth. For a valuation-reversion thesis, a falsifier might be a capital structure change that makes the original valuation basis inapplicable.
A thesis without specific falsifiers defined in advance leaves the investor without a clear basis for exit that is not driven by price movement alone. Price alone is a poor exit signal because a position can decline for reasons unrelated to the thesis (general market sell-off, sector rotation) or appreciate for reasons that also contradict the thesis (short squeeze, rumors). Falsifiers that are thesis-specific rather than price-based are more reliable guides to when the original reasoning no longer applies.
Setting review triggers
Time-based triggers (review in six months) and event-based triggers (review if the company announces a major acquisition or a key executive departs) serve different purposes. Time triggers prevent neglect of positions that have been stable for a long period. Event triggers respond to new information that may be material to the thesis.
Both should be set at the time of the decision because they reflect the investor's assessment of what information matters, before emotions about performance enter the picture. An event trigger set after a 30% drawdown is not the same as the same trigger set at entry, because the post-drawdown trigger is shaped by the desire to find a reason to hold through the loss rather than by an objective assessment of what would change the investment case.
Recording what was not known
Every investment is made with incomplete information. Recording explicitly what was unknown at the time of the decision serves two purposes. First, it prevents holding the prior decision responsible for information that was genuinely unavailable at the time. A decision made without access to information that was not yet public cannot be fairly evaluated as if that information should have been considered. Second, it highlights information gaps that might be worth filling before a similar decision is made in the future.
The unavailable information section is particularly useful in retrospect. If a bad outcome followed from information that was available but not reviewed, that is a different lesson than an outcome that followed from information that could not have been known. The distinction between bad process and bad luck matters for how to adjust future behavior.
The review rule
Never rewrite the original record after the fact. The record's value lies entirely in its pre-outcome nature. If the thesis evolves because new information emerges or the analysis changes, add a dated amendment alongside the original. The original and the amendment are then both readable, and the evolution of the thinking is preserved rather than obscured.
At each review date, compare the original process with what is now known. Note what the record got right, what it missed, and whether the distinction between evidence and assumptions was maintained. A record that called most of the evidence correctly but made an assumption that turned out to be wrong teaches a different lesson than one that mixed up evidence and assumption from the beginning.
Using decision records to improve over time
The review process is where learning occurs. A record reviewed after a favorable outcome should ask: was the outcome due to the thesis being correct, or was it partially due to luck, market conditions, or factors not anticipated in the record? A favorable result is not proof of a good process if the result came from a factor that was not identified in the original analysis.
A record reviewed after an unfavorable outcome should ask: was the decision wrong given the information available at the time, or was the thesis reasonable but the outcome fell in the loss scenario that was already identified as possible? A loss that was within the expected downside range described in the record is a different kind of outcome than a loss that fell outside what was anticipated. Distinguishing good process from good outcome is the foundation of improving investment judgment over time, and that distinction requires the pre-outcome record.
FAQ
How long should an investment decision record be?
Long enough to capture the eleven minimum components described on this page, and no longer. A record that takes an hour to write will not be written consistently. The goal is to capture the decision date, what was decided, alternatives considered, key evidence, assumptions, expected outcomes, major failure modes, falsifiers, review triggers, implementation constraints, and what was not known at the time. A record that covers those points in a few hundred words is more useful than an exhaustive document that discourages the habit.
What is the difference between a decision record and a trading journal?
A trading journal typically records execution details: what was bought or sold, at what price, at what time, and how the trade was placed. A decision record captures the pre-outcome reasoning: why the decision was made, what evidence supported it, what assumptions it required, and what would indicate it was wrong. The two serve different purposes. A trading journal improves execution; a decision record improves analysis. Both can coexist, and a good record includes execution notes as one component.
Should I keep a decision record for every position or only large ones?
A decision record is most valuable for positions that are large enough to matter to portfolio outcomes, positions in unfamiliar instruments, and any position built on a thesis that could be wrong in a specific and identifiable way. Routine contributions to a diversified index fund in a retirement account do not require the same level of documentation as a concentrated position in a single stock or a new asset class. A practical rule is to write a record for any position where the loss of the full amount would be meaningful, or where the reasoning is not trivially obvious.
What is an investment thesis falsifier and why is it useful?
A falsifier is a specific observable event or data point that would indicate the thesis is not playing out as expected. Specifying falsifiers before making an investment serves two purposes: it forces clarity about what the thesis actually requires, and it creates concrete exit conditions that are not determined by price movement alone. Without pre-specified falsifiers, exit decisions tend to be made reactively under emotional pressure, often at worse prices than a pre-committed process would have produced.
How often should I review past decision records?
At every review trigger specified in the original record, and at the time of any exit. The review should compare the original evidence and assumptions with what is now known, note what the record got right and what it missed, and assess whether the distinction between evidence and assumption held up. Reviews after favorable outcomes are as important as reviews after unfavorable ones, because a good outcome does not necessarily confirm a good process. Quarterly reviews of all open positions are a reasonable minimum for actively managed portfolios.
Can I retroactively create a decision record for an existing position?
A retroactive record has limited value because the outcome is already known and the author's memory of the reasoning will be shaped by that outcome. It is possible to write a current-state record that documents the reasoning for holding a position today, which is a genuine prospective decision with its own date, evidence, assumptions, and falsifiers. That is a legitimate and useful exercise. It is not the same as the original entry record, and should be labeled accordingly.
Educational use
This page is educational and informational. It does not tell a reader what to buy, sell, hold, or contribute, and it does not account for an individual's objectives, taxes, legal situation, benefits, debts, time horizon, or risk tolerance. Verify rules, limits, product terms, fees, and market data from current primary sources before acting.
References
The concept of pre-commitment in investment decision-making is a well-established practice in behavioral finance research. This page describes a framework for implementing it; the specific format can be adapted to individual needs.
Reviewed by the Swoopr Editorial Team in September 2026.