
Why Do Smart People Buy High and Sell Low?
Everyone knows the rule. Buy low, sell high.
Almost nobody does it.
The behavior gap shows up in study after study: retail investors as a class earn meaningfully less than the assets they own, because they buy near the top and sell near the bottom. Same instrument, same time horizon, worse return. The rule isn’t wrong. It’s fighting two of the strongest motivators in the human nervous system: greed when prices rise, fear when they fall. Buy high sell low psychology, the underlying engine behind that gap, is the most expensive blind spot in personal finance, and once you can name it, you can build around it.
What follows is the cycle, in Octalysis terms, and the mental model that lets you stop watching the dog and start watching the owner.
⚡ Speed Run Notes
- “Buy low, sell high” is the simplest rule in investing and almost nobody follows it. The reason isn’t intelligence. It’s that two Core Drives from the Octalysis Framework (Scarcity & Impatience in bubbles, Loss & Avoidance in crashes) hijack the rational play at the exact moments it matters.
- Greed shows up at the top: social proof when your taxi driver is talking about Bitcoin, FOMO when everyone else is making money, momentum bias telling you the trend will continue. Fear shows up at the bottom: panic, loss aversion, short-term thinking. Same person. Both errors.
- Buying low requires conviction in a future state. Without a 5- to 10-year belief, every dip looks like a permanent loss instead of a discount on something you wanted to own anyway.
- André Kostolany’s dog-and-master analogy is the cleanest mental model in the field: the owner (fundamental value) walks at 3 miles per hour; the dog (price) runs around but is on a leash. Most investors spend the entire day watching the dog.
- Warren Buffett’s 1986 Berkshire shareholder letter compressed the full system into one sentence: be fearful when others are greedy and greedy when others are fearful. The letter is still worth reading in 2026.
- The actual defense isn’t willpower. It’s a decision environment where the rational action and the emotionally rewarding action are the same action. That’s a behavioral-design problem, which is why I built the Geometric Utility Investing (GUI) framework on top of this analysis.
Table of Contents
- Why “Everyone Knows” Isn’t Enough
- The Greed-Fear Cycle in Octalysis Terms
- Kostolany’s Dog: A Better Mental Model
- The Variable Almost Nobody Talks About: Belief in the Future
- Three Cases Where the Cycle Played Out
- Practical Principles
- When This Becomes Speculation, Not Investing
- The Bridge: Decision Environment Beats Willpower
- Frequently Asked Questions
Author Credibility: Yu-kai Chou

Yu-kai Chou is an S-Tier Behavioral Designer and the creator of the Octalysis Framework, the gamification design system now applied to products and experiences reaching over 1.5 billion users. His book Actionable Gamification is one of the most-cited works in the field, and he has been ranked the #1 Gamification Guru in the World.
He has advised MrBeast, LEGO, Microsoft, Porsche, Tesla, Stanford, Harvard, and governments including Ukraine on turning behavioral psychology into product mechanics that actually change user behavior.
Verify: Wikipedia · Google Scholar · Wikidata · LinkedIn
The buy high sell low pattern is the most expensive case study I’ve seen for what behavioral design fixes that willpower cannot. I’ve been studying motivation systems for over a decade, and I built the Geometric Utility Investing (GUI) framework on top of this analysis precisely because the underlying greed-fear cycle is too well-engineered by evolution to be defeated with discipline alone. The post below is the diagnosis. GUI is the prescription. They’re meant to be read together, in that order.
Why “Everyone Knows” Isn’t Enough
If “buy low, sell high” were a knowledge problem, this post wouldn’t need to exist.
The phrase shows up on every personal finance website, every podcast, every brokerage onboarding flow. Even people who don’t invest can recite it. And yet, year after year, study after study finds the same behavior gap: retail investors as a class capture meaningfully less return than the assets they own, because they enter near tops and exit near bottoms.
The rule is right. The execution is broken.
Knowing the rule and acting on the rule are two different operations of the brain. The first is cheap. The second is expensive, because the second has to override emotion in real time, on a falling chart, with money you care about. Most investors lose at the second step, regardless of how confidently they nailed the first one in calm conditions.
This is a behavioral design problem, not an information problem. And once you frame it that way, you stop trying to fix it with more information.
The Greed-Fear Cycle in Octalysis Terms
Markets cycle through two phases that look opposite but share one feature: the rational action and the emotionally rewarding action point in opposite directions.
Greed at the Top: Three Stacked Core Drives
Note on terminology: In the Octalysis Framework, human motivation is broken down into 8 Core Drives (CDs). We use shorthand like CD2 for Core Drive 2 to refer to these specific motivational levers.
When prices have been rising for months, three Core Drives stack on top of each other and produce the urge to buy at the worst possible moment.
Core Drive 5 (CD5): Social Influence & Relatedness. The taxi driver is talking about Bitcoin. Your dentist asks you which crypto you’re holding. Your group chat has a screenshot of someone’s 3x. Social proof is one of the most potent CD5 triggers there is, and at market tops, social proof is screaming “buy.”
Core Drive 6 (CD6): Scarcity & Impatience. Everyone is making money except you. The window is closing. Other people are getting in and you’re standing on the sidelines. CD6 is the FOMO engine, and FOMO is what gets people to buy at the absolute worst price.
Core Drive 2 (CD2): Development & Accomplishment. Now layer in the gain itself. Watching a stock you bought go up by 30 percent isn’t just financial. It’s a measurable achievement, a level-up, a status reward. The brain interprets the green number the way it interprets a quest completion. So instead of taking profits at the top, the rational impulse to lock in is overridden by “just one more level.”
Three motivational engines firing at once, all telling you to buy. That’s why people buy at the top. It feels right.
Fear at the Bottom: One Dominant Core Drive
The crash flips the polarity. Now it’s a single Core Drive, but the strongest one in the system.
Core Drive 8 (CD8): Loss & Avoidance. Prices are falling. Your portfolio is red. Every minute you hold, the loss might get worse. Loss aversion has been measured in prospect theory research since Kahneman and Tversky’s 1979 paper, and the consensus is brutal: a loss feels roughly twice as painful as an equivalent gain feels good. So when prices crash, the brain doesn’t evaluate the asset on a 5-year horizon. It evaluates the pain of holding through one more red day.
And the most pain-relieving action available, in the moment, is to sell. Lock in the loss. End the bleeding.
That’s why people sell at the bottom. Not because they don’t know the rule. Because CD8 is dominant, and CD8 is telling them to make the pain stop.
The same person, eight months apart, executes both errors. Buys high under CD5/CD6/CD2 stacked, sells low under CD8 dominant. The result is the behavior gap.
Kostolany’s Dog: A Better Mental Model
André Kostolany was a Hungarian-born investor and writer who spent decades watching markets and writing books explaining them in plain language. His most famous mental model is what’s now called Kostolany’s Dog, and it’s the cleanest rebuttal to the greed-fear cycle I’ve ever seen.
The setup goes like this. A man is walking his dog down a country road. The man walks at a steady 3 miles per hour. The dog runs ahead, runs back, sniffs the side of the road, sprints in circles, occasionally falls behind. To anyone watching, the dog’s movements look chaotic and unpredictable. But the dog is on a leash. And no matter how much the dog runs around, the dog ends up where the man ends up, because the man is always walking forward at 3 miles per hour.
The man is the underlying economic value. The dog is the daily price.
Most retail investors spend their entire day watching the dog. They watch the price go up, watch it go down, feel each lurch, and make decisions based on the dog’s last 30 seconds of movement.
The professional investor watches the man.
This is not a metaphor that helps in the moment if your nervous system is in CD8 panic. It’s a model you have to internalize before the panic, so that when the panic arrives, you have a different question available than “is the dog still running away?” The new question is: where is the man heading?
The Variable Almost Nobody Talks About: Belief in the Future
Buying low and selling high requires a hidden variable that most coverage of the topic skips entirely.
You have to believe the asset is heading somewhere over the next 5 to 10 years.
Without that belief, every dip looks like a death spiral. Every drop is the start of a permanent loss. Every red day is the proof that you should have sold yesterday. CD8 wins by default, because there’s no counterweight on the other side of the scale.
With that belief, the same dip looks completely different. It looks like a discount on something you already wanted to own. The asset hasn’t become bad. What’s happened is that the dog has run forward of the man for a stretch, and now it’s catching back up. A drop becomes information about the price, not about the value.
This is why the practical advice “buy low” produces almost no results without a prior step: do the work on the future. What does the next decade look like for this industry? Is the demand curve growing? Does the underlying technology look durable? Are you looking at a cyclical pattern (real estate) or a structural shift (electric vehicles)?
Without that work, “buy low” is just another saying you can’t execute on, like “don’t panic.”
With that work, every CD8 panic moment is offset by a future-belief signal: this is the dog moving, not the man stopping.
Three Cases Where the Cycle Played Out
Three case studies make the framework concrete — one from 2001, one from the 2010s, and one from the last few years.
The dot-com crash and the survivors. In 2001, the NASDAQ lost more than 75 percent of its peak value. Vast numbers of internet stocks went to zero or close to it. The dominant emotional read at the time was that the internet itself had been a mistake. CD8 was screaming.
But the underlying technology was not a mistake. The internet kept growing. Demand kept compounding. Companies that had real businesses (Amazon, eBay, the early-stage Google) kept their fundamentals intact and emerged from the wreckage stronger. Investors who held conviction in the future of the internet, and who bought during the panic, were rewarded over the next decade by some of the largest stock returns in modern history.
The bubble had popped. The industry had not. Distinguishing those two things in real time was the entire game.
Real estate, 2012-2018. The U.S. housing market had been crushed by the 2008-2009 financial crisis. By 2012, fear was still dominant. Many would-be buyers stayed sidelined because the broader narrative was that housing was a structurally broken asset class.
The historical pattern told a different story. Over multi-decade timeframes, real estate in growing regions has tended to appreciate. The fundamental drivers (population growth, urbanization, finite land, mortgage tax treatment) were intact. By 2018 the market had recovered and then some.
Buyers in 2012-2013 were buying at a fear bottom in an asset class with a strong long-term trend. That’s the textbook setup. Almost no one was excited to do it at the time.
The COVID crash, March 2020. Over five weeks in February-March 2020, the S&P 500 dropped roughly 34 percent. The dominant emotional read was that a pandemic-era shutdown would permanently impair earnings for years. CD8 was screaming, and many retail investors sold. But the underlying productive capacity of the U.S. economy had not been deleted by a virus; it had been paused. By August 2020 the S&P had recovered its losses, and over the next five years it roughly doubled from the March-2020 lows. Buyers who held a 5- to 10-year future-belief and bought during the panic — even at imprecise prices — were rewarded by one of the cleanest “the man kept walking” recoveries in modern market history.
All three cases share the same shape: a real long-term trend, a panic-driven short-term price, and a small number of participants who could keep their eyes on the man instead of the dog.
Practical Principles
Five rules that survive contact with real-world greed and fear.
1. Use only money you can afford to lose. This is the single biggest CD8 mitigation in the entire system. If a 50 percent drop would damage your ability to pay rent, your nervous system will not let you hold. The rule is not optional. Position size is the precondition for behavioral discipline.
2. Stop trying to time the perfect top or bottom. You cannot do it. No one can. The most experienced investors I know are content with “quite high” or “quite low,” and they treat anyone claiming more precision than that as either lucky or selling something. Aim for a range, not a point.
3. Do the future work before the price moves. What you believe about an asset’s next 10 years has to be set when the chart is calm. If you wait until the chart is bleeding, your beliefs will be set by the chart, not by the underlying.
4. Distinguish a bubble from a trend. The dot-com bubble was a valuation bubble. The internet was a trend. Both collapsed together for 18 months and then diverged for two decades. Most asset crashes are valuations correcting, not trends ending. Treat them differently.
5. Pre-decide your action at the next dip. Behavioral economics research on implementation intentions consistently shows that decisions made under emotional load are worse than decisions made in advance with a triggering condition. “If asset X drops by 20 percent and my future-belief is intact, I will buy in tranches at the following levels” is a sentence you can write while calm and execute while panicked. That’s the move.
When This Becomes Speculation, Not Investing
I want to draw a clean line here, because it’s a line I’ve seen blurred in dangerous ways.
The framework above is about speculation, not investing in the strict sense. Speculation is making an educated bet about the future direction of an asset class or industry, based on pattern recognition and conviction in a long-term trend. Investing in the strict sense (Buffett-style fundamental analysis with margin of safety on individual securities) is a related but distinct discipline.
The greed-fear cycle is real for both. The framework helps for both. But the risk profile is different, and pretending otherwise is how people end up underwater on assets they didn’t actually understand.
If you’re going to play in markets, know which game you’re playing. Speculation is fine if you’re sized for it and clear-eyed about it. The mistake is to call it investing and be surprised when it doesn’t behave like investing.
The Bridge: Decision Environment Beats Willpower
Everything in this post is diagnosis. Naming the cycle helps. Knowing Kostolany’s Dog helps. Pre-deciding helps. But by itself, none of this defeats the cycle, because the cycle is wired deeper than self-talk reaches.
The actual defense is to build a decision environment where the rational action and the emotionally rewarding action are the same action.
That’s the entire premise of the Geometric Utility Investing (GUI) framework I built for myself and now teach. GUI doesn’t ask you to override greed or fear with willpower. It engineers the structure of your buying and selling so that every dip triggers a small, satisfying buy (CD2 reward), every overheated rally triggers a small, satisfying sell (also CD2 reward), and the conviction-based long-term trend stays the spine of the whole system.
The post you just read is the why. GUI is the how. They’re meant to be read together, in that order, because diagnosis without prescription is just an interesting conversation, and prescription without diagnosis is the cargo-cult version of behavioral design.
If buy high sell low psychology has cost you money, the path forward isn’t to read more about discipline. It’s to redesign the decision surface so the discipline is automatic.
Frequently Asked Questions
If everyone knows the rule, why does the behavior gap persist?
Because the rule is a knowledge claim and the behavior is an emotional execution. Knowing “buy low, sell high” lives in the prefrontal cortex. Acting on it during a 30 percent drawdown is fought out in the limbic system, where Loss & Avoidance has a much louder voice. Information alone never closes the gap. Decision architecture closes the gap.
Where did Warren Buffett actually say “be fearful when others are greedy”?
The line in its most-quoted form appears in Buffett’s 1986 letter to Berkshire Hathaway shareholders, where he wrote that the firm simply attempts “to be fearful when others are greedy and to be greedy only when others are fearful.” The full letter is publicly available on the Berkshire Hathaway site and is worth reading in its original context.
Who was Kostolany and is the dog analogy actually his?
André Kostolany (1906-1999) was a Hungarian-born financial commentator and investor who spent most of his career in Paris and is widely credited with the dog-and-master analogy as a way of distinguishing short-term price from long-term value. He’s much better known in European investing circles than in the U.S., which is part of why the analogy still feels fresh to American readers when they encounter it.
What if I have no idea what an asset will be worth in 10 years?
Then you have a research problem before you have an investing problem. Buying low without conviction in the future is just buying a falling knife and hoping. The work of forming a 10-year view (which industries are durable, which are cyclical, what demand curves look like) is the part most retail investors skip. It’s also the part that turns a panic dip into a buying opportunity instead of a stop-loss event.
Is this advice or is it speculation?
It’s a behavioral framework for thinking about speculation. It is not financial advice, it is not a recommendation for any specific asset, and the position-sizing rule (only money you can afford to lose) is the most important sentence on the page. Markets are not obligated to reward conviction, and history doesn’t guarantee the future. Treat the framework as a way to avoid the worst self-inflicted errors, not as a way to guarantee gains.
Stop Watching the Dog
The behavior gap is one of the most expensive consistent findings in personal finance, and it’s caused almost entirely by the greed-fear cycle running unopposed.
The diagnosis is straightforward. Three Core Drives stack at the top to push you into bad buys. One Core Drive dominates at the bottom to push you into bad sells. Same nervous system, same person, predictable result.
The defense isn’t willpower. It’s a different mental model (Kostolany’s dog), a different question (where is the man heading?), and a decision environment that converts the rational action into the rewarding action so you stop having to choose between them.
Read the post. Then read the GUI framework. The first one tells you why the cycle exists. The second one tells you how to design around it.
And the next time the chart drops 30 percent on a Tuesday, ask the question that actually matters: is this the dog, or is the man stopping?
Related Reading
- Geometric Utility Investing (GUI): My Framework for Conviction-Based Dip Buying
- The Octalysis Framework
- Actionable Gamification
- Economy Design Framework: Tokenomics That Actually Work
