Have you ever done the work carefully, gotten the framework right, logic checked out — and still lost money?
It's not because you weren't thorough. It's because investors are not machines. Even a sound framework can produce a bad decision when the person using it is biased.
The previous two articles built two analytical frameworks. The "three-force model" explains K-shaped divergence through AI supply rigidity, policy regime transition, and global division-of-labor restructuring. The "three-layer signal cross-validation" tracks the AI industry's stage through supply, capital, and earnings signals.
Both frameworks are internally consistent, their data is verifiable, their boundaries are clearly drawn. This article is about a different layer of the problem: the cognitive biases humans carry when processing information — biases that distort a framework's inputs, interfere with its execution, and ultimately turn a correct framework into an incorrect judgment.
I. Two Types of Boundaries: Structural vs. Psychological
Before discussing cognitive bias, it's worth distinguishing two different kinds of framework limitation.
Structural boundaries are the framework's own scope of applicability. The previous articles covered these — technology breakthroughs can change supply rigidity, policy shifts can alter the macro environment, geopolitics can disrupt supply chains. When objective conditions change, you adjust the framework's assumptions. That is a technical limitation.
Psychological boundaries are the mental limits of the person using the framework. Even when objective conditions haven't changed, investors can still misread signals, misjudge stages, and miss turning points — because of how the human mind processes information.
Psychological boundaries are more insidious than structural ones. You never feel yourself falling into confirmation bias — you feel yourself doing rational analysis. You never feel yourself herding — you feel yourself going with the trend.
Understanding this layer of human limitation is a prerequisite for using any analytical framework. Without that awareness, even a good analytical system can become a sophisticated way of confirming what we already want to believe.
II. How Frameworks Get Distorted: Three Cognitive Biases
Bias One: Availability Bias — Distorting the Input
Availability Bias
People more easily remember recent or vivid information, while overlooking long-term data and baseline facts.
Analytical frameworks depend on high-quality input. Availability bias systematically distorts that input quality.
When an industry trend becomes the market's main narrative, related news is everywhere — product launches, earnings beats, supply-chain price hikes, shortage alerts. Hot sectors dominate financial media headlines. When investors collect the signals their framework requires, they unconsciously over-sample hot information and under-sample cold. This is not necessarily conscious. It is a sampling problem.
Take the three-force model. The framework asks you to evaluate AI supply rigidity, policy regime transition, and global division restructuring simultaneously. But availability bias will push you toward over-weighting AI-related signals — because they appear in your information feed every day — while under-weighting policy and global signals, which don't.
In Q2 2025, TMT turnover concentration had already reached 45%. An analyst at a sell-side firm wrote in a report: "Trading concentration has not yet reached historical extremes; there is still room to rise." He didn't realize that of the ten research notes he'd read that morning, nine were bullish on electronics. The information that is easiest to see starts to look like the information that matters most. Three months later, concentration hit 49.8%, and the market entered violent turbulence.
Bias Two: Confirmation Bias — Distorting the Execution
Confirmation Bias
People tend to seek information that supports their existing views, while ignoring or discounting information that contradicts them.
If availability bias distorts the input, confirmation bias distorts the execution.
A framework asks investors to evaluate each signal objectively. But confirmation bias makes you pre-load a position — when you're already bullish on a direction, you automatically read positive signals as "framework validation" and negative signals as "short-term noise." When bearish, the interpretation reverses.
The same set of three-layer signals — supply, capital, earnings — can lead a bull and a bear to completely different conclusions. Not because the framework is broken, but because the people running it are shaped by confirmation bias.
In the age of algorithmic recommendation, this gets worse. The information investors see is increasingly what they already agree with. Over time, a feedback loop develops — belief shapes selective information intake, selective information strengthens belief. Your "framework" starts functioning like a mirror, reflecting back only what you want to see.
Bias Three: Herding — Distorting Independence
Herding
When most people are doing the same thing, it's hard for an individual to hold an independent view.
The value of an analytical framework is that it provides independent judgment. Herding systematically erodes that independence.
This is particularly important for professional investors. Being different from the consensus can create career risk even when it creates investment opportunity. A portfolio manager who underperforms alongside the market is usually easier to defend than one who underperforms because of an unconventional view. There is institutional value in being wrong with everyone else.
This is one reason crowded trades can persist longer than a purely rational framework would suggest. When a sector becomes consensus, institutions may continue increasing exposure not only because they believe the thesis, but because reducing exposure would create relative-performance risk.
In Q3 2025, fund allocation to electronics broke above 20%. Nearly every major brokerage research department published bullish reports. At that point, consensus itself becomes part of the investment environment. An analyst who raises concerns is no longer arguing only against the fundamentals — they are arguing against the professional consensus. The more people reach the same conclusion using the same evidence, the less independent that evidence becomes. And that is precisely when risk accumulates fastest.
III. How Biases Reinforce Each Other: A Four-Step Loop
These three biases don't operate in isolation. They interact, forming a self-reinforcing cycle.
Availability bias distorts input. Investors collecting signals over-attend to hot information and under-attend to cold. The framework's inputs are systematically skewed.
Confirmation bias distorts execution. Investors evaluating signals tend to confirm their existing judgment. The framework's objective evaluation process becomes a subjective filter.
Herding amplifies the result. When enough people are affected by the same biases, market prices diverge from fundamentals. The divergent price becomes "evidence" that more investors feed back into their frameworks — bias gets amplified further.
Emotional feedback locks in direction. When prices rise, investors become more optimistic, pushing prices higher; when prices fall, more pessimism drives prices lower. Bias spreads from the cognitive layer to the emotional layer, forming a positive feedback loop.
The key insight here: bias doesn't only affect the framework's user. It affects the "output environment" the framework operates in. When bias is large enough, market price itself becomes a distorted signal source — and when you fill your framework with distorted prices, the framework's output is distorted too. The error feeds back into the system.
This is why "mean reversion" always takes so long. A mispricing does not disappear simply because the fundamental argument is correct. Correcting cognitive bias requires new information, time, and a sufficient external shock to break the loop.
IV. Patching the Framework: Three Practical Principles
Patch One: The Three Contrarian Questions
When your framework produces a conclusion, force yourself to answer three questions:
Three Contrarian Questions
1. Has bias been amplified to an extreme? Check valuation percentile, trading concentration, institutional positioning. When these metrics sit at extreme levels simultaneously, the framework's output may already be contaminated by positioning.
2. Is the fundamental picture mispriced? Distinguish between "the asset is undervalued but fundamentals haven't deteriorated" and "the asset is undervalued because fundamentals are genuinely deteriorating." The former is a bias-amplified opportunity; the latter is rational pricing of a worsening reality. The distinction sounds obvious. In practice, it is one of the hardest judgments to make.
3. Does your time horizon allow you to wait? The correction of mispricing often takes a long time. A contrarian position can be fundamentally correct and still be impossible to hold. The question is not only whether the thesis is right, but whether the investment horizon allows the thesis enough time to work.
Patch Two: Probabilistic Thinking — Against the Illusion of Certainty
Frameworks easily produce an "illusion of certainty" — the structure is clean, the logic is self-consistent, and the user unconsciously concludes: "If the framework is right, my judgment must be right."
But a framework only raises the probability of a correct judgment. It doesn't guarantee it.
- Separate judgment from outcome. A good decision doesn't always produce a good result. A bad decision can produce a good result. Evaluating the quality of a decision solely by its immediate outcome encourages hindsight bias and bad process. The better question is whether the decision was reasonable given the information available at the time.
- Update probabilities continuously. When new information arrives, actively adjust your view rather than clinging to the original one. A useful test: can you name three strong reasons why your current view might be wrong? If not, you may already be in an echo chamber.
- Manage expectations, not certainty. The goal is not to eliminate surprises. It is to avoid being destroyed by them. A robust framework leaves room for outcomes that were not in the base case.
Patch Three: Self-Check List
Use alongside any analytical framework. Check each item:
- Am I only collecting signals that support my framework's conclusion? (confirmation bias)
- Is my signal sampling overly dependent on recent or vivid events? (availability bias)
- Am I more confident because "everyone is reaching the same conclusion"? (herding)
- Can I articulate three reasons against my framework's conclusion? (contrarian thinking)
- Is my judgment based on fundamental signals or on price action? (signal vs. noise)
When writing it, was I more motivated to make it persuasive — because AI has been hot and readers are paying attention — than to make it accurate? Is the conclusion here partly driven by a desire to seem useful? This kind of self-examination matters more than any checklist. A framework designed to protect investors from bias can itself become another object of belief. That is the blind spot.
Closing: The Framework Is a Starting Point, Not an Endpoint
The two frameworks from the previous articles tell you the drivers behind K-shaped divergence and the stage of the AI industry. They are good frameworks.
But a good framework is a starting point, not an endpoint.
A framework tells you what to look at, how to analyze, how to judge. It cannot guarantee that what you see is real, that your analysis is objective, or that your judgment is correct. Because the person using the framework will inevitably be shaped by availability bias, confirmation bias, and herding — biases that distort the framework's inputs, interfere with its execution, and amplify its errors.
Good research requires more than a good framework. It requires a mechanism for questioning the framework's own output. The practical answer is not to abandon structure, but to add a second layer around it: reverse questioning, probabilistic thinking, and systematic self-checks.
The purpose is not to make the framework perfect. It is to make it harder for ourselves to misuse it.
The goal is not to predict markets, but to understand them.
Understanding markets means understanding not only the economic forces that drive them, but also the psychological forces that drive the people who invest in them.
*Magic Econ · Common Sense Jack*
This article is based on public market data and behavioral finance research, intended to explore the psychological boundaries of analytical frameworks and the impact of cognitive bias on investment judgment. It is for educational exchange only and does not constitute investment advice.
If this article sharpened your thinking, follow Magic Econ
