Two investors hold a stock down 40%. One is demonstrating the patience that produces great long-term returns. The other is refusing to accept they were wrong. From inside their own heads, these feel exactly the same — both are certain, both can list reasons, both feel the market is mistaken.
A thesis is a claim, not a hope
Most theses are written vaguely enough that no outcome can falsify them. "This is a great business at a reasonable price" cannot be wrong, because nothing observable contradicts it. A usable thesis makes specific claims that reality can refute.
- “Strong management with a long runway”
- “Trading below intrinsic value”
- “Temporary problems in a great business”
- “The market does not understand this yet”
- “Margins recover to 18% within six quarters”
- “The new plant reaches 70% utilisation by FY27”
- “Receivable days fall below 60 within a year”
- “Market share stops declining within two quarters”
Distinguishing price from evidence
A falling price is not disconfirming evidence for a fundamental thesis — that is exactly what a good business at a temporarily poor price looks like. Deteriorating fundamentals are disconfirming evidence. Confusing the two produces both errors: selling good businesses on price weakness and holding failing ones because "nothing has changed" while everything has.
| What happened | Thesis status | Reasonable action |
|---|---|---|
| Price fell 30%, fundamentals unchanged | Intact | Hold, or add if sizing allows |
| Price fell 30%, margins fell with it | Damaged | Re-underwrite from scratch, not from your entry price |
| Price rose 40%, fundamentals unchanged | Intact but cheaper no longer | Reassess against the new price |
| Price unchanged, the key claim failed | Broken | Exit — the market has not noticed yet |
Why we defend positions
Once you buy, the position becomes part of your identity. You have told people. You have defended it. Every hour of research raises the cost of concluding it was wasted, and admitting error feels like a statement about your competence rather than about one company.
- 1Write the thesis before buying, with exit conditions
Two or three specific, checkable claims and what would falsify each. Written afterwards, it is a rationalisation.
- 2Diarise a review
Set a date to re-read the thesis against what has actually happened. Scheduled review beats waiting for the moment you feel like it, which never comes for losing positions.
- 3Re-underwrite from today
Ask whether you would buy this at today's price knowing what you now know. If not, the only thing keeping you in is your entry price — which the company is entirely unaware of.
- 4Argue it aloud to someone unimpressed
Explaining a thesis to a person with no stake in it exposes the parts that only sound convincing inside your own head.
Eighteen months in
You bought expecting margins to recover from 11% to 16% within two years. Eighteen months on, margins are 9%, management still cites "temporary headwinds", and the stock is down 35%. You have researched this company more than any other.
What best distinguishes conviction from stubbornness?
Dono andar se ek jaise lagte hain. Farak sirf ek hai: aapne pehle likha tha ki kya hone pe maan loge ki aap galat ho? Agar haan, aur woh nahi hua — toh yeh sabr hai. Agar aapne kabhi likha hi nahi, toh yeh zid hai, chahe kitna bhi research kiya ho.
- Conviction and stubbornness are indistinguishable from the inside; only stated exit conditions separate them.
- A thesis must make claims that reality can refute.
- A falling price is not disconfirming evidence; deteriorating fundamentals are.
- A thesis can break while the price does nothing — the hardest sell to act on.
- Re-underwrite from today's price, which is the only price the company knows nothing about.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- difference between conviction and stubbornness in investing
- Conviction has exit conditions stated in advance; stubbornness does not. Both feel identical from the inside — certainty, a list of reasons, a sense that the market has it wrong — so the strength of the belief cannot tell them apart. The only observable test is whether you wrote down beforehand what would prove you wrong, and whether that thing has now happened.
- an investment thesis that no outcome can prove wrong is described as
- Unfalsifiable. “A great business at a reasonable price” cannot be contradicted by anything observable, which makes it a hope rather than a claim. A usable thesis states things reality is able to refute — margins recovering to a stated level within a stated number of quarters, receivable days falling below a set figure, market share ceasing to decline by a given date.
- is a falling share price proof that my thesis is wrong
- No — a falling price is exactly what a sound business at a temporarily poor price looks like, so on its own it disconfirms nothing. Disconfirming evidence is deteriorating fundamentals: margins moving away from your claim, market share still slipping, receivable days climbing. Confusing the two produces both errors, selling good businesses on price weakness and holding failing ones because “nothing has changed” while everything has.
- what does re-underwriting a position mean
- Re-underwriting means asking whether you would buy this stock today, at today’s price, knowing everything you now know — with your own entry price deliberately excluded from the question. If the answer is no, the entry price is the only thing still holding you in, and the company has no knowledge of it. It is the standard check against commitment bias, because it removes the sunk research and the purchase from the decision.
- what does it mean if my reason for holding a stock keeps changing
- It usually means the conclusion came first and the reasons are being recruited afterwards. You bought for margin recovery; margins fell, so now it is the new product; the product slipped, so now it is the land on the balance sheet. Each reason can be individually reasonable while the pattern shows a position being defended rather than analysed.