Documenting Sell Criteria Before You Buy

Most investors spend hours deciding what to buy and minutes deciding when to sell. That imbalance is where a lot of damage happens: holding a broken idea because it’s down, or dumping a good one because it’s noisy. The fix is simple and unglamorous — write your sell criteria before you buy, while you’re still thinking clearly.

Educational only — not financial advice. Every example below is hypothetical. This post goes deeper on one countermeasure from behavioural biases that hurt stock pickers most: pre-committed exit rules.

Direct answer

Before you place a buy order, write down (1) the thesis in two sentences, (2) the catalysts you expect and roughly when, and (3) kill criteria — specific, checkable facts that would prove the thesis wrong. Add a valuation ceiling, a position-size limit, and a time stop. Then review only at defined triggers (earnings, major filings, large price moves) and ask one question: has a kill criterion been met? If yes, exit or cut back unless you can write down why the criterion was wrong — before looking at your gain or loss.

Why pre-commitment works

Bias	What it does to sell decisions	How written criteria counter it
Disposition effect	Sell winners too early, hold losers too long	The criterion, not the P&L, decides
Anchoring	Wait to “get back to even”	Your purchase price doesn’t appear in any criterion
Confirmation bias	Explain away bad news	The disproof was defined before you were invested
Sunk cost	“I’ve done so much research”	Research time isn’t a reason to own
Recency	Panic on one bad quarter	Kill criteria usually require a trend, not one data point

The five types of sell criteria

Type	What it tests	Example wording (hypothetical)
Thesis break (kill criterion)	Did the reason to own disappear?	“Exit if same-store sales growth is negative for two consecutive quarters.”
Valuation ceiling	Is the upside priced in?	“Trim if price exceeds my high-case value of $68.”
Time stop	Did the catalyst arrive?	“If the new mine isn’t producing by Q4 next year, reassess from zero.”
Size limit	Has it grown too big?	“Trim back to 6% if the position exceeds 10% of the portfolio.” See position sizing.
Opportunity cost	Is there a clearly better use of the capital?	“Replace only if the new idea scores higher on my brief template and fits the sector cap.”

Price drops are deliberately missing from the list. A falling price is a review trigger, not a sell criterion — it tells you to check the kill criteria, not to sell.

Writing kill criteria that actually work

Good kill criteria share four properties:

Property	Weak version	Strong version
Specific	“If growth slows”	“If organic revenue growth falls below 3% for two quarters”
Observable	“If management loses focus”	“If the CEO or CFO leaves without a named successor”
Linked to the thesis	“If the stock falls 20%”	“If the dividend payout ratio on free cash flow exceeds 100% for a full year”
Sourced	“If things look bad”	“As reported in the MD&A or quarterly results” — see how to read an MD&A

Aim for two or three kill criteria per stock. More than five usually means the thesis itself is fuzzy.

Worked example: a hypothetical Canadian pipeline holding

Field	Entry (written before buying)
Thesis	Contracted, fee-based cash flows support a growing dividend; a large expansion project adds cash flow in two years.
Catalysts	Project in service by end of next year; annual dividend increase.
Kill criteria	(1) Distributable cash flow payout ratio above 90% for four straight quarters. (2) Net debt to EBITDA above the company’s own stated ceiling with no plan to fix it. (3) Project cancelled or written down.
Valuation ceiling	Trim if price exceeds my high-case value.
Size limit	Max 8% of portfolio.
Time stop	If the project slips more than 12 months, re-underwrite from scratch.
Review triggers	Quarterly results, any material change report, a 20% price move.

Planned vs improvised exits

Keeping a log of why you sold is as useful as the criteria themselves. After a year or two you’ll see which criteria fire, which you ignore, and whether improvised exits helped or hurt.

A review routine that respects the plan

Trigger	What to do
Quarterly results	Check each kill criterion against the new numbers; update the brief
Material change report or big news	Ask whether a criterion is affected; most news isn’t decisive
Price down 20%+ (hypothetical threshold)	Re-read the thesis and kill criteria before looking at the chart again
Criterion met	Exit or reduce unless you can write a specific reason the criterion was wrong
Changing a criterion	Allowed only with a written reason, dated, and not on the day of a big price move

In a taxable account, also note the tax side before selling — capital gains, adjusted cost base and, if you plan to buy back, the superficial loss rule. Tax should inform the how of an exit, rarely the whether.

Common mistakes

Mistake	Why it hurts	Fix
Using only a price stop	Sells good businesses on noise, holds bad ones that drift	Pair price triggers with thesis-based criteria
Vague criteria	Easy to rationalize	Make each one specific and observable
Moving the goalposts	Turns criteria into decoration	Change only with a dated, written reason
No time stop	Dead money ties up capital for years	Put a date on the catalyst
Criteria in your head	Memory is edited by outcomes	Keep them in the stock’s brief or journal

Go deeper

A one-page “Thesis → Catalyst → Kill Criteria” card is planned for the DIY Process ebook. You can build your own today with the free bull / bear / watch brief template.

Related reading

• Behavioural biases that hurt stock pickers most
• How to build a simple stock brief template (bull / bear / watch)
• Position sizing for DIY stock pickers
• SEDAR+ alerts and filing types: monitoring stocks you already own

Sources

• Shefrin, H. and Statman, M. (1985), “The Disposition to Sell Winners Too Early and Ride Losers Too Long,” Journal of Finance 40(3).
• Odean, T. (1998), “Are Investors Reluctant to Realize Their Losses?,” Journal of Finance 53(5).

Media credits

• Hero: “Emergency exit sign 2016.jpg” by Santeri Viinamäki, CC BY-SA 4.0, via Wikimedia Commons. Cropped and resized to 1280×720. Real photograph, not AI-generated.

By Andrew · stocksummarizer.com