Post-mortemLessonInvestingDecision-makingPsychology and behaviour
A bad year separated into its causes: which mistakes were mine and which were the market
François Rochon · Giverny Capital's 2025 annual letter, on a year of underperformance · 2026
The idea
Giverny Capital's 2025 letter is a rare full account of underperformance from a long-term manager. The causes are separated: two stock-specific mistakes, and a market that sold software companies on fear of artificial intelligence while their earnings kept growing. On one holding the manager names two errors of his own, not selling when the chief executive left, and buying it at all instead of adding to a stronger business he already owned. The letter then compares the year with 2006 and 2007, when avoiding popular energy stocks also cost relative performance for a time. The structure is the lesson: own the mistakes precisely, and distinguish them from a market you have chosen not to follow.
Separate a bad year into your errors and the market's mood, name the errors specifically, and say whether the method changes.
When it works
Reviewing your own record, reading any manager's bad year, and the interview question about a mistake.
When it may fail
The letter's rights do not allow an excerpt here. And a manager's division of a bad year between self and market is a judgement a reader should test, because the second column is the comfortable one.
How to use it
Write your own version of the letter after any bad period: a column for decisions you would reverse with the same information, and a column for prices that moved against a view you still hold. Act only on the first column.
Where it came from
Letter Giverny Capital's 2025 annual letter, on a year of underperformance · François Rochon
What the source says. Rochon separates two stock-specific mistakes from a market that sold software companies on fear of AI while earnings grew, names two errors of his own on one holding, and compares 2025 with 2006 and 2007 when avoiding popular energy stocks also cost relative performance.
Connected ideas
Says something similar
State, for each holding, what would have to happen for the thesis to be rightWrite the thesis with its test attached: the variant view, the evidence, and the event that would confirm or refute it.
Puts it to work
A good decision and a good outcome are separate thingsJudge a decision by what was knowable when it was made, and keep the outcome in a separate column.
Concepts
Put it to work on L3VLUP
More like this
- FrameworkPsychology and behaviourBook
Anchoring and base rate neglect are the two biases that cost money
Before estimating, ask what number is anchoring you and what the base rate is for cases like this.
From Thinking, Fast and Slow · Daniel Kahneman
- Case studyCrises and market historyDocumentary
Judge a decision on what was knowable at the time, not on how it turned out
The fair test of a decision is the information available when it was made; hindsight is a different exam.
From Inside the Meltdown · 2009
- Decision principleInvestingBook
One person's record is a story about survivorship until proven otherwise
Ask how many people ran the same method and are not in the sample before treating a record as evidence of skill.
From How I Made $2,000,000 in the Stock Market · Nicolas Darvas
- Post-mortemLeadershipLetter
The mistake was how quickly we moved, not where: a chief executive locates the error
In a post-mortem, separate the direction from the pace and the explanation, and say which one failed.
From Netflix's third-quarter 2011 letter to shareholders · Reed Hastings and David Wells