Why Chart Analysis 'Works' — 4 Mechanisms Behind the Lines¶

For / Key Points

For: Investors curious about technical analysis but unconvinced by its theoretical foundations. Those who suspect chart reading might be closer to astrology than science.

Key Points:

Chart lines themselves have no predictive power — they work because participants' behavioral patterns repeat
Four mechanisms explain why: self-fulfilling prophecy, behavioral biases, institutional execution constraints, and information delay
Instead of memorizing pattern names, ask: "Will asymmetric order flow follow this shape?"

"It's a double bottom — time to buy." You've seen this in countless investment commentaries. But why should a shape on a chart predict future prices? Press for evidence and you usually get: "Because it worked before."

The question this article answers: Why do chart "patterns" repeat? The conclusion, upfront: Charts don't make you win. If you can read the "traces of others' trading behavior" left on charts slightly better than average, an edge might emerge.

Here are the four mechanisms that explain why.

A Thought Experiment¶

Why is looking at lines alone meaningless?

Imagine you're the only person in the world who looks at charts. You identify a "support line" — but nothing happens. No other participant recognizes it, so there's no reason for buy orders to cluster at that price.

Chart lines have no intrinsic power. Lines only gain meaning when many participants see the same line and act similarly.

So what happens when millions of participants watch the same moving averages and support levels? That's the real question.

What Charts Actually Record¶

What is encoded in a price chart?

Each candlestick on a chart is the composite result of all trades executed during that time period. Behind it lie expectations, fears, stop-losses, profit-taking, trend-following, institutional execution plans, and liquidity imbalances — the accumulated trades of participants with countless different motives.

The fundamentals of price formation are covered in "How Prices Form." The key takeaway: a stock price is a record of agreement between participants, not an objective reflection of corporate value.

If chart analysis works at all, what you're reading isn't lines — it's the repetition of supply-demand dynamics and behavior. The question is: why does behavior repeat? Four mechanisms provide the answer.

Four Mechanisms That Make Charts "Work"¶

Each mechanism is examined below.

Mechanism 1: Self-Fulfilling Prophecy¶

What happens when many people see the same line and act the same way?

The pattern didn't "predict correctly." It came true because everyone believed it and acted on it. That's the self-fulfilling prophecy mechanism.

Trace the causal chain. Many traders recognize the same price level as "support." As price approaches that level, people thinking "it should bounce here" place buy orders. The concentration of buy orders creates actual excess demand, and the price bounces.

The critical point: the support line had no "intrinsic force." The more widely technical analysis textbooks are read, the more participants recognize the same levels, and the stronger the prophecy's self-fulfilling power becomes¹.

However, self-fulfillment has limits. At levels that diverge sharply from fundamentals, large contrarian orders can break through the self-fulfilling barrier. Self-fulfillment alone doesn't mean "it works everywhere, always."

Are there behavioral patterns that repeat regardless of whether anyone looks at charts? The next mechanism answers this.

Mechanism 2: Behavioral Biases (Human Tendencies)¶

Why does the same "shape" appear in markets repeatedly?

Because human psychological tendencies don't change. As Kahneman & Tversky's Prospect Theory demonstrates, humans react roughly twice as strongly to losses as to equivalent gains². This asymmetry repeatedly manifests as "shapes" on charts.

Three typical patterns illustrate this.

Return selling ("relief selling"): An investor who bought at the high endures unrealized losses. The moment price recovers to their purchase price, relief triggers a sell order. Odean's research documented that investors sell winning stocks early while holding losers — the "disposition effect" that drives return selling pressure³
FOMO buying (Fear of Missing Out): Investors without positions see prices breaking to new highs and panic-buy. "If it keeps rising without me, I'll miss out" — this fear fuels breakout momentum
Capitulation exhaustion: When panic sellers during a crash have all exited, selling pressure drops sharply. Remaining buyers gain the advantage, and a reversal pattern forms

These patterns occur regardless of whether anyone is looking at charts. As long as loss aversion bias persists in human psychology, the same shapes will repeat.

Behavioral biases are most pronounced among individual investors, but institutional investors have their own distinct "source of repetition."

Mechanism 3: Institutional Execution Constraints¶

Why do "trends" emerge even when no one watches charts?

Buying $100 million worth of stock at once would spike the price with your own order. So institutional investors split purchases over many days. This physical constraint is one source of "trends."

Almgren & Chriss's optimal execution theory showed that time-spreading is optimal for minimizing market impact of large orders⁴. For example, splitting $100 million evenly over six days means roughly $17 million of sustained buying pressure each day. On a chart, this appears as a "gradual uptrend."

The reverse holds equally. When large sellers unwind positions, selling pressure continues over days to weeks.

The key characteristic of this mechanism: it exists completely independently of technical analysis. Even if no one looked at charts, trends would structurally emerge from institutional block trading.

One more structure arises independently of charts: the fact that information doesn't reach everyone simultaneously.

Mechanism 4: Information Spreads Slowly¶

Why doesn't a good earnings report immediately reflect in the stock price?

Because not everyone reacts at the same time. Say a company reports strong earnings.

Day 0-1: Professional analysts and algorithms react immediately. They have the capability and speed to parse earnings data
Day 3-5: Retail investors react after receiving brokerage reports or reading investment blogs
Day 7-14: The slowest group learns from TV news or social media buzz and enters the market

Hong & Stein's model showed that the non-uniform diffusion of information among market participants provides the theoretical foundation for momentum — the persistence of price direction⁵. Jegadeesh & Titman empirically confirmed that a strategy of buying past 3-12 month winners and selling losers generates statistically significant returns⁶.

This time lag contradicts the efficient market hypothesis's assumption that "information is instantly reflected in prices." As long as information diffusion speeds differ, momentum has structural room to persist.

"If Everyone Knows This, Can Anyone Still Win?"¶

Would the four mechanisms be arbitraged away once widely known?

This question is half right and half wrong.

The right half: Overly famous patterns are easily arbitraged. Simple rules like "buy on a golden cross" often lose their statistical edge when transaction costs are included. Algorithmic trading has shortened the lifespan of simple patterns.

The wrong half: Knowing something and executing on it are different. Even if everyone knows Prospect Theory, few remain calm when holding unrealized losses. Institutional block execution is a physical constraint that "knowledge" doesn't solve. The structural gap in information diffusion speed can't be bridged by individual knowledge.

Easily arbitraged	Hard to arbitrage
Simple trading rules (crossover-based)	Human psychological biases
Patterns anyone can find via backtesting	Physical institutional execution constraints
Low-cost strategies easy to replicate	Structural differences in information speed

As long as differences in capital size, time horizon, stop-loss discipline, and execution capability exist, biases won't completely vanish.

The Real Question to Ask When Reading Charts¶

What should you ask instead of memorizing pattern names?

"Will asymmetric order flow follow this shape?" This is the question to ask when looking at charts.

Memorizing pattern names and deciding "the shape appeared, so buy" skips the causal chain. The right approach is to reinterpret patterns as "hypotheses about the order flow they predict."

Breakout = Hypothesis that new highs will trigger trend-following buys + short covering
Double bottom = Hypothesis that capitulation selling is exhausted and return selling is absorbable
Symmetrical triangle = Hypothesis that converging sell/buy pressure will release in one direction

And edge isn't determined by win rate alone. Evaluate by gain per win x loss per loss x frequency x costs. A 60% win rate loses money overall if each loss is three times the average gain.

Chart analysis works when	Chart analysis fails when
Subsequent order flow is biased	It's just after-the-fact interpretation
Edge survives after costs	Pattern definition is vague
Participant behavioral biases repeat	The regime (market environment) shifts

Summary¶

Chart lines have no magic. They work because participants' behavioral patterns behind the lines repeat
Four mechanisms — self-fulfilling prophecy, behavioral biases, institutional constraints, information delay — generate the structure of repetition
Use patterns as "hypotheses about predicted order flow" and evaluate by expected value, not win rate

One more point worth adding: the strength of these four mechanisms varies by market regime. In low-volatility calm markets versus high-volatility crisis periods, both the manifestation of behavioral biases and institutional execution methods change. Recognizing when chart patterns "break" is as important as knowing how to use them.

De Long, J. B., Shleifer, A., Summers, L. H., & Waldmann, R. J. (1990). "Noise Trader Risk in Financial Markets." Journal of Political Economy, 98(4), 703-738. A foundational paper theorizing self-fulfilling price dynamics in markets. See also Shiller, R. J. (2000). Irrational Exuberance. Princeton University Press. ↩
Kahneman, D., & Tversky, A. (1979). "Prospect Theory: An Analysis of Decision under Risk." Econometrica, 47(2), 263-291. Demonstrated that losses have roughly twice the psychological impact of equivalent gains. ↩
Odean, T. (1998). "Are Investors Reluctant to Realize Their Losses?" The Journal of Finance, 53(5), 1775-1798. Documented the "disposition effect" — investors' tendency to sell winners early and hold losers — using large-scale brokerage data. ↩
Almgren, R., & Chriss, N. (2000). "Optimal Execution of Portfolio Transactions." Journal of Risk, 3(2), 5-39. Formalized the optimal time-spreading strategy for large order execution. ↩
Hong, H., & Stein, J. C. (1999). "A Unified Theory of Underreaction, Momentum Trading, and Overreaction in Asset Markets." The Journal of Finance, 54(6), 2143-2184. Theorized how gradual information diffusion generates momentum. ↩
Jegadeesh, N., & Titman, S. (1993). "Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency." The Journal of Finance, 48(1), 65-91. The classic empirical study demonstrating momentum strategy effectiveness. ↩